Category Archives: Games

11×11 Shogi, Part I: Shinjuu Shogi

If you’ve read this blog in recent years, you’ll be well aware that I consider Shogi one of the finest games ever created, surpassed only by its larger cousin Chu Shogi, which I believe to be the greatest traditional Chess-like game on the planet.  As I’ve been working with Stephen Tavener and his Ai Ai software to slowly and laboriously bring Shogi and its modern variants to life as well as the fascinating historical variants, I’ve been inspired to build on what I’ve learned about the historical Shogi variants and create a modern variant that brings the unique flavour of the ancient large Shogi games into the 21st century.

In this article I’ll talk a bit about the process of creating this variant, my design goals for the game, and my own evaluation of what makes a Shogi game feel like a Shogi game.  As this variant has evolved through constant, endless playtesting and experimentation, I feel I’ve learned a lot about what makes the Shogi family work so well, so I decided to share this here to generate discussion and hopefully help others who may be experimenting with Shogi variants of their own.

The Birth of Shinjuu Shogi

While testing the various historical Shogi variants in Ai Ai, I spent quite a bit of time with Wa Shogi.  Wa Shogi is an 11×11 game with 27 pieces in each player’s army, and like Tori Shogi it uses a completely different set of pieces from the normal Shogi assortment, including different Pawns and Kings (though with the same moves).  Wa Shogi is an excellent game in its modern form, in which it’s played with drops, but I felt the game lacks a certain intensity compared to the other historical variants, because the pieces are generally weak:

Wa Shogi pieces

While there are a few pieces with ranged moves in Wa Shogi, the majority of the pieces have short-range stepping moves only.  That means the game works very well with drops as in modern Shogi, but it also gives the game a slower, less intense tactical experience than the other historical variants, which are generally pretty packed with powerful pieces.  I sometimes felt that tactical fights in Wa Shogi were too localised, and the more powerful pieces could mostly just avoid the masses of much weaker stepping pieces.  In perusing the few recorded games of Wa Shogi between strong players, I was also surprised to find that most of them were shorter than a typical game of 9×9 modern Shogi (the average tournament Shogi game lasts about 120 moves).

That inspired me to try to make my own take on an 11×11 modern Shogi variant, but this time incorporating a range of pieces from the large historical variants.  My goal was to generally increase the power level of the pieces, allowing for sharp tactical play, but to balance the piece array so that the games would also be a bit longer than Wa Shogi.  Ideally, the game would serve a dual purpose as both a fun game in its own right, and an enticing preview of the creative and unusual piece types present in the old Shogi variants; with luck, perhaps players of this new variant might be inspired to take a look at the games that originally housed these interesting pieces.

I felt the best way to achieve the second goal would be to capture some of the interesting cultural elements that exist in the old Shogi variants, so I decided to centre the game’s armies on four mythological creatures we find on the board in Tai Shogi: the Blue Dragon, White Tiger, Vermillion Sparrow and Turtle-Snake.  In Japan these beings are called Divine Beasts, or Shinjuu, and hence Shinjuu Shogi was born.

Design Goals

The dual design goals of developing an enjoyable modern Shogi variant and an inspiring selection of ancient Shogi pieces meant that I’d need to break down the essential properties that make for a good Shogi game, and the aspects of the historical variants that could feasibly be transferred into a modern Shogi framework.  In my view the key aspects that make Shogi so compelling are these:

Shogi pieces
  1. Asymmetric pieces biased toward forward movement — If we take a look at the piece types present in the starting array of modern Shogi, we can see that only the Bishop and Rook have symmetric moves.  The Pawn, Lance, Silver General, Gold General, and Knight are all asymmetric, and most or all of their movements are pointed directly forward.  This aspect contributes to the generally aggressive feel of the game, because retreat is often not really an option.
  2. Slow opening and aggressive endgame — Part of the appeal of Shogi stems from the distinctive way in which play evolves over the course of the game.  In Chess, the opening is very sharp, because the board is small and packed with long-range pieces that can start causing trouble instantly.  In contrast, endings tend to be intricate, strategic affairs requiring intense calculation due to the lack of mating material left on the board.  In Shogi, however, openings tend to be slower and more flexible; players aim for advantageous middlegame positions, but due to the size of the board and the slower pieces, following precise move orders is less critical than in Chess.  Meanwhile, Shogi endings are savage and explosive — both players have most or all of their pieces still on the board, thanks to the drop rule, so the end of a Shogi game is typically a tense race to checkmate as both players mount vicious attacks with drops and promoted pieces.
  3. Minor and major piece promotion — In Shogi every piece except the King and Gold General is able to promote upon reaching the enemy camp, unlike in Chess where only the Pawns promote.  Promotions happen frequently, given that reaching any of the opponent’s three ranks allows a promotion, and pieces can be placed in the promotion zone when dropped.  The ability of nearly every piece to power itself up again increases the intensity of Shogi tactics, and also complicates the concept of material value in the game as compared to Chess.  Promoted pieces hold extra value because if your opponent captures them they get a weaker piece instead.  We see this reflected in Shogi proverbs like ‘a central Tokin [promoted Pawn] cannot lose’.  In other words, the Tokin is a great way to hold the centre of the board, since it controls a number of squares, but when taken it only gives your opponent a Pawn.
  4. Promotion can be deferred — Sometimes, a Shogi player will defer promotion of a piece, because not all promotions are directly upwards-compatible with the original piece; in other words, the promoted piece may not include the moves of its ancestor in its arsenal.  For example, a promotable Knight may be left as-is even if its eligible for promotion, because the owner needs the Knight’s aerial threat, which goes away if the Knight promotes to Gold General.  In fact, deferring Knight promotion is so commonplace that there’s another Shogi proverb about that, which simply states ‘Don’t promote the Knight!’

So, in order to make a fun modern Shogi variant that would capture and retain the interest of Shogi players, I felt I would need to capture all of these elements in some fashion in Shinjuu Shogi.  However, the ancient Shogi variants (other than Wa Shogi) are all played without drops, and simply adopting the drop rule transforms those games into endless slogs; the many powerful pieces keep returning to the board and building powerful defensive castles around their King, making checkmate exceedingly hard to achieve.  That meant I’d need to think about the fundamental aspects of large Shogi that give these ancient games their distinctive character, and try to work out which of those could be transferred intact to the modern Shogi battlefield:

  1. Lots of powerful pieces — As we can see in the summary of the Chu Shogi army above, the initial position of the game is filled with powerful attackers.  The Rook and Bishop of 9×9 Shogi are present, but this time in pairs, along with pairs of their promoted forms, and multiple other pieces that are substantially more powerful still.  The same dynamic persists in the other large variants, with each larger game bringing in even more powerful pieces.  These power pieces help maintain tactical excitement in these games even when playing on these gigantic battlefields.
  2. Thematic richness — Many of the larger Shogi variants incorporate pieces linked to classical Japanese and Chinese mythology.  Alongside the Divine Beasts, we find pieces like the White Elephant (bringing to mind the white elephant the Buddha’s mother saw in a dream before his birth), the Lion Dog (the creature that defends many temples throughout Japan), the Teaching King (thought to symbolise the Lotus Sutra) and many more.  This thematic element and the tendency to link powerful pieces to mythological or religious figures immerses the players even more in the battle unfolding on the board.
  3. Even more asymmetry — While modern Shogi does have an asymmetric starting position, due to the Rook and Bishop only being present as singletons rather than pairs like the other pieces, the large variants take this much further.  Wa Shogi is the most extreme example, where only the Sparrow Pawns are present in multiples in the starting position; Dai Dai Shogi is similar, with most of the starting position’s 64 piece types existing as singletons, leading to a heavily asymmetric setup.  This asymmetry not only adds tactical variety and intricacy, it also lends a certain additional Japanese-ness to the game; asymmetry is highly valued in traditional Japanese arts.
  4. Different approaches to promotion — In modern Shogi all pieces promote to Gold General, except for the Rook and Bishop.  In the ancient variants, we see much more varied takes on promotion.  For example, we tend to see weaker pieces gain very significantly in strength when promoted, as a reward for marching them all the way across these huge boards.  In a few cases we have pieces that actually demote, pushing players to maximise the utility of the piece’s original form.  We also see pieces that are very direct enhancements of the original piece, like in Maka Dai Dai Shogi where single-step pieces promote to ‘free’ versions of themselves that may slide unlimited squares in the same directions their original form could step.  Maka Dai Dai also has the unusual Buddhist Spirit and Teaching King, which if captured transform the capturing piece into a copy of themselves, ensuring that they hardly ever leave the board permanently!

In order to bring the ancient Shogi feel to the modern Shogi battlefield, I’d need to somehow combine these different properties of these two sub-genres of the Shogi family.  However, I was unsure whether some of these properties could be reconciled at all: could I increase the power levels of all the pieces without unbalancing the game or trivialising Shogi’s rich tactics?  Would it be feasible to adopt different approaches to promotion in a drops-based game, or would Shogi’s explosive endgames be too quickly curtailed by the opportunity to generate strong promotions shortly after a drop?

The only way to judge these points would be through experimentation.  So I began constructing possible initial positions for the game, with the four Shinjuu pieces taking up flank positions and the centre providing a home for strong pieces that could both mount damaging attacks and defend the King when called upon to do so.  Then I embarked on a long, meticulous testing process, investigating dozens of possible starting arrangements and playtesting over and over again, discarding piece selections and positions that failed to produce appropriately modern-Shogi-like gameplay.

Thankfully I’m going to spare you all the details of that process, and instead show you the result!  What you’ll see below is the 62nd version of Shinjuu Shogi (!), but I’m finally at a point where I’m ready to release it soon.

Meet Shinjuu Shogi

Let’s be clear on this from the start — Shinjuu Shogi is a big game.  The board is an 11×11 grid with 121 squares, substantially bigger than modern Shogi’s 81-square board.  Players start with 27 pieces in their armies, nearly all of which can promote, giving us 28 different moves to remember — the same number as in 12×12 Chu Shogi.  Wa Shogi is also an 11×11 game and has 25 moves to remember; the additional moves exist here thanks to the presence of the Dogs in addition to the normal Pawns, and the two additional pieces in the starting array.

The selection of piece types references most of the historical variants:

  • The Pawn, Gold General and the King are present across all variants (except Tori and Wa).
  • The Phoenix, Kirin, Ferocious Leopard are everywhere in the large variants, but for me they mostly bring to mind Chu Shogi.
  • The Divine Beasts — Blue Dragon, White Tiger, Vermillion Sparrow and Turtle-Snake — come to us from Tai Shogi, but here we use the Taikyoku Shogi moves for the Blue Dragon and White Tiger, and the Tai Shogi moves for the Vermillion Sparrow and Turtle-Snake.  The White Tiger and Blue Dragon promotions also come from Taikyoku Shogi.
  • The White Elephant and Fragrant Elephant come from Dai Dai Shogi, and their moves are vertically mirrored versions of each other.
  • The Dog comes to us from Tenjiku Shogi, as well as its promotion to Multi General.
  • The Fierce Eagle and Old Kite join us from Tai Shogi, with their moves drawn from the Japanese sources rather than the move promoted in The Shogi Association rules.
  • The Mountain Witch and Wizard Stork join us from Dai Dai/Maka Dai Dai, as the promotions for the Vermillion Sparrow and Turtle-Snake, respectively.
  • Various pieces and promotions come from Taikyoku Shogi: Great Standard; Walking Heron (promoted Fierce Eagle); Bird of Paradise (promoted Old Kite); Wooden Dove (promoted Phoenix); Free Boar (promoted Gold General); and Copper Elephant (promoted Ferocious Leopard).
  • The Golden Bird, the strongest promotion in the game, originates in Dai Dai/Maka Dai Dai but uses the Taikyoku move, which gives it the ability to leap over up to three pieces while sliding along the forward diagonals.

Shinjuu Shogi therefore contains references to most of the large Shogi variants, with the notable exception of Ko Shogi.  Ko Shogi is a fascinating game with incredibly creative and complex piece movements, but the most interesting pieces in that game are simply too powerful to have a place in a game the size of Shinjuu.  Similarly, the infamous Fire Demons and jumping Generals of Tenjiku Shogi can’t appear here either, as they would dominate the board completely.

I’ve built this game to try to capture the quirky character of the large Shogi variants:

  • Asymmetry — The starting position is asymmetrical, with the Blue Dragon, Turtle Snake, and Old Kite starting on the left side, and the White Tiger, Vermillion Sparrow and Fierce Eagle on the right.  The pieces themselves also have asymmetries in their movements, including both the step-moving pieces like the Old Kite and Fierce Eagle and the various sliding pieces like the Divine Beasts and the Elephants.
  • Promotions — The Pawns promote to Tokin (Gold General), which is a nearly universal truth across the whole Shogi family.  The other promotions here are specifically inspired by the large Shogi variants.  The short-range pieces tend to gain significant power on promotion, graduating from single-step movers to long-range attackers.  The Dog becomes a sliding piece that slides in the same directions it could step.  The four Divine Beasts are straight upgrades of the unpromoted pieces, gaining additional sliding directions and retaining their previous moves as well.
  • Promotion quirks — The Fragrant Elephant and White Elephant are vertically mirrored versions of each other, and their promotions are both strong but have a trade-off: in each case, the promoted form gains substantial power but loses their original two sliding moves, meaning that in some cases you may want to defer promotion.  Similarly, the Kirin and Phoenix gain enormous mobility on promotion, but their jumping moves change.  The Old Kite and Fierce Eagle gain vertical slides and some new options, but the Old Kite loses horizontal movement while the Fierce Eagle loses the rear diagonals.  All these quirks mean that promotions sometimes must be deferred to make use of the unpromoted piece’s subtle differences.
  • Powerful pieces — The amount of attacking power on the board is massively higher than in Wa Shogi.  The starting array includes multiple-square step-movers, several different sliding pieces, and two leapers.  The promoted pieces add some powerful threats: the Wooden Dove is able to leap and then slide afterward, providing dangerous long-distance attacking potential; the Golden Bird is able to leap over up to three pieces in certain directions, enabling deadly cross-board strikes; and the Great Dragon has a sideways 2- or 3-square leap that is lethal in certain close-quarters situations.
  • Pacing — Piece placement has been optimised to generate a pace of play that follows the modern Shogi model.  The opening position offers 39 possible first moves, which will hopefully reduce the likelihood of very deep opening analysis developing.  All Pawns are protected at the start of the game, preventing any super-early tactical shots, and the longer-range pieces are tucked behind the Pawns and thus require additional moves to be developed.  The result is a relatively deliberate and strategic opening phase leading up to a very active middlegame.  The middlegame itself is more intricate than Wa Shogi, given the higher piece density and greater piece mobility, which allow for complex and varied tactics to emerge.  The endgame is characterised by dynamic and aggressive play, with frequent drops supported by powerful long-range attacks from promoted pieces.
  • Thematic appeal — The Divine Beasts are the thematic core of the game, and the other pieces have been chosen to amplify the feel of an epic battle between mythological forces.  The Dogs, Old Kites and Fierce Eagles are the vanguards, leading these armies of beasts and gods into battle.  The King is protected by a Great Standard, which effectively guards the second rank and has strong forward attacking potential when given the opportunity.  The Phoenix and Kirin (Unicorn) provide nimble leaping attacks that can complicate enemy attempts to secure their King.  The Fragrant and White Elephants are stalwart defenders that blossom into vicious airborne attackers when given an opening; their promotions are thematic, with the White Elephant of Buddhist mythology transforming into a Golden Bird symbolising enlightenment, while the Fragrant Elephant matures into a Great Elephant, losing some forward attacking power in exchange for strong rearward sliding capabilities.

The result of all this is, I hope, a game that combines the dynamism of modern Shogi with the creative character of the ancient large Shogi variants.  To give it a distinct personality I included pieces with unusual moves and abilities:

These kinds of unique moves are characteristic of these old games.  Perhaps the most iconic piece of the large Shogi variants is the powerful Lion that can move and capture twice in a turn, but the Lion and its many cousins are simply too strong to live happily in a drops-based game.  So instead I selected pieces with unusual sliding modes from Tai and Taikyoku Shogi, which are still strong and unique but not overpowering.  These pieces also add further dynamism to the game by providing moves that neatly ignore dense defensive formations, forcing players to adopt flexible defences.  I found that having a few leapers in this game also helped counterbalance the much higher attack density, and prevented games from dragging on for too long due to an overabundance of defensive power.

I should note that I made a possibly controversial decision with this game to have nearly all pieces promote to a piece type that is not present in the initial position. Pawns promote to Tokin, which moves identically to the Gold General, but otherwise all pieces upgrade to something new. This was a conscious choice, which I made to ensure that Shinjuu would maximise the opportunity to showcase a wide variety of different pieces from the ancient variants. This does mean players need to familiarise themselves with more moves, but I tried to choose promotions that have some vaguely logical connection to their unpromoted forms, particularly for the shinjuu themselves. As you’ll see in future posts, some of my other 11×11 designs do have more pieces promote to something present in the initial array.

In my defence, many Shogi variants have promotions that don’t appear in the initial position, including Shogi itself (the Dragon King/Dragon Horse); here I’ve just extended that to its logical extreme!

 

Gameplay

Shinjuu Shogi is a big game, much bigger than standard Shogi, but much like Chu Shogi and its larger cousins, a few games will quickly solidify the piece movements in your mind.  The major conceptual obstacle to overcome is the substantially increased asymmetry in the moves compared to standard Shogi; many pieces in this game have odd mixes of stepping, jumping and sliding moves.  To play this game well, players must master all these distinct movement modes for these pieces, and this can be a challenge at first.

I was careful to select pieces that all have weaknesses in certain directions; we have no equivalent of the Queen here.  Many pieces in Shinjuu cannot move in some directions at all, and even the strongest pieces have some directions with limited movement capability, which opens them up to long-range attack.  These weaknesses mean that pieces generally have to be used in concert for maximum effect, and no one piece is able to fully dominate the board.  In addition, these weak spots open up opportunities for fun tactics — a clever drop in an opposing piece’s blind spot can be an effective way to regain the initiative. 

The general feel of the game is high intensity, and very dense.  Despite the piece density being about the same as modern Shogi — in modern Shogi 40 of the 81 squares are occupied at the start, and here 58 of 121 are occupied — the game feels more dense, perhaps because the sheer number of pieces creates the impression of being constantly surrounded by threats.  However, in practice the game board offers plenty of space for manoeuvre, and rarely will a player struggle to find a way to improve the positions of their pieces.  In addition, once players hold a few pieces in hand the number of possible moves each turn escalates massively (over 500 possible moves is not uncommon!).  This means that strategic and positional thinking is critical, because in a complex middlegame position it will be impossible to calculate out all the possible tactical variations your opponent may choose.  Players have to survive by developing sound positional fundamentals and maintaining whole-board vision.

Unfortunately, I’ve yet to have the opportunity to play this game against strong opposition, but in my test games so far I’ve found a typical game will last between 200-300 moves.  This is substantially more than Shogi and Wa Shogi, but on a similar level to Chu Shogi.  This game length is right around the 250 move mark that I was aiming for throughout the design process.  The typical game of Go is about 250 moves long, and something like ~50 million people play Go on this planet, so I felt that was a good target to aim for that would satiate my thirst for long, involved games but would not feel particularly excessive.  On rare occasions I had games just over 400 moves long, but in subsequent analysis I found they could’ve ended much sooner with some sharper play.

On the whole I’m sure opinions will vary on this game and my design principles, but so far I enjoy the result.  The game is big, complex, and jammed full of possibilities, but it also feels like Shogi and never slows to a crawl.  The larger board and longer games create an expansive feel, and the thematic pieces drawn from ancient Shogi help create a quirky narrative feel in the game.  Realistically I suspect very few people will ever play this game, but if nothing else I feel I’ve made progress on my design goals and have learned a lot in the process.

The Pieces of Shinjuu Shogi

Shinjuu Shogi is a big game, and given that I want it to be a showcase for unusual historical Shogi pieces it contains few pieces that will be familiar to modern Shogi players (just the King, Pawns and Gold Generals).  That means there’s a bit of a learning curve, so I’ve put together a quick reference sheet, where each piece is paired with its promoted form:

A PDF version for printing or reference is available here.  Note that the top of the sheet also has some reminders of the pieces with special moves, which are various types of leaping moves.

The four shinjuu themselves are perhaps the hardest pieces to remember initially.  However, it may help to realise that the moves of the Turtle Snake and Vermillion Sparrow are related — just take the forward slide and backward step of the Turtle Snake and invert them, and you have the Sparrow!  Likewise, the White Tiger has the same move as the Blue Dragon, just rotated 90 degrees counterclockwise, and the same is true of their promoted forms.  The Turtle Snake and Vermillion Sparrow promotions are easy to remember too, as they both simply replace the backward diagonal steps with unlimited slides, while keeping the rest of their unpromoted moves the same. Similarly, the promoted White Tiger and Blue Dragon replace one two-step move with an unlimited slide.

Players familiar with large Shogi will notice that some of the promoted pieces here are quite powerful, in particular the Golden Bird and its ability to leap over up to 3 pieces while sliding diagonally forward.  However, promotions are more difficult to achieve here than in the ancient games, thanks to drops; throughout the game the number of pieces on the board remains largely constant, so any piece attempting to promote is likely to encounter a heavily-guarded promotion zone.  Attackers should be more cautious about pushing for promotion here, and spend time setting up substantial weaknesses in the opponent’s defensive formations beforehand. 

The generally stronger promotions mean that minor pieces are more valuable in Shinjuu than in standard Shogi.  In particular the Gold Generals gain six unlimited sliding moves when they promote, but they’re also very useful for defending one’s King so deciding when or if to promote them is often challenging.  Promoting a Gold General commits you to the attack, because it can only retreat one square. Meanwhile, the Kirin and Phoenix are interesting to use in this game, as their unpromoted moves include very useful leaps. Their promoted forms are dangerous aerial attackers, given they can leap 3 squares over defending pieces.  Beware of these pieces when defending; the King will need escape squares, or you may end up in a smothered mate situation.

Of the major pieces, the Fragrant Elephant and White Elephant are perhaps the most dangerous.  Their standard moves are strong already, but in promoted form they become highly mobile as well.  Players must always be aware of any Elephants attempting to promote, and in the endgame having a couple of Pawns in hand to buttress the King’s castle against airborne Golden Bird threats can be helpful.  The Great Standard is very strong but also has limited retreating moves, plus it is one of the few pieces that can slide sideways and is therefore useful for protecting the promotion zone.  The White Dragon can also serve well in this role, while the Blue Dragon can guard against edge attacks and support advancing Pawns.  The Turtle Snake and Vermillion Sparrow are perhaps the most active attacking major pieces, as their forward-facing slides allow them to generate a variety of long-range threats, and in promoted form they can retreat very effectively too.  The shinjuu have fairly mild promotions in general, so players may be more willing to use them for daring sacrifices; in contrast, handing your opponent extra Elephants can be seriously risky!   

At this early stage in the game’s life it’s difficult to judge the best approach to defence.  One’s instinct may be to surround the King with a thick wall of defensive pieces, but in practice this can be very dangerous.  Leaping pieces can bully the King around even from outside your defensive structure, and handing the initiative to your opponent without generating counterplay may allow them to dominate elsewhere on the board and develop opportunities for deadly promotions.  My approach has been to shift the King to one of the flanks of the board, keeping him guarded by the Leopards, Golds, White Tiger and Great Standard.  My frontline attackers have been the Turtle Snake, Vermillion Sparrow, and Fragrant Elephant, with the Kirins and Phoenixes providing support.  

On the whole, the pieces in this game have unusual moves, and don’t fit cleanly into our pre-existing notions of what we might do with more typical Chess or Shogi pieces.  I’ve found it enjoyable to try and understand the best use of these pieces, as each has a distinct personality and clear weaknesses that must be taken into account.  Over time I hope to deepen my understanding of how to use these pieces effectively.

A Closer Look

Let’s take a moment to examine the last portion of one of the sample games shown in the GIFs above. This game was played against the AI (120 seconds/move thinking time), and playing as Sente I won after 181 moves:

Shinjuu Shogi sample game (181 moves).

Bear in mind I’ve played this game a lot, so I’m quite familiar with the AI’s foibles, and I’d emphasise that Gote’s defence is not optimal. Nevertheless, hopefully this brief discussion of a sample endgame can give you a bit more of a idea of how the game plays.

In this game I decided to press home my advantage and play for the win around move 139:

Until this point I’d been focused on securing the centre of the board, with an eye toward prying open Gote’s position on the left side. Gote’s right flank has the two Kirins and several Dogs providing additional protection, so the more lightly-defended left offered a bit more of an opportunity. I’d also developed a strong material advantage, with six pieces in hand ready for dropping.

I started my plan by pushing the freshly-dropped Pawn on d6 forward, while Gote responded by advancing pieces out of the right flank to pressure me from the other side:

On move 145 my Pawn has promoted to Tokin, putting pressure on Gote to respond. Gote elects to exchange the Vermillion Sparrows, so I recaptured with my Fragrant Elephant, which Gote then took with their White Tiger. Rather than take that, I elected to press the Tokin forward, punching a hole in Gote’s defensive formation around their King:

Here the AI elects to press forward with its White Tiger, threatening to fatally weaken my Pawn line in the centre. I prevent this by taking with the Turtle Snake, and then the AI attempts a distraction by dropping a Fragrant Elephant near my King. I ignore this and take the Ferocious Leopard with my Tokin, then sacrifice my Turtle Snake in exchange for the Gold General, creating a serious weakness in Gote’s defences:

The White Tiger now pins the Phoenix to Gote’s King, and it’s time to start hurling my massive material advantage into the meat grinder. I do have to be cautious, because Gote does hold several good defenders in hand, and if I allow Gote any room to breathe they could still cause trouble around my lightly-defended King.

Gote proceeds to attempt exactly that, taking my Gold General with their Fragrant Elephant and forcing me to retake, then dropping a Turtle Snake in prime position in the centre. However, this is fine for me, as I can take Gote’s Phoenix with my White Tiger, forcing a recapture with the Kirin and opening a diagonal toward Gote’s King. Rather than exploit this immediately I drop a White Elephant next to the Phoenix, which can’t attack orthogonally adjacent pieces, and promote to a deadly Golden Bird:

This forces Gote to drop a Gold to block, and then I can drop a Vermillion Sparrow on that key diagonal. Gote is now firmly on the back foot, with only two droppable pieces left, and it’s time for me to end the game in a hail of drops.

There’s no way out for Gote, and their King ends up checkmated by a deadly combo of Mountain Witch and Great Elephant:

Granted, I’m no Shogi master, and the AI is struggling on without a specialised evaluation function for this game, but hopefully you can see that Shinjuu offers interesting checkmating sequences that feel very Shogi-like. Despite the larger board, bigger armies and wackier pieces, the general strategic themes feel very reminiscent of the modern game. Battles for central control, prying open weaknesses in the position with drops, and building up pieces in hand for a final assault — all these aspects of Shogi play are still present.

The moment-to-moment tactics do feel distinct though, as the pieces are generally quite powerful, so choosing the right moment to attack is crucial. Pieces like the Elephants can protect a large number of squares, so using your leaping pieces and strong promoted pieces is important to get around solid defences.

The other sample game has an example of this aspect. In that game, a crucial moment happens on move 250:

Here Sente has had its King forced into the centre, but it’s completely cocooned by a deep fortress of strong defenders. However, Gote drops a Phoenix and promotes to Wooden Dove, which puts the King in check via its 3-square leaping slide! This shows us the danger of having a dense fortress — the King can be forced out via leaping attacks, and in this case if the King hadn’t had this one escape square, this move would have been checkmate.

From there, Gote launches a vicious attack that completely bypasses Gote’s massive central castle, and Sente’s desperate counterattack fizzles due to Gote’s more flexible defensive structure on the back ranks. As Sente’s offensive splutters and dies, Gote seizes the moment and forces an unusual checkmate with Blue Dragon, Great Dragon and Fierce Eagle:

So, while some basic Shogi principles do transfer over, the presence of many strong pieces and leaping attackers changes attack and defence significantly. While Sente could have defended better here, that check on move 250 was a nice refutation of Sente’s attempt at King safety, and once the King was isolated on the back ranks it was difficult to defend against Gote’s assaults. Gote, in contrast, kept the King out of the centre with a looser defensive formation, enabling them to absorb counterattacks well and maintain more escape squares to protect against leaping attacks.

In any case, I hope that some Shogi fans out there might be convinced to give this game a try, as I’d love to see what strong players could do with this game. On more than one occasion I’ve had very exciting endgame attacks in this game, where the path to victory was a spectacular sequence of 20+ drops with both Kings getting pushed all over the board. A tense battle between two strong players could provide quite a spectacle!

Lessons Learned

When I started working on this game, I hoped that I’d have a chance to really test some of the standard assumptions one encounters in discussions about Shogi variants.  There are a couple of widely repeated concerns I’ve seen relating to drops-based variants:

  1. Games with drops should have weak pieces — This point is often justified by pointing at Shogi, Tori Shogi and Wa Shogi, all of which are excellent games and all of which have generally weak pieces.  However, a few other examples of larger drops-based games like Okisaki Shogi and Futashikana Shogi seem to show that more powerful pieces don’t break the game, and of course Crazyhouse uses the strong Chess army on the tiny 8×8 board and is still great fun to play (though White has a nice advantage).
  2. Drops-based games won’t work on large boards — I’ve seen this brought up often in passing; people will say things like ‘imagine playing 12×12 Chu Shogi with drops, that would be preposterous!’  Of course they’re completely right about this, but I wasn’t sure that the general point follows; after all, Chu Shogi is completely crammed with super-powered pieces and the start position is much higher density than modern Shogi (66.7% vs 49.4%).

Having spent ridiculous amounts of time on Shinjuu Shogi at this point, and a number of related drops-based variations during testing, I’ve convinced myself that both these issues are not as bad as they seem and the pitfalls they identify can be avoided.

Firstly, powerful pieces seem quite at home in drops-based games, so long as they have weaknesses that prevent them from dominating the board.  In Shinjuu, I accomplished this by using pieces with clear blind spots.  The selection of weaker pieces helps here, and ideally some of those pieces should provide an effective countermeasure to the strong pieces. Okisaki and Futashikana Shogi replace the Shogi Knight with the strong Chess Knight, which I think works well because the Knight can attack the Queen from squares it can’t see (and conversely, Knight + Queen is a potent attacking combination).

As far as large boards go, again I found this to be a matter of balance and piece density.  The general approach in large drops-based variants seems to be to lower the density relative to modern Shogi, but I’ve found it equally playable to keep the piece density around 50%.  Having done extensive testing on 11×11 with anywhere from 23 to 33 pieces per player, I found that lower densities tended to lead to less exciting tactics and somewhat shorter games, while higher densities were packed with tactical shots but were either overly long (due to an overabundance of defensive resources) or overly short (due to an overabundance of powerful pieces that could checkmate with a single drop), depending on other factors like piece selection.  The sweet spot was right in the middle, at or near the 50% density of modern Shogi.

I’ve since begun experimenting with a 12×12 drops-based game called Yuudai Shogi (Grand Shogi) which follows these principles, and the results have been interesting:

The sample game above is actually unusually long; most test games so far have been under 300 moves (against AI with 120 seconds of thinking time).  The piece density is much less than Chu Shogi (precisely 50% in Yuudai), and the piece selection is very different as well.  Again I selected pieces with limitations, and there are no super-powered monstrosities like the Lion of Chu Shogi. 

This game is still very much in development, but so far it’s quite playable.  Games are over in a reasonable amount of time, given the board size and number of pieces, and play maintains momentum and intensity despite the huge armies protecting each King.  I imagine that between two very strong players, some games might top 500 moves.  Having said that, if someone is willing to spend the time necessary to master a huge 12×12 game with 35 types of pieces, I suspect they’ll have the patience needed to fight such an epic battle!

I should stress that I’m still at the early stages of these experiments, and I may end up finding that there are pretty hard limits on board size or other elements when building drops-based games.  But for the time being at least, I think there is a lot of promising and mostly untapped design space here that should be investigated further.

If I were to revise the general wisdom on drops-based games, I would just soften it a bit, and say that when building games of this type on large boards with strong pieces, just be prepared to do a lot of work.  Chess is a generally remarkable game category in that the basic framework remains robust and playable even when we use extremely powerful pieces and/or go nuts with board sizes and other aspects (Tenjiku Shogi is a fine example of this phenomenon).  For better or worse, this has allowed Chess variant design to be pretty wide open, given that one can place just about anything on the board and the resulting game will probably still work.  But when we introduce drops and pieces keep coming back to life, suddenly things get a lot more brittle, and relatively small changes can have large effects.

Given the nature of drops, perhaps this isn’t so surprising; a capture in a drops-based game constitutes a significant swing in material value between the two players, and allowing any piece to suddenly materialise nearly anywhere on the board necessarily amplifies its potential impact.  Combine that with complicated positions allowing hundreds or even thousands of possible moves (when using large boards), and we can see why certain non-linearities of effect start coming into play.

So, at this point my (very) basic suggestions are:

  • Drops-based games on large boards need armies where pieces have weaknesses and blind spots (note that pieces having weaknesses isn’t the same as a piece being weak!). 
  • Piece density should be at or around 50% to avoid overcrowding and games dragging on too long, or ending too suddenly under a hail of super-powered drops.
  • Even small changes in the composition of the armies can have large effects when drops are involved, so when you make a change, test, test, and test again!  

Over time I hope to build on these principles and develop more detailed tips, but for now, I hope at least some other Shogi variant designers might be inspired to give larger drops-based games a try and share their experiences.

Of course, these ideas are built on the assumption that we will keep drops exactly as they are in Shogi, but this doesn’t need to be the case.  What might happen, for example, if we took Chu Shogi, chopped out a few pieces to reduce density and added drops, but Lions, Free Kings, Dragon Kings and Dragon Horses were permanently eliminated from the game when captured?  Or what if we introduced a system like Kamikaze Mortal Shogi, where captured pieces degrade in power until eventually they fade away completely?  Or what if we restrict drops to only one’s own half of the board?  Or what if we allowed only a small number of captured pieces to be held in hand for drops, with any additional captured pieces being removed from the game instead?  So far as I know, few of these ideas have been tried with large-board drops-based variants.  At this point I’m fairly convinced that, with a bit of clever engineering, drops could be used in nearly any size of Shogi variant.

Next Moves

Now that testing is mostly done, I plan to release Shinjuu Shogi in a future Ai Ai update.  At that point, I’d welcome any feedback from players, and of course I’d be happy to play games with readers of this blog via Ai Ai’s online functionality.

During the construction of Shinjuu Shogi I’ve learned a great deal about the challenges of designing larger drops-based Shogi variants, so I’ve been applying this experience to my development of 12×12 Yuudai Shogi.  Over the next few months I plan to put this game through an intensive testing regime as I did Shinjuu Shogi, and hopefully at the end of all that I’ll have a game worthy of release.

While working on Shinjuu I also constructed several other 11×11 modern Shogi variants, all of which incorporated pieces from Taikyoku Shogi.  Each game focuses on a particular thematic category of pieces, like dragons, cats, and so on.  I’ve been testing these in tandem with Shinjuu, and have found them to be fun introductions to the wild piece selection of the enormous Taikyoku Shogi.  In the future I will feature these games as well, in particular Shouryu Shogi (Rising Dragon Shogi) and Puppy Shogi, which I believe to be the most promising games.

Tagged , , ,

Courier Chess, Part II: Modern Variations

Modern Takes on Wider Chess

In my previous post on Courier Chess, we took a look at the classic 12th century game of extended 12×8 Chess, which maintained a player base for hundreds of years before finally being overcome by the modern game. Courier is perhaps a bit deliberately paced compared to the 8×8 game we’re all familiar with, but I maintain that it has a great deal of charm and definite strategic interest.

We also met Courier Spiel, a 19th-century revamp of the classic game that added modern touches like a full-power Queen and double-stepping Pawns:

Note that in the Ai Ai implementation, Courier-Spiel uses more standard promotion rules — Pawns must promote on the last rank to any previously-captured piece of the same colour — rather than the strange triple-backward-hopping rules associated with the historical game. If there is demand I can look into providing the old rules, but frankly I think it makes endgames much slower and more difficult to convert, so I find the modernised promotion rules more enjoyable.

Below I’ll introduce you to a few interesting modern Chess variants that take Courier as a jumping-off point. All of these games are playable in Ai Ai against humans or against the AI. Feel free to give them a try, and let me know in the comments which ones you prefer.

Courier Chess with Alibabas

We’ll start here with a very simple variant — this one is exactly how it sounds: Courier Chess, but with the Alfils swapped for Alibabas. The Alibaba is a compound piece that combines the moves of the Alfil — jumping two squares diagonally — with the Dabbabah, which jumps two squares orthogonally. That gives the piece significantly more power, while still restricting it to a subset of the squares on the board.

I have to say I quite enjoy this simple piece swap; the resulting game is still deliberate and strategic like the original, but the Alibabas are stronger and exert more influence on the game. In the sample game, you can see how they prove very useful in the endgame for Black! I’m a fan of Chess variants with intricate endgame play, and this certainly provides that.

ArchCourier Chess

ArchCourier, invented by Eric Greenwood, is a highly-playable modern take on the Courier experience. Eric went to great trouble to tweak and test the game to ensure the piece mix was suitable and that the initial setup was balanced, and I feel he did a very good job.

Rules-wise the game is easy to grasp — all the familiar pieces from Chess move as in the modern game, including the Queen and Pawns. No castling is allowed. Pawns that reach the far rank must promote to any previously-captured piece of the same colour.

However, four new piece types join the fray in this game that give it a very distinct personality from classic Courier:

  • Centaur: moves as a Knight or a Guard (non-royal King)
  • Squirrel: may jump as a Knight, or jump two squares in any diagonal or orthogonal direction — in other words, it can jump to any square two steps away from its starting point
  • Dragon Horse: moves as a Bishop or steps one square orthogonally
  • Dragon King: moves as a Rook or steps one square diagonally

The Guard parked in front of the King provides useful defensive cover, particularly since there is no castling.

The new pieces are well-suited to this game and to Courier’s distinctive 12×8 board. I’m particularly a fan of the Squirrel, a flexible and fun leaping piece that can crack open some surprising tactical opportunities when used well. The Dragon King and Dragon Horse, which will be familiar to Shogi fans, are helpful additions on this larger board.

Alas, ArchCourier does not seem to have received much attention since its creation, which is unfortunate. I highly recommend it for fans of Courier who want something with a bit more zip, but with no real additional rules overhead to worry about.

Courier de los Combinados

Courier de los Combinados, created by prolific Chess inventor Charles Gilman, is essentially a combination of ancient Courier with Wildebeest Chess, a popular modern variant on an unusual 10×11 board. Wildebeest Chess extended the Royal Game by incorporating new leaping pieces: the Camel, an extended Knight found in the ancient large variant Tamerlane Chess; and the Wildebeest (also known as the Gnu), a combination of Camel and Knight:

Wildebeest Chess (which I’ll cover in a future post) is known for a wide-open style of play, thanks to these long-distance leaping pieces and a fairly empty board compared to standard Chess. Courier de los Combinados goes the opposite direction, packing Courier’s 12×8 board with additional pieces:

Alongside these two Camels and a Wildebeest on each side, the Wazir and Ferz pieces have been doubled up and parked in front of the King, creating an advanced forward line of Pawns. The Alfils are gone entirely, and the Queen is upgraded to its modern power levels. The King may also castle with his Rooks, and Pawns must promote at the last rank to Queen, Wildebeest or Guard.

The play that results is quite interesting; while the initial setup is fairly dense, the long-distance leapers provide some lethal early threats, making the opening surprisingly sharp. In fact, be careful which Pawns you move in the opening, as very quick Wildebeest mates are possible if you open up the wrong squares! Note that in the sample game, an AI vs AI test game, both sides elected to allow the opponent to take a Rook early on; I suspect this is very much not an optimal opening! Clearly I need to do some tweaking to Ai Ai’s engine parameters for this game.

Charles Gilman produced a shocking number of variants during his time on the Chess Variant pages, and was at times criticised for not actually playtesting enough of his games. I can say however that this take on Courier has proven to be quite interesting, and may provide some intriguing challenges for opening theoreticians!

Northern Ecumenical Chess

Here we have another interesting Charles Gilman creation, this time extending Courier’s board to 16×8, or the equivalent of two standard Chess boards laid side-by-side. Northern Ecumenical Chess takes Courier de los Combinados to the next level by adding in compound pieces of the Knight and Camel:

  • Cardinal: moves as a Knight or a Bishop
  • Marshal: moves as a Knight or a Rook
  • Caliph: moves as a Camel or a Bishop
  • Canvasser: moves as a Camel or a Rook

Rules-wise, we have castling in this game (with the Rooks), Pawns may take an initial double step, and must promote to any non-Pawn piece upon reaching the last rank.

I believe this was one of the first games I added to Ai Ai once the Camel compound pieces were available, which tells you something about my opinion of the game. I have a certain fondness for oblong boards in Chess, and going all the way to 16×8 is even more entertaining in my opinion; such boards can be easily constructed by mashing together two standard boards, which is a practical benefit, and the odd shape creates a substantially different feel in play. As you might expect, pieces that can slide horizontally — Rooks, Marshals, Canvassers and Queens — are even more useful here, as they can reposition more easily than the diagonal pieces, which take multiple moves to cross the longer axis of this board. The Knight feels less powerful on a board like this one, given how long it takes to get it to a useful position, but in contrast the Wildebeest is a flexible and dangerous attacker.

All told, I enjoy this variant quite a bit and find it a nice balance between rough-and-tumble tactics and slow-burn positional play. If you fancy experimenting with a very long board and some unusual pieces, I can highly recommend giving this a try.

Other Oblong Oddities: Duke of Rutland’s Chess

As I got into Courier Chess and its variants, I soon discovered that Courier was not the only historical Northern European expanded Chess game on an oblong board. This game was invented by John Manners, the Third Duke of Rutland in 1747, and it managed to have a brief spell of popularity until his death in 1779. The game has languished in obscurity since then, but we still honour the Duke’s creative accomplishment by calling it Duke of Rutland’s Chess.

Rutland’s Chess is an odd game, not just for its unusual board size, but also due to the selection of pieces:

Click on the image of the starting position and you’ll see the strangeness for yourself (note I use the starting setup provided in Jean-Louis Cazaux’s book A World of Chess, rather than the setup on the Chess Variant Pages which is the mirror image of this one). First, we have four Bishops instead of two, and three Knights instead of two, which introduces an unusual imbalance between the minor pieces. Second, we have Dragon Kings present — moving as Rook or King — yet we have no Dragon Horses (Bishop or King). Third, we have a Marshal (Rook + Knight), but no Cardinal (Bishop + Knight).

This strange starting army raises a number of questions. Why are two of the Bishops not Dragon Horses instead? Did the Duke not know about them? If he didn’t, one would expect he would have easily extended the Dragon King’s move to the Bishop and reached the same conclusion anyway, but it seems he didn’t. Why the third Knight? We know that the Cardinal and Marshal (Bishop + Knight and Rook + Knight) were known at least since 1617 in Carrera’s Chess, and a Marshal is present here, so why no Cardinal? I find those oddities a bit fascinating; perhaps we may never know why the Duke made these particular design choices.

All that aside, the resulting game is nevertheless quite entertaining. The extended board is deeper than that of Courier Chess, providing more room for manoeuvre. The Pawns in this game have the option of an initial double- or triple-step as their first move, meaning that the deeper board doesn’t particularly bog down the opening. The presence of the modern Queen, Marshal and Dragon King further open things up, providing opportunities for sharp tactical play. After a few plays one can see why the game found a certain following, at least for a brief time; apparently the legendary Philidor was quite the fan of this game.

Modern Manners

As one might expect, Duke John Manners’ unusual game has provoked some interest in the Chess variant community, and various inventors have tried to resolve the imbalances in the original game’s starting position. A simple and effective example is Modern Manners, coming to us from the ever-prolific Charles Gilman.

Modern Manners uses the same board as Rutland’s Chess (though the image in the linked CVP entry shows a 14×8 board instead, for some reason), but alters the starting setup in the most logical way:

Modern Manners replaces the odd third Knight with a Cardinal (Bishop + Knight), and swaps out two Bishops for two Dragon Kings, giving us a logically complete setup and a slightly increased overall power level. As in Rutland’s Chess, Pawns have the option of an initial triple-step forward, and they may promote to any non-Pawn piece upon reaching the last rank.

I highly recommend this game as a modern alternative to Rutland’s Chess; the new setup is easier to understand and use, and the addition of the Cardinal in particular livens up the tactics even further. Of course you should try Rutland’s innovation first, as you may well find its quirkiness to your liking! Personally I feel these simple changes transform the game into a more engaging and more logical game, and yet it retains the wide-open, fast-paced feel of the original, so I prefer this to Rutland’s effort.

Manners Kamil

Charles Gilman didn’t stop there — on the very same CVP entry you can find Manners Kamil, a 16×8 game that adds two Camels and swaps the Cardinal and Marshal for two Wildebeests:

The lengthened board and altered pieces substantially change the feel of this game. The powerful long-distance attacks of the Cardinal and Marshal are replaced with the airborne threat of the Wildebeests, leading to some tricky tactics in the early stages of the game. In fact I’m slightly uncertain as to whether the game is meant to be played on a 16×10 board; the CVP entry is ambiguous, given that the Modern Manners board is shown as 14×8 despite the clear statement in the text that it uses the Rutland Chess board. The Manners Kamil description states that 2 files are added but does not explicitly mention the board’s depth, so perhaps 16×10 should be used instead of 16×8.

In any case, I will experiment more with Manners Kamil on 16×10 and provide the .mgl file in a future release of Ai Ai. For now I do recommend giving this game a try on 16×8 and seeing how you feel about it. The compressed vertical dimension is somewhat compensated by the reduction in long-range pieces, and on the whole the game still feels fairly open. However, I sense that the player best able to develop their Rooks and Dragon Kings may enjoy a good advantage; without them in play, only the Queen can hold sway over the board’s longer axis.

In my admittedly limited test plays thus far, I sense that the 16×10 version may be slightly better. The open playing area is more reminiscent of Modern Manners and its historical predecessor, and the additional two rows give the opening a bit more room to breathe, while also reducing the early mate threats of the Camels and Wildebeests.

Sneak Preview: Giga Rutland Chess

Here’s a quick sneak peek at one of my own creations — Giga Rutland Chess, a huge expansion of the original’s rather sparse starting position:

An entire extra row of pieces is added to each side, which includes types of pieces completely new to Rutland’s Chess:

  • Gryphon: a bent slider piece that steps one square as a Bishop, then slides any distance as a Rook
  • Rhino: the Gryphon’s counterpart that steps one square as a Rook, then slides any distance as a Bishop
  • Paovao: based on the wonderful Cannon of Xiangqi, this piece must jump over exactly one piece to capture something on the other side
  • Bison: a combination of the Camel and the Zebra, which is another extended Knight that jumps 3 squares outward then 2 squares to either side
  • And more besides!

I won’t go into much detail here because the game is still being tested, but my intent has been to provide a denser, more hectic game on the 14×10 board, and while it may be rather too much for some, I find it enjoyable so far. More to come on this game once it’s finalised.

Alekhine Chess

This game is a bit of a mystery in some ways — despite the name being Alekhine Chess it has nothing to do with the famous Chess player Alekhine. Why is it called Alekhine Chess then, you ask? I have no idea. But that aside, it’s an intense and dramatic variant on a 14×8 board, and a good choice for fans of brutal tactics.

As large variants go, the setup here is pretty simple: all the standard Chess pieces are here, plus an extra Queen, two Marshals, two Cardinals, and an Amazon (Queen + Knight). The presence of all these powerful pieces generates a game of intense tactics; most of the pieces on the board are incredibly mobile, and the Amazon can actually checkmate the enemy King completely on its own, without the help of any other piece!

I must say I was a bit sceptical of this game initially, as the Cardinal, Marshal and Amazon are very common pieces in Chess variants and not that fascinating, and the sheer power density on the board seemed like it might reduce the game to a simplistic slugfest. But in practice, it’s anything but simplistic. The entire game will balance on a knife edge, certainly, and a single blunder could result in a horrendous slaughter, but there’s absolutely room for positional thinking thanks to the large dimensions of the board. As you can see in the GIF, each game of Alekhine is high intensity, and when playing you must get comfortable with being under constant threat at all times! I enjoy this variant a lot, and would be happy to see it gain more traction.

Double Chess

Here we have a game that doesn’t quite fit the mould we’ve established here so far. Double Chess is exactly what it says on the tin: Chess, except doubled! We use a double-length 16×8 board, give each player two full Chess sets for their side (but replace the extra King with a Queen), and off we go:

Now, some may quite rightly take issue with me including this game in a list of Courier-inspired games. Courier, Rutland’s Chess, and the others in this article are full-on extensions of Chess, expanding both the board and the pieces available to each army. Double Chess extends only the board, using only standard pieces. However, in my mind the use of the double board places it in a similar category, if only because of the distinctive feel of playing on these elongated boards. The use of only standard pieces also makes the game extremely easy to play in real life — just buy two cheap tournament Chess sets with those roll-up plastic boards, trim the edges so you can fit the board edges snugly together, plonk your two sets of pieces down, and off you go!

The resulting game is surprisingly different from normal Chess, and of course can run substantially longer. I usually prefer games with new pieces, but this game is one of the few using only the standard army that I find just about as good as the more adventurous designs. Creator David Short also provides some sample games and a fully-commented game to help the beginner, which is fantastic for those looking to give this game a try.

On the whole it’s an excellent game, and well worth your time.

Full Double Chess

Full Double Chess takes the basic idea of Double Chess and extends it, so that the board now includes the compounds of the Knight with the Bishop, Rook and Queen:

Full Double Chess has one immediately obvious deviation from Double Chess — each side has two Kings! Each King can also castle with either the Marshal or Rook on either side, meaning that when implementing this game I had to allow for a huge assortment of possible castling options. Both Kings must be eliminated to win, which is a nice counterbalance to the huge power of the additional Cardinals, Marshals and Amazons on the board. Thanks to the double Kings, the game won’t end immediately after a single blunder — though of course the Amazon could easily checkmate both Kings at once if you’re not careful!

Full Double Chess is quite a fun variation on the oblong Chess theme; the double Kings are an unusual feature that works really well in this context. The additional Knight compound pieces are again not the most exciting pieces, but fit perfectly in this game, and make it easy to introduce this game to someone who may be new to the concept of Chess variants as a whole. I think this game co-exists quite nicely with Double Chess; they play entirely differently and present really divergent takes on the double-board theme. David Short, creator of Double Chess, was rather disparaging of Full Double Chess on the basis of its ‘new fangled’ pieces; however, the added pieces in this game were in fact invented centuries ago, so the only way they could really be ‘new fangled’ is if David is several hundred years old!

My only issue is the name, really. The creator insists the game includes all the compounds of the basic Chess pieces, but we are actually missing Rook + Knight + King and Bishop + Knight + King, which can be seen in Jean-Louis Cazaux’s Very Heavy Chess (more in this game in a future article). In my opinion, a truly ‘Full’ Double Chess perhaps should include those pieces and widen the board even further. Besides that very pedantic gripe, I strongly recommend giving this game a try!

Final Thoughts

I hope you’ve enjoyed these brief introductions to a few games that build on the oblong foundations of the ancient game of Courier Chess. There are more games in this category that I haven’t covered, of course, but I’ve tried to include a good variety of options within Ai Ai that give players some interesting and distinct games of this type.

If you’re new to these types of variants, I’d recommend that you start with the historical games first — Courier Chess, Courier-Spiel and Duke of Rutland’s Chess. These games are very distinct from modern Chess but are easy to pick up and play as a newbie, and they’ll be a nice introduction to the feel of playing on a widened board. From there, you can progress to modern takes on each, with the natural next step from Courier being ArchCourier in my opinion, and of course Modern Manners is the natural follow-up to Rutland’s Chess.

If you want a modern take that diverges further from the Courier model, Double Chess is an excellent and highly accessible double-board game to try out. Alekhine Chess is a good one to try as well, particularly for the tacticians out there. Finally, if you hunger for some new pieces on the longer boards, I can recommend Northern Ecumenical Chess as an enjoyable and balanced double-board game with interesting yet intuitive new pieces.

From here I will move on to some other approaches to large board Chess and give some more comments and reviews on games in these categories. First I plan to cover games that extend the Chess army through compounds of classical pieces, like we saw in Alekhine and Full Double Chess; there are a LOT of games like this however, so I’ll be quite selective in my coverage. I’ll also include some games of mine in this category, which take the compound-piece approach to its logical extreme.

I should mention that in the previous Courier article, I’d promised to provide some detailed commentary on sample games of Courier and Courier-Spiel; unfortunately I have to put this to one side for now, as I did a tonne of work for that post and managed to lose nearly all of it! I do plan to get back to that, but I’ll need to start my analysis again, so it will take some time.

Tagged , , , , , , , , , , , ,

Reviewing (almost) all the games on MindSports

UPDATE 6 Nov 2021 — Oust review updated, thanks to finally having a breakthrough!


Christian Freeling’s MindSports site is an essential destination if you want to play some good abstract strategy games.  Christian’s own games take centre stage, of course, but numerous other games, both traditional and modern, are also featured.

During the pandemic I’ve taken the time to try out most of the games on MindSports, either against AI or using my collection of various printed boards.  I was keeping my impressions in a little text file, mainly as a reminder of what to focus on in future blog posts, but have decided to put them up here in case they might help anyone looking for something to play but perhaps lacking the time to try everything until they find a hit.

Before I start I’m going to do something I wish other reviewers did, which is provide a summary of my general perspectives on games.  Hopefully this will give you an idea of where I’m coming from, and will help you interpret my mini-reviews in that light.

  1. I value games with tradition, that have survived centuries of scrutiny, and that have a culture that has grown around them.  However, I also love experimenting with new rules, and with variants of existing games.  So in that sense I probably would annoy both the traditionalists and the cult-of-the-new types.
  2. Simple rules are nice, but as a regular player of very complex board games,  I tend not to consider rules simplicity a particularly important factor in my evaluations of games.  A good game is good regardless of how long the rules document happens to be.
  3. I have dabbled in game design but above all I’m a player.  With that in mind I’m trying to avoid terms like ‘arbitrariness’ or ‘inelegance’, which are ill-defined terms that relate to mostly to rulesets and not gameplay.

And my scoring system, which uses the entire 0-10 scale:

  • 0-2/10: I do not like this.
  • 3-4/10: There are other similar, and better options available.
  • 5-6/10: Good but not earth-shattering.  Worth a try if you like that sort of thing.
  • 7-8/10: Very good games that deserve your attention, even if you’re not normally into the genre.
  • 9-10/10: Games that are either modern classics or traditional classics.  Potential lifestyle games.

While I’m here, a disclaimer: these opinions are intentionally brief, to-the-point, and somewhat flippant.  I don’t claim any particular authority here, so do yourself a favour and don’t take these overly seriously!

With that out of the way, let’s start with the games in The Arena.

The Arena

Christian’s six essential games

grand-chess-initpos

Grand Chess Christian’s 10×10 Chess variant adding the Cardinal (Bishop + Knight) and Marshal (Rook + Knight) to the traditional lineup.  On the whole this is a decent variant, but I have some issues with this game — the Cardinal/Marshal were first used by Pietro Carrera in 1617 and at this point are not very exciting; the initial position allows Marshals to be traded off too easily; the removal of castling is unnecessary in my opinion; the presence of three power-pieces of similar value dilutes the excitement of the game somewhat; and the piece density is too low for my taste, making for a big game that doesn’t actually expand Chess’ strategic/tactical landscape as much as I would like.  What really bugs me is the promotion rule, which needlessly punishes early promotions and leads to the awkward rule forcing Pawns to remain in suspended animation if no captured pieces are available.  I don’t like that rule in the historical variants that include it, and I don’t like it here either; just let me promote stuff, promoting stuff is fun!  Honestly I would probably score this game higher, except that in its wake this promotion rule appears to have become more popular, which is a problem for me.  In any case, Grand Chess is fun but there are other 10×10 variants I prefer over this one, and I still plan to feature some of these in a future post eventually (a recent favourite of mine is Expanded Chess).  4/10.

dameo-10s-mv121

Dameo Christian’s highly-regarded Draughts variant, often touted as a replacement for 10×10 International Draughts.  Unfortunately, this game doesn’t quite work for me.  The linear movement aspect feels out of place in a Draughts game, and the kings are a bit too strong for my taste.  Various people have told me I’m wrong about this, but I can’t shake the impression that breakthrough is the critical factor in this game, and that getting a king first is a very strong indication that player will win.  I can, however, see why people like this game, so I give it a good score on that basis; certainly opening play will be varied, and the draw rate is low.  Personally I’d rather play Croda or Turkish Draughts though.  6/10.

emergo-5s-1

Emergo Here Christian boils down column checkers variants like Bashni and Stappeldammen to their essentials.  The result is a frantic tactical game characterised by a somewhat confusing placement phase, followed by a violent explosion of captures in all different directions.  The game is most definitely unique and at times feels brilliant, but I’m not much for placement phases in Draughts (or Chess, for that matter), and I do find myself missing the incessant forward motion inherent in Bashni and Stappeldammen.  Definitely worth a go if you haven’t tried it.  If I can ever get past the placement aspect, I feel this game could become a favourite of mine.  7/10.

Sygo A Go variant that uses the Symple move protocol combined with Othello-style toggle-capture.  I started out a bit skeptical of this game, given my love for Go itself; most Go variants are a fun distraction at best.  Sygo however does have a distinct personality, and the Symple protocol is a slice of genius.  As Go variants go, it’s better than most, but given the choice I’ll take Symple every time.  7/10.

Symple — Formidable.  Once I sunk my teeth into this game I was blown away by how singularly unique it feels.  In Symple one can either place a single stone on its own to start a new group, or grow every group of yours on the board by one stone; the goal incentivises building small numbers of large groups of stones.  The game creates a feeling of constant tension that builds until the endgame, where precise calculation is required to squeeze out the last few points, or to force your opponent into bad placements.  Symple is one of the most remarkable and exciting discoveries to be found in modern abstract gaming, in my opinion.  10/10. (read more here)

Storisende — An extremely unusual game of territory, featuring pieces that move and split, and walls that form, and there’s just a lot going on here.  I’m not sure I can really rate this game properly, as I simply don’t understand it.  Christian has made various valiant efforts to explain this game, but I still find it confusing.  I trust his judgment, generally speaking, so I believe something interesting must be happening here, but I just can’t find it yet.  Note that Christian has written quite a lot about this game and its strategic complexities, with another article forthcoming soon; I plan to study these articles and revisit the game again in the future.   5/10.

The Arena: The Other Seventeen Freeling Games

yari_shogi_inpos

Yari Shogi — A 7×9 Shogi variant that uses pieces with lots of Lance-like forward movement.  Any Shogi variant is a tough sell for me, because the historical variants are so incredibly interesting.  In contrast to something like Tori Shogi or Wa Shogi, Yari feels a bit pedestrian.  Also, only one piece type can move backward at all, meaning many more games of Yari will end in an impasse as compared to standard Shogi, as the King will be very safe from attack if it enters the opponent’s camp.  I also admit to a certain dislike for attempts to ‘Westernise’ Shogi; Shogi is a Japanese game made by Japanese people, and it bothers me that we so often try to strip its Japanese-ness away, rather than meet the game on its own terms.  4/10.

Dragonfly — This game is reasonably fun, but if I’m going to play a 7×7 game with drops I’ll stick with Tori Shogi.  Dragonfly also has a hovering unpromotable Pawn thing that’s similar to the promotion rule I dislike in Grand Chess, and the Pawns can’t be dropped, which diminishes the excitement somewhat.  I understand why that rule exists in this particular game, but Christian’s claim that Chess Pawns aren’t suitable for drops is odd to me — they work perfectly well in Crazyhouse, Bughouse, Chessgi, etc.  3/10.

Chess+ I appreciate the design goal of this game, which was to create a version of Chess which eliminated opening theory without requiring some external list of starting positions like Chess960.  Unfortunately it doesn’t work that well, in my opinion.  The opening protocol adds significant cognitive load for the new player.  Messing up is very easy, and early losses can happen because the long-range power of the Chess army makes it relatively easy to punish bad placements quickly.  If you want to play Chess with placement, I recommend the traditional Burmese variant Sittuyin; this game works better, in my opinion, as the pieces are weaker and bad placements are less immediately dangerous thanks to the fixed Pawn structure.   4/10.

Chad, Rotary — I haven’t played these.

Hexdame A direct translation of International Draughts to the hex board.  Quite enjoyable, although a lot of the play seems to bunch up around the edges; granted, that happens in other Draughts variants too, but I notice it more here for some reason.  The tactics are enjoyable, and I wouldn’t turn down a game of this, but it’s perhaps a bit *too* tactical.  6/10.

Bushka — A Draughts-adjacent game inspired by the classic Fanorona.  Here pieces capture by approach rather than by leaping.  This game also originated the linear movement mechanic which made its way into Dameo.  The idea is intriguing, but in actual play I find it a bit confusing.  The games I’ve played also ended surprisingly quickly, which was a bit disconcerting given the size of the board and the number of pieces.  Colour me skeptical, but interested to learn more.  6/10.

Pit of Pillars — haven’t played this yet.

Io — An excellent showcase for the under-appreciated one-bound, one-free opening protocol.  Io is an Othello variant that starts with a placement phase, but the innovative one-bound, one-free procedure ensures that the pieces of both sides are placed in interesting ways and tends to create balanced positions for the subsequent capture-fest.  A very good game that deserves more attention.  8/10.

PhalanxThis game has several elements not usually seen in combination, bringing together placement, movement, capture, movement of whole groups, all in service of a territorial goal.  The result is complex and intriguing.  I’m intrigued by the capture mechanic, which ends up creating impassible walls around the board that gradually constrict and focus play until the inevitable conclusion.  I can’t pretend to have a good grasp on this game yet, but so far I like what I see.  8/10.

Mu VeloxThis game scares me, so I haven’t tried it yet.

InertiaThis game is a rare example of the unification genre, in which players strive to be the first to bring all their pieces together into a single connected group.  While reminiscent of the venerable Lines of Action, Inertia feels quite different due to the variable opening position produced by the one-bound, one-free opening protocol, and the movement mechanics which are easier to grasp than in LoA.  Enjoyable.  7/10.

KnightVision This very recent creation melds the classic goal and rhombic board of Hex with a focus on a hexagonal version of the Knight’s leap.  Christian’s pitch says ‘it adds drama to an incredibly deep game without affecting its depth’, but I slightly disagree; the Knight’s-move placement pushes more locality on a game that thrives on global strategy.  Having said that, the chucking of axes is very fun, and the game overall is a good experience.  But if someone breaks out a Hex board I’m going to play Hex on it over this every time.  7/10.

havannah-elaboratefork

Havannah — I respect Christian enormously, but his decision to exclude Havannah from his ‘games that matter’ list is forever baffling to me.  A classic by any measure, Havannah is easy to understand yet blessed with bottomless depth.  The multiple goals create a sense of limitlessness that few games can muster.  I can understand the impulse to push his fans toward his later games, and certainly he’s produced many great games after Havannah.  But equally, there’s nothing wrong with getting it right the first time.  Brilliant.  10/10. (read more here)

Starweb — A game of connecting corner cells together, with point scores growing rapidly as groups encompass more and more corners.  A beautiful melding of connective and territorial impulses, this game has been a favourite of mine since I first heard of it.  I continue to hope more players will discover the strategic delicacy that emerges on this unique star-shaped board.  I like this game so much that I asked Stephen to add even bigger boards to Ai Ai, so that I could immerse myself in the game for even longer sessions (in the collage of sample games above, the top left one is playing on the standard size-10 board, the others are on the bigger size-11, 12 and 13 boards).  Essential.  10/10. (read more here)

Scware — Here we find a noble attempt to combine the Symple protocol with a connective goal.  There’s fun to be had here; however, the restrictions in place feel rather *too* restrictive to me in the context of a connection goal.  In the sub-genre of connective Symplistic games, I think SympleHex just squeaks past this one.  6/10.

The Glass Bead Game — I’m not much of a mancala player, so I haven’t played this one yet.

The Arena: Other Contributions

mitrofanov-5

Chess — Still the greatest source of drama available on an 8×8 square grid.  This game has captured the gaming world’s imagination to such an extent that it has birthed an entire global genre of checkmate-based games and variants.  I learned this game in my youth, like many people, but only truly learned to appreciate it in my old age.  Now it’s a fixture of my daily life.  10/10.

shogi-set-cropped

Shogi — Possibly the greatest game ever invented by humans.  The rules, piece movements, and aesthetics marry perfectly to create a glorious experience — a game of delicate beauty and intricate manoeuvre, combined with vicious tactics and constant aggression.  I love everything about this game, and my Shogi board and traditional pieces with Minase calligraphy (pictured above) are perhaps my favourite gaming item in my home.  If I could give this 11/10, I would.  You know what, I’m going to, this is my blog, dammit!  11/10. (read more here)

int-draughts-2s-mv123

10×10 Draughts As a kid I played Checkers, Anglo-American style, and like many other kids I played it incorrectly (with optional captures instead of mandatory).  Years later I found out that A) Checkers is unbelievably better when you play it properly, and B) a 10×10 version exists with long-range kings and backwards captures!  Boy did I feel silly once I realised I’d been playing Checkers fundamentally wrong all that time.

Incidentally, since that embarrassing realisation I’ve been somewhat relieved to find that apparently a huge number of other people also never played Checkers correctly.  Here are a few reviews of the app All-in-One Checkers on Android:

checkers-morons-1

So, on the plus side, at least I figured out my mistake eventually!  I feel for all these people who never had that moment, though.  So far as I’m aware there’s not any established Checkers/Draughts variants out there with optional capture, and thank goodness for that, because from that rule comes all kinds of tactical excitement.

Anyway, the discovery of 10×10 International Draughts was quite a moment in my gaming life.  I fell in love with this glorious game of spectacular combinations.  Christian often criticises the game for being drawish at the high levels of play, and while that’s completely true, I’m not good enough for that to be a major problem.  10/10.

Constitutional Draughts This variant of 10×10 Draughts reduces drawishness by restricting kings, forcing them to avoid crossing squares where they could be captured.  For me this substantially screws up my endgame tactical vision, and feels more awkward than Killer Draughts’ simple restriction, which also has predecessors in historical Draughts variants.  The constitutional restriction reminds me of Caissa Britannia, a Chess variant invented in 2003, which is where I first saw this concept; perhaps unfairly, Constitutional Draughts feels weirdly un-Draughts-like to me as a result.  6/10.

frisian1

Frisian Draughts — Frisian Draughts massively expands the tactical craziness of 10×10 Draughts by allowing pieces and kings to capture orthogonally as well as diagonally.  The result is very hectic, with spectacular combos appearing frequently, and often in surprising ways.  Initially I found this game too confusing, but recently it finally clicked; I played a game against a strong AI (on Lidraughts), and suddenly I found myself seeing patterns and creating robust structures that I had never thought of before.  I’m now convinced this is an excellent take on Draughts, and the tiny 5% draw rate in competitive play is a great bonus as well.  Download this 59 page guide for a great intro to the game.  9/10.

stappeldammen-end

Stappeldammen — 10×10 Draughts extended into verticality.  Like Bashni or Emergo, here pieces build stacks as they capture, and subsequent captures can re-expose stacks, leading to frequent changes in ownership.  A unique element here as that there is no promotion, meaning stacks can get stuck on the last rank until freed by the appearance of an adjacent capture target.  This seems weird at first, but I agree with Christian that it adds an interesting strategic wrinkle.  Bashni is more fun though.  7/10.

oust7-win-10s-mv43

Oust After numerous attempts, I finally had a breakthrough moment with this game.  Oust is a game with the unusual property that each game both starts and ends with an empty board.  Players may either place a stone anywhere, so long as it’s not adjacent to their own stones, or grow one of their groups on the board in order to capture an adjacent opposing group of lesser size.  The winner is the player who completely annihilates his opponent’s pieces.  Critically, each capturing move must be followed by another placement, which means that wins can appear in sudden and shocking fashion as one capture chains into another.  I found the game utterly impenetrable at first, but at last I figured out some basic strategic principles and could begin to place my pieces on the board in some sort of structured way.  Now I find the game innovative and surprising, rather than confusing and opaque.  However, I probably played about 30 games to reach this point, so be aware that you may not ‘get it’ right away.  Whether it’s worth it to you to face that learning curve is another question; in any case I would say the game deserves a few tries so you can judge that for yourself.  7/10.

Many people praise this odd game of placement and elimination, but I find it utterly baffling.  Some other games have boggled me in the past, but something kept me motivated to try again until a lightbulb eventually went off.  Here, for whatever reason, that didn’t happen for me.  I’ve no doubt this will be intriguing for others, and the uniqueness is obvious, but for me it’s an alienating, confusing experience.  Nick Bentley’s Bug is similar to Oust, and that game confuses the heck out of me too!  I played both games a bunch of times and never got the sense I was unlocking any additional understanding or improving my play in any meaningful way.  3/10.

IMG_20200930_180728

A treasured part of my gaming collection — a Go board made from Japanese Kaya, with the lines drawn via a sword dipped in black laquer (yes, really). The stones are the traditional slate and clamshell.

Go — The quintessential game of territory, one of the oldest games on the planet, and perhaps the most revered.  What can I say about this game that hasn’t already been said?  Go is staggeringly deep, tactically complex and strategically varied, and capable of making hours fly by as you end up completely absorbed in the infinite possibilities in front of you on that 19×19 grid.  Many people praise it for its simple rules, but I tend to stay away from that characterisation; in theory, the rules are simple, but the learning curve is steep, so it feels anything but simple as a beginner.  Go can take a while to reveal its character and richness to the newbie; beginners are often urged to dive in and lose 100 games as quickly as possible, as it takes a lot of bitter experience to grasp the basic concepts.  But if Go does end up hitting the mark for you, you may well find it takes over your gaming life.  10/10. (read more here)

rosette-sample

Rosette Go on a hex board has been a goal for many a designer, and Rosette makes it work through a wonderfully simple mechanism: if a player occupies all six points of a single hexagon, that group of stones cannot be captured, ever.  In a single stroke this rule compensates for the reduced number of liberties on the hex board, and allows a Go-like intricacy to flourish on the hexagonal grid.  Yes, I know it’s played on the intersections so it’s actually the dual of the hex board, but it’s hexagonal enough for me, and a great game besides!  8/10.

stigmergy-sample

Stigmergy This game started life as a Tumbleweed variant, and has subsequently branched off into its own thing.  I actually came up with the name, which apparently everyone hates, but as yet no one has come up with a better one.  Anyway, Stigmergy is a good game, substantially easier to play OTB than Tumbleweed, but also feels a bit more sterile.  I’ve yet to play the Stigmergised connection game that recently appeared on MindSports, but I suspect that might suit me a little better.  7/10.

IMG_20200321_190157

Ready to play a quick game of Ayu on my 9×9 Go board.

Ayu Another example of the underserved unification genre, Ayu (‘attach your units’) forces the game toward inevitable unification via an easy-to-understand movement restriction.  I’m very bad at this game but I still love it, which in my view is a clear indication that a game has something special about it.  Strongly recommended.  9/10.

IMG_20200312_113228

Playing 19×19 Hex on a board of my own design.

Hex — The game which looms so large in the world of modern abstract games that any connection game is inevitably compared to it.  Hex is incredibly simple to explain — take turns placing a single stone of your colour, win by being the first to connect your coloured sides with a continuous line of stones — but in play one finds endless intricacy and an innate focus on global strategy.  The more I play Hex, the more I admire it, and unlike some other games I admire, I actually thoroughly *enjoy* it as well.  I strongly recommend starting on 13×13 boards, then progress to 15×15 and 19×19, where the territorial aspects become significantly more important.  A must play.  10/10. (read more here)

The Pit

There are many games on this page, so I’m going to simply skip over those I haven’t played, to save some space!

Chess Variants

I like my Chess variants expansive and with unusual pieces, and most of these games don’t hit that mark for me.  Loonybird is fun, but divergent pieces are pretty common in Chess variants, so I didn’t stick with it for long.

Chess960 — Christian hates this game, and his complaints about it from a design perspective are reasonable, but as a player I don’t find them to be a problem.  FischerRandom/Chess960 had as its goal the elimination of opening theory, and it succeeds admirably at this without much in the way of rules changes, and now with official FIDE backing it’s easily the most important Chess variant on the planet (not the best, but the most impactful).  I’d quite like to ditch castling from FRC/960, which is awkwardly implemented in my opinion, but otherwise it’s a simple variant that allows interesting Chess to happen without 20 moves of opening prep needing to happen first.  8/10.

xiangqi-initpos

Xiangqi — China’s traditional form of Chess, virtually unchanged over the last 900 years.  Played on a larger 9×10 board, yet is more tactical rather than strategic.  The feel is substantially different from FIDE Chess, in part because the concept of material value here is much less important than positional advantage.  I very much enjoy playing Xiangqi, but I prefer Janggi, which removes the river and allows the Elephants to roam free (plus they have a more interesting move).  Xiangqi deserves additional credit for gifting us the wonderful Cannon piece, which is extremely interesting to use and is an excellent addition to numerous large Chess variants (like the superb Shako and its larger relatives Zanzibar, Maasai, Gigachess, and Terachess).  The downside for me is the relative lack of Pawn structure and promotion (except for slightly stronger Pawns); however, I must admit the game works fine without these, thanks to the confinement of the King in the palace, which allows checkmate to occur even when little attacking power is left on the board.  If you want to learn more about Xiangqi, Jim Png’s excellent XQ in English site has everything you need; do check out his excellent introductory books and translations of ancient Xiangqi texts as well.  8/10.

Games of Annihilation

turkish-draughts-prob1

Turkish Draughts — For some reason, despite having played Checkers in my youth I never considered the idea of an orthogonal version, so seeing this game for the first time was a real eureka moment.  In Turkish Draughts, players start with 16 pieces each on the 8×8 board, and men may move one square straight ahead or in either sideways direction; kings move like Rooks in Chess and capture via the long leap.  Unlike Draughts, pieces are removed as soon as they are jumped rather than at the end of the sequence, which can allow some spectacularly long capturing sequences.  The tactical problem above (White to move and win) is a great example; the final move captures 12 checkers at once!  I love this game and play it constantly on my phone, and I feel it should be just as revered as Draughts and Checkers. 10/10.

Armenian Draughts This variant of Turkish Draughts spices things up by allowing men to move diagonally.  This seemingly small change actually completely alters the gameplay, removing the concept of opposition.  For fans of Draughts this is worth a try.  Kings are strong as in Dameo, but men here are also more mobile, which feels a bit better in my opinion.  However, I’m a little unsure about the rules reported in the English-speaking world; in Russian sources, it is said that pieces are removed at the end of a capture sequence, rather than immediately.  I’m trying to find Armenian sources to get a definitive answer, but as yet have been unsuccessful.  Most of the Russian-language information seems to come from the book Checkers: 60 Unusual Games on a Classic Board by Alexander Pavlovich, so I’m working on obtaining this book.  8/10.

bashni-problem-1

Bashni — The columnar version of Russian Draughts, where captured pieces are continually stacked under the capturer.  Stacks can be unveiled again by re-capturing the top piece off a stack, leading to amazingly complex combinations and exciting back-and-forth action.  Promoted stacks capture with the long leap, too, adding to the hectic nature of the game.  An absolute blast.  Abstract Games Magazine featured this game in a number of articles, and included numerous useful tactical and strategic tips, as well as some nice problems like the above (White to move and win).  10/10.

croda-30s-mv148

Croda — This ingenious Draughts variant takes Turkish Draughts, replaces the sideways moves of the men with diagonal ones, but keeps capture orthogonal only… then stuffs the 8×8 board with 48 pieces.  The result is an action-packed game with lots of opportunities for exciting combinations.  Dameo was inspired by this game, but I actually prefer Croda; the lack of linear movement and the weaker kings produces a simpler game that feels more balanced to me.  Admittedly the average Croda game will be much longer than a typical Dameo game, but all other things being equal I usually prefer longer games to shorter ones.  A viable alternative to 10×10 Draughts, if you’re worried about drawishness.  9/10.

Fanorona A traditional elimination game that uses contact capture, and that later inspired Freeling’s Bushka (see above).  Very fun to play casually, as the action starts immediately, but as with Bushka I find it quite confusing.  I like a more subtle opening phase, and as any reader of this blog will know, I always like larger boards.  Still, well worth a try!  There’s no denying it’s a unique game that’s packed with excitement.  7/10.

Killer Draughts — An attempt to alleviate the drawishness of Draughts, Killer forces kings to stop immediately on the square just beyond the piece it captures at the end of a capturing sequence, but only if that piece was also a king.  This simple change, derived from a similar rule in Thai Draughts, allows two kings to win against one in the endgame (as compared to four kings being required in the standard game).  The rule is simple to understand, and large endgame tablebases have already been calculated for this variant, making it possible to adjust one’s endgame knowledge through study.  In my view, a straightforward and effective option for tightening up the Draughts endgame.  9/10.

Lasca — The famous Chess player Emanuel Lasker invented this game, which is the columnar adaptation of English Draughts (or Straight Checkers, for my American compatriots).  Sadly the result is too constrained and dull, lacking all the verve and vitality of Bashni.  Still better than a kick in the head, and I imagine this game could be substantially improved if someone with design chops spent some time with it.  5/10.

Loca This tricky little number flips one’s Draughts expectations around, allowing men to capture like kings.  The result is a weird experience, with carnage starting very quickly.  I’ve only played this twice, and I don’t think I could possibly take this game seriously, but it was pretty fun.  7/10.

Territory Games

Amazons — This is one of those games that I admire, but don’t actually like playing that much.  Players move their Amazons around the board like Chess Queens, blocking off an adjacent square after each move; the last player to move wins.  In play the game becomes a territorial battle, and there is ample scope for delicate strategic play.  I just find it a bit dull in practice; the interaction between the Amazons is indirect, and slightly unsatisfying for me.  An amazing invention and clearly a great game, just not one that suits my temperament.  7/10.

cannons-bullets-1

Cannons and Bullets — This clever game allows players to place pieces on vacant hexes, but only if that hex is directly visible to a friendly piece; if a piece is visible to three or more friendlies, you may place a double-stack, which then can be fired as a cannon to capture enemy pieces.  I find this game easy to get into and remarkably fun in play; games are quick, generally speaking, but there is scope for some cute tactics.  Recommended.  8/10.

desdemona-1

Desdemona — This odd combination of Amazons and Othello is actually more fun for me than Amazons itself.  The addition of toggle-capture adds some much-needed interactivity to the game that I find enjoyable.  Desdemona has been through a few revisions to avoid degenerate play, but the final version works well and is recommended if, like me, you find Amazons just a little bit sterile.  8/10.

HexSygo — One of the few, elite members of the hexagonal-Go-games-that-actually-work club.  A straight-up translation of the Sygo rules to the hexagonal board results in a game with Go-like territorial vibes but a very unique tactical feel.  I haven’t played this game all that much but it’s on my list to study more deeply.  8/10.

Keil — I don’t like to be this negative, but I really, really dislike Keil.  Keil takes Go, known for its minimalistic visuals and high clarity, and transforms it into a game that makes no visual sense, where links between cells and stones are obscure and hard to follow.  In essence, in my opinion it takes Go and removes virtually all of the clarity and aesthetic beauty.  This is perhaps the only abstract game I’ve tried so far that actually upset me when I attempted to play it.  0/10.

Lotus-01

Lotus I mentioned this game in a previous post as an under-appreciated gem, and I stand by that assessment.  The odd board and Othello-style capture give this game a unique impression, and it’s certainly a better use of the Kensington board.  Check out the other post for more.  8/10.

MacBeth — Not much to say about this game other than the fact that it’s a translation of Othello to a hexagonal grid, and it works great.  I’ve had a few games of this, one of which ended in a last-second victory for me that had all the thrills of a close scrape in its parent game.  I liked this enough to program it for Ludii, so clearly I like it a lot.  9/10.

Medusa-10-01

Medusa — Another unusual territory game with an unusual board, Medusa again features in my other post with brief reviews of cool abstracts.  Medusa is like Lotus’ big brother, with a more deliberate, strategic feel.  Both games make fine use of Rosette’s clever conceit of granting eternal life to hexagonal formations of a single colour.  Give it a whirl!  8/10. (read more here)

othello1

Othello This game is of course a classic, present in millions of homes throughout the world, and is particularly popular in Japan.  Othello is one of the rare modern-ish abstracts to actually have a robust tournament scene, with regular world championships and ample high-level online opponents available.  The flip-flopping capture mechanism can be a bit opaque at first, but Othello is so instantly accessible and fun that it’s easy to break this out with non-gamers and have a good time.  My only lament is that the 10×10 version is so hard to find anywhere; I still occasionally trawl Japanese auction sites looking for a copy, and someday I hope to grab one at last.  This game often gets denigrated in the abstract community for some reason, but there’s a lot of depth here and it’s absolutely possible to take up Othello as a lifestyle game.  9/10.

tumbleweed-sample

Tumbleweed A rare example of a game birthed on the BGG forums that has gained some kind of community around itself, thanks to the herculean promotional efforts of the creator Mike Zapawa and superfan Alek Erickson.  Tumbleweed is a territorial game centred on a line-of-sight placement mechanic and the use of stacks of counters.  I initially was excited about this game, as it felt novel while still being comprehensible, but in subsequent play something bugs me about it.  I haven’t quite worked out what that is, but I suspect a few factors play into it: it’s hard to justify expending lots of effort on another territory game when I’ve invested so much in Go; OTB play is tough to organise due to the need for hundreds upon hundreds of counters; and large board sizes produce a much more interesting game, IMO, but most players prefer smaller sizes.  However, it’s definitely a significant and enjoyable game and I would recommend trying it; plus it has a tournament scene, which is a rare and precious thing.  8/10. (read more here)

Largest Group Cascading

Catchup-new-attempt-hex9-01

My order-9 Catchup board design.

Catchup — Easily the best game made by Nick Bentley, which regularly gets criticised by the BGG abstracts community, and Nick himself, because of the need to keep score.  This is unjust in my view, as the Catchup mechanism, in which players may place 3 stones in a turn instead of 2 if the opponent matches or exceeds their current score, is a brilliant innovation that gives this game an immediately apparent uniqueness and flair.  For a very long time it was only played on order-5 hex boards, which means games were so short that there was not much room for strategy.  Despite my love for this game I don’t enjoy playing it on the order-5 board.  Thankfully, now people seem to be playing on order-7 boards more often, which is a significant improvement.  I printed an order-9 board for myself, and I sometimes play at even larger sizes using Ai Ai.  In any case, I consider this one of the finest modern abstracts, and in a more sensible world it would have been published and would be getting played by millions of adoring fans.  10/10. (read more here)

Migong-test1

Migong — Here’s part of a post I wrote on this game in the BGG abstracts forum, in relation to the game captured in the screenshot above: “That was pretty fun. Shutting down opposing groups is enjoyable and there’s a lot of back-and-forth fights. I’m curious what high-level Migong would look like, and what sort of strategic play might develop.  Weirdly, I think I’d enjoy this much more as a physical game than as a MindSports app. It’d be visually satisfying on a clean board with acrylic pieces, and it’d be appealing to people who might not normally engage in a combinatorial territory game. There’s a satisfying ’embed the rules’ thing going on here that I think makes for a good candidate for a physical product.”    To add to that, I’m not normally a big fan of tile-laying games, but this one I enjoyed, not least because the patterns produced during play are visually attractive.  Digging up that post made me want to play again, which is a good sign.  Let’s call that a solid 7/10.

Permute 12x12 -- move40 -- Ac6d7-d7-01

Permute This is my game, and I like it, and I’m good at it.  But I’m not going to give it a numerical rating on my own blog post, because people do that on BGG sometimes and give themselves 10/10 and say their game is better than Go, and every time I see that I roll my eyes so hard I need to go see an ophthalmologist.  So instead I’ll just say it’s a fun game, inspired by the Rubik’s Cube, in which players build up groups of pieces by rotating them like the face of a twisty puzzle.  I’ve written a long post about this game which took a huge amount of effort, so please go take a look.  12×12 is the ‘beginner’ size, I recommend the 16×16 game for a deeper strategic battle.

Qascade — The sole member of the exclusive games-inspired-by-Permute club, Qascade adds the twisting mechanic to a placement game on the hex grid.  Combined with Christian’s one-bound, one-free opening protocol, we get a nice game of group-building with some tactical surprises due to the twisting element.  Fun to play, but in my extremely objective, not-biased-at-all opinion, Permute is more purely about twisting and therefore has a stronger identity.  7/10.

slyde16-10s-1

Slyde — One of the inspirations for Permute, this game tasks players with building up groups of pieces on the square grid by swapping them.  Each swap locks a piece in place, which was the mechanic I later used in Permute to ensure termination.  The game seemed to be doing relatively well at attracting players, until the creator stopped promoting it in favour of Tumbleweed.  8/10.

Connection Games

Gonnect A fine game that results from a very simple idea: what if we kept most of the rules of Go (except passing) and just changed the goal to a connective one?  The result is an interesting experience, and a nice example of how connection games can incorporate capture successfully.  Some players dislike how close games enter a ‘cold war’ phase where the game essentially becomes No-Pass Go, but I guess as a Go fan this doesn’t bother me too much.  Trying this is a no-brainer if you have a Go set lying around (and you should, you can play about a billion different games using a Go set).  8/10.

slither19-10s-1

Slither — Conceptually this game has great appeal; stones are placed and then slide around each other as players attempt to connect opposite sides of the board.  In practice however, I find it hard to get a grip on this game.  I made a bit of an effort to understand it with the help of David Ploog’s excellent guide, but still it didn’t quite click.  Ultimately I gave up, because there are many connection games out there that I can comprehend much better than this one.  There is a nice community of players on Little Golem, which is a big point in its favour.  I suspect I’d do better with this game if I could stare at it over a Go board with a real human opponent, somehow.  6/10.

Symple Hex — Combining Symple and Hex gives us this game, which applies Symple’s remarkable move protocol to the connection genre.  The result is enjoyable, but somehow not as compelling as the other connection games I like.  I think it may be that the growth mechanic of Symple feels a bit strange in a connective context, as my brain wants to make big groups but that’s not necessarily the right thing to do to actually win the game!  In any case, I’m glad this game was made, and I hope another Symplistic connection game might iterate on this idea in the future.  7/10.

Unification Games

loa-10s-1

Lines of Action This bonafide modern classic birthed a genre, and for many remains the best example of it.  Here players try to unite their checkers into a single group, using a strange movement mechanic in which pieces may slide according to the number of pieces present in the direction they wish to move along.  The game is simple to learn but getting one’s head around the basics can be challenging initially.  I intend to spend more time with this game, having learned a lot about it from David Ploog’s guide.  As usual I’d love to see this expanded to a larger board with more pieces too; the original is easily playable with a Checkers set, so I want one playable with a Draughts set!  8/10.

Looping Games

coil1

I’ll save you some more time here too — honestly, just play Havannah or Coil instead!  Granted they have other win conditions besides loop-building, but they’re excellent games so who cares.  Coil (above) is such a cool game to me that I even programmed it for Ludii… only for the BGG abstracts forum to start exploding shortly afterward with endless loop-based game designs which explicitly aim to replace it.  That’s some bad timing.

Race Games

Ordo — I’m not generally a huge fan of race/traversal games, but Ordo is a good one.  Designed by Dieter Stein, creator of a number of interesting games, this one features movement of whole groups, which is always satisfying.  I recommend it, but again I’m no expert on this genre.  7/10.

Blockade Games

Stalemate games.

Monkey Trap A cute Amazons variant themed around monkeys flitting around the board dropping coconuts on the board behind them.  Christian presents this as a game for kids, but I think it’s fun enough for adults, too.  The reduction of the decision space relative to Amazons makes it more accessible, but there are still enough options on a typical turn to allow for surprising tactics to happen.  A good game, and maybe a commercially viable one — a jungle-themed board with cute plastic monkeys and big chunky coconut pieces would go down a treat amongst younger players.  6/10.

pilare-initpos-MS

Pilare Eventually tried this thanks to encouragement from David Ploog.  An ingenious 2D Mancala-like game designed by Jorge Gomez Arrausi, creator of the also-ingenious Unlur.  In Pilare, players pick up stacks topped with pieces of their colour, and sow their contents over the board, until someone is left without any stacks they can sow and thus loses.  Each move is potentially long and filled with possibilities, so the game ends up feeling tactically rich and loaded with mysteries.  Someday I hope to play more of this.  9/10.

Configuration Games

A hugely underrepresented category on MindSports, and in the abstracts community in general.  The n-in-a-row genre is ripe for some innovation, but the abstracts community seems to have largely abandoned it as a lost cause in recent years, which is a shame.

Hexade — Christian Freeling’s take on the n-in-a-row genre follows Havannah and grants the win under multiple conditions: victory comes after connecting six of one’s stones in a line, triangle or hexagon.  Capture occurs exactly as in Pente, but capture is not an alternate win condition as in that game.  The result is an appealing, varied game of interesting tactics.  Loses a couple points from me on the basis that Pente doesn’t feel much less rich despite the simpler win condition.  Must play more though to determine whether this is an accurate impression.  7/10.

pente1

Pente Once a popular game with a robust tournament community, Pente is a bit obscure nowadays but players still congregate and many games still take place at Pente.org.  Players compete to be the first to get five stones of their colour in a line, but in Pente players may also capture pairs of enemy stones if they place their stones at either end of that pair.  Captures also provide an alternate win condition for the game.  I like this game a lot, to the point that I programmed it for Ludii, alongside the excellent variant Keryo-Pente.  9/10.

Games With Various Other Goals

hexsymple-sz12-p10-sample-game1

HexSymple — Translating the Symple protocol to the hex grid works great, and produces a game with a strikingly different feel; the additional connectivity options present on the hexagonal grid shift the focus more toward strategy than tactics.  As a devoted fan of Symple I enjoy this game just as much, and sometimes I think it may even be better than the original, but that feeling fluctuates depending on my mood.  Like Symple, it’s deep, engaging, tense, and hugely flexible and scalable.  Just play it.  10/10. (read more here)

Multiplicity — Some of you will know of Omega, a clever game that uses multiplicative scoring, leading to interesting consequences when working out how to build groups that achieve the best score.  Multiplicity uses this same scoring mechanism, but unlike Omega, players place one stone per turn, may play stones only of their own colour, and the game opens using Christian’s one-bound, one-free opening protocol.  The resulting game is tense and fun, and shares with Omega the strange feeling of needing to avoid certain connections or one’s score can drop significantly.  One downside is that keeping accurate score is a bit annoying in OTB play, but on the other hand, just playing using basic concepts of trying to build groups of the optimum sizes can get you pretty far.  7/10.

Xodd/Yodd — This pair of games tasks players with building less groups of their colour on the board, but players may drop stones of either colour, and at the end of the turn the total number of groups on the board must always be odd.  The result is a very unique playing experience, with some highly unusual tactics and appealing whole-board strategies.  Xodd (on the square grid) is generally considered more tactical, and Yodd (on the hex grid) more strategic, but both are very worth playing.  For what it’s worth, I slightly prefer Yodd, not least because the use of the hex grid helps stop my brain from mistakenly applying Go concepts to this game!  Elsewhere I proposed Snodd, a variant played on the snub-square tiling which has five adjacencies on each point, precisely between Xodd’s four adjacencies and Yodd’s six.  Like most of my ideas it had no impact at all, but I still think it has potential.  9/10.

Draughts Games Not Appearing On MindSports

I thought Checkers, Russian Draughts, and some other Draughts variants used to be listed on MindSports, but they don’t seem to be available for play anymore (or maybe they never were?).  I decided to review some of these games anyway, since I play several of them regularly.

checkers-snip-1

Excerpt from the monumental Checkers opening guide by GM Richard Pask, Complete Checkers (2020).

Checkers — Also known as Draughts, English Draughts, or Straight Checkers, this 8×8 classic suffered unjustly from Jonathan Schaeffer and his engine Chinook weakly solving the game in 2007 (it’s a draw).  This game-theoretic result has no real bearing on actual play between humans, and alternative opening options like 11-Man Ballot ensure the game has plenty of life left in it.  GM Richard Pask has written some excellent books on the game in recent years (like Complete Checkers above), and great YouTubers like AZCheckers continue to play, analyse and promote this venerable game.  The upcoming World Championship has a $32,000 prize fund, so previous reports  that competitive Checkers is ‘as dead as the dodo’ were a bit exaggerated.  I urge players out there to download Pask’s Checkers for the Novice and give the game a chance; Checkers remains a playable and enjoyable game to this day, and strong opponents are easy to find.  8/10.

Italian DraughtsDama Italiana is a unique take on 8×8 English Draughts with one massive rules difference: kings cannot be captured by men.  This single change has a profound impact on endgame play especially, and for that reason alone is well worth checking out for fans of Draughts/Straight Checkers.  There are some additional wrinkles to the majority capturing rule, which take some getting used to.  The Italian Draughts Federation seems highly active, and supports OTB tournament play in both the Italian game and 10×10 International Draughts.  I’ve heard tales of 10×10 Dama Italiana being played in some parts of the country, which intrigues me, but I’ve yet to find confirmation.  8/10.

Unfortunately, Christian chose to describe this game in an oddly nasty way:

“It bears testimony to the idea that Italians either prefer ‘complicated’ to ‘simple’ or cannot distinguish between ‘complicated’ and ‘complex’, and in either case are obsessed with hierarchy.  Checkers is simple and complex. Italian Draughts is complicated and complex. There’s nothing gained except the loss of simplicity.” (here)

I don’t believe we can generalise about an entire nation on the basis of their Draughts preferences.  Besides, the Italian Draughts Federation runs more events for International Draughts than they do for Italian, so this claim doesn’t make any sense — clearly Italians can, and do, appreciate the ‘simpler’ side of the Draughts rules spectrum.  Also Dama Italiana does gain something unique from the powerful kings, and the more complex capturing rules work in service of that new dynamic.

Russian Checkers Known as Shashki in its native land, Russian Checkers has a distinguished history of high-level competition and sophisticated analysis.  Russian Draughts is nearly 8×8 International Draughts, but men can promote to king in mid-move (!), and players are not obligated to take the capture sequence of maximum length.  These two simple changes create a game with a distinct character, well worth playing.  Russian Checkers also has a large literature associated with it, but unsurprisingly nearly all of it is in the Russian language; a large archive of classic books on the game can be found here.  Tournament play occurs all over the world and is supported by the IDF64.  Online opponents are readily found on PlayOK.  9/10.

pool-checkers-1

American Pool Checkers — A game criminally underrated in its home country, American Pool Checkers is a close relative of Russian Checkers, where men no longer promote in mid-move.  The game found some popularity for a time, particularly among black men in the American South.  I hope this game can maintain a foothold, particularly because it’s so rare that we recognise and support the contribution of minority communities to the world of abstract games.  Captivating articles like this feature in The Bitter Southerner certainly can’t hurt; how could we not love a game featuring top players with nicknames like ‘Iron Claw’ and ‘Big Willie’?  In the meantime, give it a try on Ludoteka9/10.

Brazilian Draughts — This is a straightforward reduction of 10×10 Draughts to the 8×8 board, all rules are otherwise the same.  The result is a quick and action-packed game, great as an introduction to the world of 10×10 Draughts, a quick blitz between longer games, or as a deep and rewarding pursuit in its own right.  Like Russian Draughts, Brazilian Draughts is played all over the world and high-level tournament play exists both online and offline.  9/10.

Russian and Brazilian Draughts have one noteworthy difference from 10×10 Draughts, besides the obvious: in 10×10, four kings are required to ensure victory against one, but in these 8×8 games, three kings can trap a lone king so long as they can occupy the long diagonal.  The key formation is known as Petrov’s Triangle:


So there we go, a great big pile of reviews and impressions of many of the games on MindSports.  I hope someone out there finds this useful, at least fodder for discussion if nothing else; but if not, at least I’ve collected all these thoughts in one place for my own future reference.

As you can tell from this sizeable list, there are quite a lot of games on MindSports.  Ideally the Chess and Shogi sections would be more comprehensive — it’s a bit funny to me that MindSports has stuff like Armenian Draughts but not Shatranj, Makruk or Sittuyin — but there’s tremendous variety there nonetheless.  I highly recommend checking it out for yourself, and I’m happy to accept challenges on the site; just be aware I sometimes disappear for long periods due to being overworked.

I’ve resolved before to write future articles about certain things, and have often failed to follow through, partly due to a general lack of interest from others and partly from my own lack of time and energy.  About all I’m willing to promise anymore is that, assuming I survive the next pandemic wave(s), I’ll keep writing about games here and there, when I can.

Tagged , , , , ,

Ancient Shogi Revival, Part II: The Big Ones

UPDATE 6 Nov 2021More sample games updated — Dai Dai Shogi, Tai Shogi, Mini Tai Shogi.

UPDATE 25/10/2021 — Some of the sample game GIFs on this page broke, possibly from WordPress again changing the size limit for displayed images.  I’m slowly replacing the sample games for Maka Dai Dai Shogi, Tai Shogi and Mini Tai Shogi with new, smaller GIFs.  These games were generated with much longer thinking times (120 seconds) so the games are also higher quality.

USEFUL TIP: WordPress handles GIFs in a weird way. When you want to see/download the full-size GIFs, click the image and then click the (i) button, and finally click the link to view the full-size GIF in another tab.


Since my previous post on our addition of multiple variants of Shogi to the Ai Ai general game-playing software, Stephen and I have been hard at work implementing even more ancient Shogi games.  If you thought the giant 15×15 Dai Shogi was over-the-top, you’ll be amazed what we’ve brought to you this time!

Before I show off what we’ve done, a few points to keep in mind when you download the new Ai Ai version with all these amazing games:

  1. We’ve done a huge amount of work in a relatively short period of time, and have done everything we can to catch mistakes and bugs, but some bugs will still be present!  If you find a bug or incorrect piece movement or something, please do point it out to us, and we’ll do our best to resolve the problem.  Also, some rules are still not quite there (repetition rules, for one), so please be patient.
  2. The AI is very much a work in progress!  Some of these games are truly gigantic, and as a result getting strong play out of the Shogi engine in Ai Ai is really, really difficult.  When you do play against the AI, be sure to give it plenty of thinking time; at an absolute minimum, give it 30 seconds per move for the 16×16/17×17 games, 1 minute for 19×19, and 2 minutes or more for 25×25.  If you have an old computer with a slow CPU you may need to amp those times up significantly.  If the thinking times are too low, you may find the AI repeats moves a lot as it struggles to find useful continuations.
  3. For those games which have a playing community and multiple possible rulesets (Tenjiku Shogi), we have not attempted to resolve any long-standing rules disputes, but instead have opted for being comprehensive.  All the games have options to choose different rules interpretations, piece movements, etc.  Essentially I decided to include a rule proposal or piece movement when that interpretation had been played before, and/or had some historical plausibility, and tried to remain neutral on the long-standing rules disputes in some of these games.  The hope is that players may try all these different variations in Ai Ai, and hopefully use those experiences to choose the best ruleset to play as a community.

With that out of the way, let’s get to the games!  For each one I’ve included some details about the options available to players in Ai Ai.  At the end of the article there is an FAQ section, so please check that for some additional useful tips and tricks for using Ai Ai, and some general questions about our Shogi implementations.

Tenjiku Shogi (16×16)

Tenjiku Shogi (天竺大将棋, or Tenjiku Dai Shogi in Japanese) is a remarkable historical Shogi variant that is one of the most popular of these ancient games among modern players (alongside Chu and Dai Shogi).  Tenjiku Shogi is played on a 16×16 board with each player starting with an army of 78 pieces of 36 different types.  The game stands out not only among Shogi variants, but among Chess-like games in general for its extreme tactical sharpness and super-powered pieces.  The word ‘Tenjiku’ means ‘Indian’ but is more often translated as ‘exotic’, because at the time of Tenjiku’s invention (sometime in the 16th-17th century, most likely), India was seen as a mystical and exotic place, the birthplace of Buddhism and the home of marvellous creatures like tigers and elephants.  Tenjiku certainly lives up to its name, and has some very unique pieces that give the game a distinctive flavour:

  1. Fire Demon — for a start, this incredibly powerful piece is very mobile; it can slide unlimited squares in six directions, or take three successive King-like steps in any direction to dodge around obstructions.  But much more notable is its ability to burn opposing pieces — when the Fire Demon ends its move, all adjacent enemy pieces are immediately removed from play.  This means the Fire Demon may kill up to eight pieces in a single turn (one on its destination square, and seven in the surrounding squares).  This ability even functions on the opponent’s turn; any enemy piece that ends its move next to your Fire Demon is immediately burned and removed from play, and that doesn’t count as your turn!  If your opponent lands their Fire Demon next to yours, their Fire Demon is burned, and it doesn’t get to burn any of your pieces before it is removed from the board.  Fire Demons can still be captured by any piece that lands directly on its square.
  2. Range-Jumping Generals — Four pieces in Tenjiku, the Great General, Vice General, Rook General, and Bishop General, have the ability to make range-jumping captures, where they may leap over any number of friendly or opposing pieces to land on an enemy piece and capture it.  To keep this amazing ability under control, these pieces are subject to a hierarchy, and may not jump over pieces above them in that hierarchy; this allows players to block these powerful jumps with careful defensive placement.  In the two main Tenjiku rulesets there are two different interpretations of this ability; more on that in the discussion of game options below.
  3. More powerful Lion pieces — Tenjiku includes all the pieces present in Chu Shogi including the remarkable Lion, which may move and capture twice in a turn.  In Tenjiku the Lion can promote to an even more powerful piece, the Lion Hawk, which may move as a Lion or a Bishop.  Tenjiku also includes the Free Eagle, a piece that may slide an unlimited number of squares in any direction, or may make a Lion-like double-move but only diagonally.

Combine all these powerful pieces and you get a game that despite its size is remarkably fast-paced and violent, with deadly tactical combinations possible right from the beginning of the game.  Most large Shogi variants have quite long and subtle openings, but in Tenjiku one opening mistake might lead to an embarrassingly early checkmate.  These unusual properties have enabled the game to develop a small but dedicated Western playing community, and as a result numerous resources are available for prospective players, including opening guides.

Rulesets and Options

Tenjiku Shogi has had a lot of attention over the years since it was introduced to the West in the 1980s by George Hodges.  Unfortunately, as with many of these ancient games, some important rules questions remain unresolved.  Modern players have developed two main rulesets that are in use, both of which are available in Ai Ai:

  1. Richard’s PBEM Server ruleset: the yearly Tenjiku Shogi Championship tournament is played on Richard’s PBEM Server, and uses this ruleset.  In brief, the main distinguishing features here are: Fire Demons may slide unlimited squares diagonally or vertically; range-jumping generals may capture Kings or Princes while jumping, allowing for some very early smothered mates to occur; and the Heavenly Tetrarchs do not have a vertical sliding move (and in fact use an odd set of movements that seems to be a mistake, so you may want to use the TSA Tetrarch variant instead).  Early checkmates are very common in this ruleset.
  2. Wikipedia/Chess Variant Pages ruleset: these rules have been promoted by HG Muller, and produce a somewhat less violent Tenjiku experience than the PBEM ruleset.  In this ruleset: Fire Demons may slide unlimited squares diagonally or horizontally; range-jumping generals may not capture Kings or Princes while jumping, eliminating the early smothered mates; and the Heavenly Tetrarchs have a vertical sliding move, making them a clear upgrade over the Chariot Soldiers from which they promote.

Various spirited arguments have taken place, and continue to take place, over these key rule interpretations.  In Ai Ai we have chosen instead to implement both options, to give players a choice of how they want to play — simply load TenjikuShogi(PBEM).mgl for the PBEM ruleset, or TenjikuShogi(WP).mgl for the Wikipedia/Chess Variant Pages rules.  In the hope of helping the community to resolve these rules discussions, we have also included a customisable version (TenjikuShogi(Custom).mgl), which allows players to customise their ruleset with the following options:

  1. Fire Demon move — you may choose whether the Fire Demon slides diagonally and vertically (PBEM version), or diagonally and horizontally (WP version).
  2. Range-Jumping Generals — you may choose whether they can capture the Royal pieces (King and Crown Prince) by jumping (PBEM) or not (WP).
  3. Heavenly Tetrarchs move — you may choose from four (!) different move options for the Heavenly Tetrarchs:
    1. PBEM move — no vertical slide, may never move to the eight adjacent squares, may capture without moving on adjacent squares, adds a three-square vertical jump (which seems like a mistake)
    2. Wikipedia/Chess Variant Pages move — vertical slide, may never move to the eight adjacent squares, may capture without moving on adjacent squares
    3. TSA move — the original move advocated by The Shogi Association — no vertical slide or vertical jump, may never move to the eight adjacent squares, may capture without moving on adjacent squares
    4. Sho Shogi Zushiki move — vertical slide, may never move to the eight adjacent squares, and may NOT capture without moving on the adjacent squares
  4. Lion Hawk move — you may choose two variants of the Lion Hawk:
    1. TSA move — the Lion Hawk does NOT have Lion Power, but instead may move as a Bishop or step twice as a King in any direction, but may only capture once during this two-step ‘area move’
    2. Modern move — the Lion Hawk moves as a Bishop or a Lion, with full Lion Power
  5. Free Eagle move — you may choose two variants of the Free Eagle:
    1. TSA move — the Free Eagle may move as a Free King, or may jump two squares orthogonally
    2. Modern move — the Free Eagle may move as a Free King, or may make two successive one-square diagonal steps, which can include making two captures, capturing an adjacent piece and moving back to the starting square (igui capture), or passing a turn by moving to a diagonally adjacent square then back to its starting point.
  6. Allow Zone-Passing — you may choose whether or not to allow zone-passing; if zone-passing is allowed, that means that Lion-Power pieces may promote if they make a Lion-style move into the promotion zone and immediately back out again on the same turn.
  7. Water Buffalo promotion burn — if Water Buffalo promotion burns are allowed, that means that when a Water Buffalo promotes to Fire Demon, it may immediately burn all adjacent enemy pieces on the move where it promotes.  This is suggested in the Wikipedia/Chess Variant Pages ruleset.

That’s a lot of options!  I’m 100% sure that some members of the Tenjiku community would prefer that I adopted only their preferred ruleset/piece movements, but again I have tried to avoid wading into any long-standing rules disputes here.  I opted to go over all Tenjiku-related materials I have and simply include any rules/movements that have some historical backing, or have been played before, or advocated seriously as a proposed change to the rules, so that players may try them all and make an informed choice.  I did decide not to include the original TSA rules for the range-jumping generals, however, as these have been shown to give an easy win to Sente.

Promotion Rules and 50 Move Rule

Note that I have decided to offer all three main Tenjiku variants with the Chu Shogi promotion rules; this is mainly because there seems to be some momentum lately to adopt these rules in the PBEM Server version, and the Japanese Wikipedia rules state the Chu Shogi promotion rules are applied in this game.  If anyone wishes to use modern Shogi promotion rules instead, simply open the appropriate .mgl file in a text editor, and find a line that looks like this:

“promotionRule” : { “rule” : “ChuShogi”, “allowIguiPromotion”:false },

Remove the “rule” : “ChuShogi”, part and you will be using modern Shogi promotion rules.

I also decided not to include any equivalent to the 50-move rule in Chess, mainly because these rules are a modern invention and definitely not part of ancient Shogi, and also because we have no idea what a good move threshold would be in these huge games.  However, if the community does want to use a 50-move rule equivalent, this can be added to any Shogi variant by opening the appropriate .mgl file, finding the section called “endCondition”, and within the square brackets adding another line that looks like this:

{ “condition”:”boredom”, “who”:”all”, “result”:”draw”, “moves”:50 },

Simply change the move number to your preferred option, and that’s it!

In any case, I’m a strong believer that debating game rules without playtesting them is pretty pointless, so I’m hoping that having all these options available will help the various rules debates to eventually get resolved.  With these different rules variations now playable online as well as against the AI, players may thoroughly test them and decide together how they want to play Tenjiku in the modern age.

Tenjiku Sub-Variants

For players who are interested in playing Tenjiku but find the full-sized game intimidating, we’ve also included two smaller modern variants of Tenjiku Shogi:

Nutty Shogi (13×13)

Nutty Shogi was invented by HG Muller in 2015, and is a very enjoyable reduction of Tenjiku Shogi to a more compact 13×13 board.  All rules remain the same, so it’s a straight subset of the full game.  In this implementation players may choose from the options present in the custom Tenjiku Shogi version, in order to match their preferred ruleset for the larger game.   Note that the game seems to have been designed to use the Wikipedia/CVP rules, so I recommend using those options.

Makyou Shogi (12×12)

I’m a fan of Nutty Shogi, but find it mildly inconvenient in that no historical Shogi variants use a 13×13 board, so it’s not easy to play physically even with my collection of Shogi variant pieces and boards.  So I created Makyou Shogi, a reduced version of Tenjiku designed to fit on a 12×12 Chu Shogi board.  Makyou is a Japanese word meaning ‘demon-infested place’, which felt appropriate for this tight board packed with powerful pieces!  Note that the piece mix is a bit different from Nutty Shogi.  Also, this game is still being tested, so the starting array may change over time.

Makyou Shogi allows players to tweak any of the rules/piece options present in the Tenjiku Shogi customisable implementation; having said that, I strongly recommend using the Wikipedia/CVP rules in this game, as the PBEM pieces make the game too tactical with little room for strategy.

Dai Dai Shogi (17×17)

With Dai Dai Shogi, we enter the realm of the truly gigantic Shogi variants.  Players start the game on this huge 17×17 board with 96 pieces each of 64 different types (4 additional types appear only via promotion).  Among those pieces are the ultra-powerful hook-moving pieces — the Tengu and the Hook-Mover — which may make two successive moves as a Bishop or a Rook respectively, making a right-angle turn after the first one (but may only capture once).

Dai Dai Shogi is the smallest of the ancient Shogi games to use a promotion-by-capture rule.  Instead of promoting upon reaching the enemy camp, pieces in Dai Dai Shogi promote immediately after they capture any opposing piece (if the piece has Lion Power, the promotion happens at the end of their complete move).  Promotion by capture is mandatory and cannot be deferred.  Unusually for a Shogi variant, most pieces in Dai Dai Shogi do not promote, including Pawns; this relative lack of promotable pieces together with the promotion-by-capture rule gives the game a quite different feel from the other historical Shogi variants.  The promotion-by-capture rule has a profound impact on tactics, too, as players must now be careful that any capture exchanges don’t leave the opponent a strong promotion at the end of the fight.

The Dai Dai Shogi starting array is packed with pieces and very asymmetric, with many pieces only appearing once rather than in pairs.  The initial position is also arranged quite differently from the other large Shogi variants (with the exception of Tai Shogi, to an extent) — the strongest pieces are at the back of the player’s army, rather than the front. This means opening play is quite slow and subtle, as players try to develop their pieces soundly without leaving any easy captures for the opponent to use for promotion, while also trying to free their most powerful pieces.  The endgame can become pretty hectic, as the hook-moving pieces finally come out to play and their incredible mobility makes them a deadly threat.  All told the game is thoroughly engrossing, deeply strategic and yet filled with rich tactics:

”This is the writer’s personal favourite of the [large Shogi variants], owing to the tremendous variety of pieces, the wealth of strong pieces and weak pieces with strong promotions, and the asymmetrical opening setup…. What often happens in practice is that the entire board opens up into the most complex tactical struggle seen in any of the forms of Shogi.” (R. Wayne Schmittberger, writing in Shogi Magazine in 1981)

Rulesets and Options

Fortunately, unlike Tenjiku Shogi there are relatively few major rules disputes in Dai Dai Shogi, if only because very few people have played it.  As a result our implementation has just two main incarnations: one which includes the troublesome Great Elephant piece in four variations (DaiDaiShogi(complete).mgl); and another which excludes the Great Elephant piece entirely, which apparently is the case in two historical sources (DaiDaiShogi(noGE).mgl).  In both versions the player may choose between two variations of the Furious Fiend’s move, as well.

Alongside the Elephant variants described in the translation notes article, in this final version we also included a variant move sometimes used in Tai Shogi,  which allows the Great Elephant to  move as a Lion Dog or slide up to 5 squares horizontally or diagonally backward.  This move is recommended by the Japanese Chu Shogi Association (Chu Shogi Renmei).

Sub-Variant: Cashew Shogi

Invented by HG Muller in 2015, Cashew Shogi is a reduction of Dai Dai Shogi onto a smaller 13×13 board.  This variant includes most of the characteristic pieces of the larger game, but substantially reduces the number of moves one has to remember.  Despite the size reduction, the AI vs AI test games I’ve tried are often not that much shorter than the full-sized games of Dai Dai Shogi, and sometimes are substantially longer!  I suspect this is mostly due to the size difference, however; the much larger board size in Dai Dai Shogi gives the AI a hard time, so it tends to make more serious blunders during play.  In Cashew Shogi the AI can find stronger moves, and is less likely to blunder its way into a quicker loss.

Our Cashew Shogi implementation allows players to choose their preferred form of the Great Elephant and Furious Fiend, as in Dai Dai Shogi.

Maka Dai Dai Shogi (19×19)

Let’s get my biases out there straight away — Maka Dai Dai Shogi is a very cool game, and I’m overjoyed that it’s now playable in Ai Ai.  This is a truly immense game, played on a 19×19 board with 96 pieces of 50 types in each player’s starting army, and it uses a promotion-by-capture rule like Dai Dai Shogi (more on this below).  What makes Maka Dai Dai Shogi stand out is the presence of several unique mechanics and pieces that give this enormous game the feel of a giant mythical battle; this is no coincidence, as the game is clearly heavily influenced by Buddhist mythology:

  1. The All-Powerful Emperor — unlike the other Shogi games, in Maka Dai Dai the King can promote!  If your King captures an enemy piece, he may promote to Emperor, perhaps the most powerful piece in any Chess-like game.  The Emperor can instantly jump to any square on the board, including squares occupied by enemy pieces, but it may not enter protected squares or capture a protected enemy piece.  Checkmating the Emperor seems impossible, but the hook-moving pieces can make it happen.
  2. The Deva/Dark Spirit/Teaching King/Buddhist Spirit — At the start of the game the Kings are flanked by a Deva and a Buddhist Spirit, two incredibly weak pieces with very awkward asymmetric 1-square moves.  But when a Deva captures an enemy piece, it becomes a Teaching King, a super-powered piece that may move as a Lion Dog or a Free King; likewise, the Dark Spirit becomes a Buddhist Spirit, which moves as a Lion or a Free King.  More intriguingly, these pieces are contagious — that means that if an enemy piece captures your Deva or your Teaching King, it immediately becomes a Teaching King, or if it captures your Dark Spirit or Buddhist Spirit it immediately becomes a Buddhist Spirit.  Consequently these pieces are very difficult to eliminate from the board permanently, unless they are taken by another Teaching King/Buddhist Spirit or by a royal piece (King, Emperor or Prince, who ignore the contagious aspect and simply promote normally).  Some Shogi historians believe the Buddhist Spirit and Teaching King may represent the Buddha and the Lotus Sutra, respectively.
  3. Promotion by Capture — as in Dai Dai Shogi, pieces in Maka Dai Dai Shogi promote when they capture something, so there are no promotion zones (or, alternatively, we can consider the entire board the promotion zone).  However, there are two variants of this rule: in one version, suggested by Wikipedia and the Chess Variant Pages, the capturer may choose whether to promote or not, unless the captured piece is promoted, in which case they must promote; in the other, suggested by the TSA rules pamphlet and Japanese Wikipedia, pieces must promote when they capture, as in Dai Dai Shogi.  In either case, Maka Dai Dai moves away from the Dai Dai Shogi model and allows nearly all pieces to promote, more like the other historical variants.  Many short-range pieces promote to ‘Free’ versions of themselves, which move in the same directions but extend all one-step moves into unlimited sliding moves.

These elements combined give Maka Dai Dai Shogi the strategic depth and nuance of the other large games, but punctuated with moments of extreme dynamism: the Emperor can teleport around the board savaging his enemies; the two powerful, mystical spirits stalk the board and never stop hounding one player or the other; and even the weakest pieces can promote into powerful board-spanning threats.  I honestly can’t recommend this game enough, it’s huge and crazy but very very interesting and enjoyable to play, and just packed full of cool pieces.  Here are my picks for the top ten coolest pieces in Maka Dai Dai Shogi, from strongest (Emperor) to weakest (Free Bear):

Fortunately, it seems I’m not the only one who developed a bit of an obsession with this game.  In the historical documents there is a long chant given that helps players to remember the opening setup, which suggests the game was popular enough that players developed these kind of mnemonics to help them get the game started more quickly.  A research group in Osaka, led by Professor Tomoyuki Takami, has been studying this game for years, generating a lot of lively debate about its origins, influences and rules; Professor Takami has also invested a lot of effort in promoting the game.  I have a fervent hope that one day these pockets of interest in Maka Dai Dai Shogi will blossom into a full-blown playing community, and this Ai Ai implementation is my attempt to help that process along.

As a side note, Prof Takami’s group has developed a set of rules for what they believe to be an earlier form of Maka Dai Dai Shogi, which they call Maka Dai Shogi.  This game also seems very interesting, but we haven’t included this in Ai Ai because it would involve programming numerous variant rules and pieces, and the rules are not stable and seem to change fairly frequently as Prof Takami finds new evidence in the historical documents.  At some point we may include it, when I have time to reach out to Prof Takami and get a comprehensive set of rules.

Rulesets and Options

Maka Dai Dai Shogi has relatively few options to worry about when starting a game:

  1. Promotion Rule — players may choose whether promotion on a capture is compulsory, or whether it is generally optional but becomes compulsory when the captured piece is promoted.
  2. Furious Fiend move — as in the other large games featuring this piece, players may choose the old TSA Furious Fiend move (Lion + 3-step slide move) or the more current interpretation (Lion + Lion Dog).

That’s it!  Note that the old TSA rules included a variant Teaching King move which was nonsensical, as it had the piece moving like a Free King or using a 3-step slide, which completely overlap.  The one historical Japanese source I have available clearly indicates a Lion Dog move plus Free King move, and all other sources I can find suggest this move, so I’ve opted not to include the TSA move here.

Sub-Variant: Macadamia Shogi (13×13)

Invented by — you guessed it — HG Muller in 2015, this 13×13 reduction of Maka Dai Dai Shogi retains all the most distinctive pieces in the game and removes most of the weaker ones.  The consequence is a quite intense game, where a nerve-wracking opening phase tends to develop into a rather intricate middlegame as the players try to outfox one another with their remaining nimble power-pieces.  I’ve enjoyed my time with this game quite a bit, and definitely recommend trying it.

Our implementation of Macadamia Shogi allows players to select the same options as the main game (promotion rule and Furious Fiend move).

Sub-Variant: Hishigata Shogi (19×19)

Hishigata Shogi was invented by Sean Humby in 2005, and attempts to slim down Maka Dai Dai Shogi in a different way.  In this game the full-size 19×19 board is retained, but the initial position excludes nearly all the weaker pieces, and the King is placed much closer to the centre of the board and surrounded by his army.  The result is an extremely fast-moving game, as the power-pieces swoop dangerously around the board, picking away at the enemy King’s dense pack of protectors until they can find a fatal weakness.

Personally I substantially prefer the original game to this variant; Maka Dai Dai Shogi shines partly because the deadly power-pieces are embedded in vast armies of weak-yet-still-important short-range pieces, balancing out their ridiculous power somewhat thanks to the levelling effect (referred to in my first post on our Ai Ai implementations).  Hishigata abandons much of that, and I feel the result is a bit less nuanced.  However, I can imagine some players will vehemently disagree with me and will enjoy the frenetic tactical explosion this variant offers, so of course I include it here for everyone to try for themselves!

Upon starting a game of Hishigata Shogi players may choose the promotion rule to use, but there is no choice of Furious Fiend moves as that piece doesn’t appear in this game.

Tai Shogi (25×25)

Tai Shogi is a truly immense game, and is the second-largest Chess-like game ever created (36×36 Taikyoku Shogi being the largest).  On its vast board of 625 squares, players battle it out with dense armies of 177 pieces each, with 93 different piece types present in the initial position.  Tai Shogi essentially combines the pieces present in Dai Dai Shogi and Maka Dai Dai Shogi, and adds an additional nine new types of pieces.

Until now I haven’t had the opportunity to play a full game of Tai Shogi, so I’d always seen it as phenomenally impressive, but probably too big to be truly playable.  However, since implementing it in Ai Ai I’ve played with it quite a bit, and I have to say I was wrong; Tai Shogi may be huge, but it’s definitely playable!  Not only that, the sheer size and scale makes the game feel quite epic, and I’ve enjoyed my time with it so far.  For those of you out there who are familiar with Dai Dai Shogi and Maka Dai Dai Shogi, please do give this game a try; the learning curve will be small, since there are only nine new pieces, and you’ll be able to develop some basic strategies based on your experiences with those games.

Tai Shogi has a couple unique properties of note:

  1. No Kings on the board — instead of starting with a King, players in Tai Shogi start with both an Emperor (!) and a Crown Prince on the board, both of which must be captured to win.  Since the Emperor can jump instantly to anywhere on the board, that means it effectively serves as a potential extra protector for every friendly piece, adding an additional tactical wrinkle to any battles taking place.  Games also tend to run long (2,000 moves is a typical length for a game), since the Emperor is far too powerful to get checkmated in the early stages when few pieces are developed and available to attack it.
  2. Full complement of hook-movers — Tai Shogi players have at their disposal the Hook Mover (double Rook), Capricorn (double Bishop), Tengu (double Bishop + single orthogonal step) and Peacock (double Bishop, but only forward).  With six hook-moving pieces on each side of the board, long-distance captures are always a threat, especially later in the game as the board empties out a bit.  However, players must use these pieces cautiously — some will demote on capture, and they can’t be spent carelessly as otherwise checkmating the opposing Emperor will become very difficult!

Playing Tai Shogi is quite an experience, and it’s most definitely the most wargame-like of the many Shogi variants I’ve played so far.  The starting armies are so huge that the board tends to have multiple local skirmishes happening at the same time, and the advancing Pawn lines backed by swarms of generals and mythical beasts reminds one of the clash of phalanxes of ancient soldiers.  The game clearly has scope for incredibly varied strategies, so I’m looking forward to playing more so I can begin to appreciate its subtleties.

Rulesets and Options

Tai Shogi has two main rulesets in use, which substantially change the gameplay:

  1. Japanese Wikipedia rules — These rules essentially combine Dai Dai Shogi and Maka Dai Dai Shogi in their entirety.  Most of the pieces from both games are included, and they promote as they do in their parent games, meaning that nearly all pieces in the game promote.  Promotions are compulsory upon capturing an enemy piece, as in Dai Dai Shogi.
  2. TSA rules — In the TSA version of the game, promotions are limited to about 1/3 of the pieces in the starting array, similar to Dai Dai Shogi (Pawns don’t promote here, either).  Promotions are also compulsory on capture, like Dai Dai Shogi and the Japanese Wikipedia rules.

The additional 60 or so promotions available in (1) make the game feel substantially more hectic than the TSA rules, and since I don’t have the historical sources available to decide on way or the other, I decided to include both rulesets in separate files so that players may choose for themselves.  Ruleset (1) is in TaiShogi(JWP).mgl, and ruleset (2) is in TaiShogi(TSA).mgl.  Try them both and see what you think!

Note that English Wikipedia has yet another set of promotions, but the entire page appears to be in a state of flux and so I don’t consider those suggested promotions reliable enough to include as a third option.  Similarly, the German Chu Shogi Association appears to suggest yet another variant where Crown Princes may promote to Emperors, but I can’t find much support for this idea elsewhere.

Players may also choose between variant piece movements for several pieces:

  1. Great Elephant — this has the same four move options as in Dai Dai Shogi
  2. Furious Fiend — this has the same two variants as in other games including this piece
  3. Fierce Eagle — players may choose between the TSA move given for this piece and the Japanese Wikipedia move
  4. Great Dragon — players may choose between the TSA move and Japanese Wikipedia move

Given the sheer size of this game, I wouldn’t worry too much about which moves you choose; the changes between each option are minor and will have little to no impact on the course of a full game on this massive board, particularly when there are numerous other pieces with substantially more power flying around.

Sub-Variant: Mini Tai Shogi (15×15)

This work-in-progress variant is my own creation, and is an attempt to shrink down the Tai Shogi experience onto a smaller 15×15 board.  The most exciting pieces from Tai Shogi are all present, and players start with Emperors on the board and multiple hook-movers, plus Poisonous Snakes and Old Kites that can promote to hook-moving pieces in the endgame.  My test games thus far have lasted anywhere between 400 and 1,000 moves, so still a long haul but substantially shorter than Tai Shogi.

At the start of the game, players may choose variant moves for the Great Dragon, Furious Fiend and Great Elephant.  Players can also decide to replace the Emperors with Princes, if they want a less mind-bending endgame; however, I strongly recommend using Emperors to get the full Tai Shogi experience!

As with Tai Shogi, there are two versions of Mini Tai Shogi — one following the promotions set out in Japanese Wikipedia, the other following the TSA rules.  The Japanese Wikipedia version has nearly all pieces promoting, while the TSA version has more limited promotions available.

Mini Tai Shogi is definitely a work-in-progress, but so far I’ve found it enjoyable to play.  Tai Shogi is such a huge game that reducing it down to 15×15 means eliminating a huge number of pieces from the board; over time I will experiment with different piece choices and arrangements for Mini Tai, and may even produce a second smaller variant with a very different set of pieces drawn from Tai Shogi.

Emperor Endgame Trainer

Maka Dai Dai Shogi and Tai Shogi add a new endgame twist over the other ancient Shogi games — the mighty Emperor, who can jump instantly to any unprotected square on the board.  At first this piece is very confusing to deal with, and checkmating it can seem almost impossible.  To make the transition to Emperor endgames a bit easier, and to open up a way to have some quick fun with all these crazy Shogi pieces, we added an Emperor Endgame Trainer to Ai Ai that helps you learn how to cope with Emperor endgames.

When you load up the trainer (EmperorEndgameTrainer.mgl), you’ll be given a menu where you may choose five different pieces for Sente to use to try and checkmate Gote, who starts with an Emperor and two Gold Generals.  If you have Ai Ai play as the Emperor side, you can experiment with different piece combinations and learn how you can checkmate in these wild new endgames.

Rulesets and Options

Upon loading the trainer, you will be able to choose five pieces for Sente to use, from the entire current list of available Shogi pieces in Ai Ai.  However, to keep things applicable to the actual games that use the Emperor, only the pieces available in Maka Dai Dai and Tai Shogi can promote, and they use the promotion-by-capture rule.  At the start you may also select whether to have promotion-by-capture as compulsory (like Tai Shogi) or optional (like Maka Dai Dai Shogi).

You may also choose whether one or both sides have a King or an Emperor, if you just would like to gain some experience with some of the Shogi variant pieces in more normal endgame situations.  The board size is configurable too.

Note that if you leave piece selections on ‘Random’, you may occasionally get starting positions where the enemy Emperor/King can be captured immediately; for example, selecting a Bishop hook-mover like a Capricorn, Tengu or Peacock as Sente Piece #2 will cause this.  If that happens, you can click the Game menu and select Game Parameters to choose different pieces and try again.

Shogi-Inspired Chess Variants

For those die-hard Chess players who miss the classic FIDE pieces and those wonderful FIDE Pawns, don’t fret!  I’ve also put together some highly experimental Chess variants that try to capture the large Shogi feel in a distinctly Chess-flavoured package.  These games use Chess pieces and popular fairy pieces, but adopt large Shogi conventions like very dense starting setups, promotions for many other pieces besides Pawns, and eschewing castling in favour of starting the Kings in strong fortresses right from the beginning.

All three games are really just intended for fun, so please don’t take them too seriously; having said that, I have playtested them quite a bit and they’re playable and not obviously broken.  The starting positions have been tweaked and tuned to avoid any obviously weak squares for either side to target in the opening, the piece mixes have been adjusted following test games, and promotions have been swapped around in each game to try and encourage the use of certain pieces.

Rider Romp (10×10)

This game is sort of a bridge between 10×10 Chess variants like Grand Chess and Opulent Chess and the large Shogi experience.  The opening setup is denser than most 10×10 Chess variants, and the pieces are more powerful and many are capable of promotion.  The four middle Pawns have been replaced with Stewards, variant Pawns that can move and attack in four directions, which are quite robust defenders; in early testing, a line of ten normal Pawns proved to be somewhat feeble resistance against the onslaught of all these power-pieces.

The name comes from the game’s theme of promoting one’s pieces to stronger forms that include additional ranging moves.  Some of the promoted pieces are ‘rider’ pieces, a term originating from fairy Chess which means they extend a basic move by allowing the piece to perform it repeatedly.  Nightriders are a classic example — they are Knights that can perform repeated Knight leaps in the same direction in one turn.

Rider Romp starts out fairly tactical, with long-range attacks flying everywhere; just mind the Steward Pawns in the middle four files.  Once the board empties a bit promotion really becomes a focus.  These games use promotion zones like the large Shogi games, so players will have to try to advance their own pieces to achieve strong promotions, while trying to guard their own huge promotion zone against incursions from enemy pieces with high-impact promoted forms.  In my tests a game played out to the bitter end takes about 150-250 moves (using Shogi-style move counting, not Chess style!).

Neutronium Chess (12×12)

Now things get a bit crazier.  Neutronium Chess gets its name from the ridiculously dense material that makes up neutron stars, because this game is absolutely packed to the brim with pieces at the start of the game, and there are 33 distinct piece types available for use.  The piece lineup includes various strong leaping pieces, some with additional orthogonal or diagonal step options to add flexibility, and powerful sliding pieces as well.

My favourite pieces in this game are the Griffon and Manticore, which are bent slider pieces — the Griffon takes one step as a Bishop then can slide unlimited squares outward as a Rook, while the Manticore takes one step as a Rook and then can slide outward as a Bishop.  This gives them some very surprising attack options, but also gives them a fundamental asymmetry; Griffons and Manticores can’t retreat the same way they advanced, so they must be deployed with finesse.   This game (and Dai Chess) also contains other bent pieces, including the Crooked Rook and Crooked Bishop that move in zig-zag patterns, and the Ship (invented by Jean-Louis Cazaux here), a vertical-only Griffon.

As with Rider Romp, many pieces promote in Neutronium Chess, so watch out for strong pieces getting even stronger in the endgame.  In play this game has an unusual feel, with the phases of the game proceeding in a Shogi-like progression, but the high-powered pieces create complex, long-range tactical shots more reminiscent of Chess.  I’m still not sure it’s ‘good’, as such, but I do have a lot of fun playing it.  A typical game lasts anywhere from 200-400 moves; the sample game shown above is on the shorter side, finishing at 261 moves.

Dai Chess (15×15)

Dai Chess takes the next logical step and increases the board size to a massive 15×15.  This game uses the exact same starting squares for pieces as Dai Shogi, but with all different pieces.  All Pawns start the game protected, so there are no immediately obvious weak points to target in the opening.   Players start play with 65 pieces each, and there are a total of 46 piece types available for use.  Given the larger board size I wanted to try for a more strategic feel in this game, so some of the more powerful pieces from Rider Romp and Neutronium (like the Griffon and Manticore) now only appear via promotion, calming the early stages of the game somewhat.  One aspect I’m still working in is the placement of the colourbound pieces, which is troublesome due to the board having odd dimensions rather than even, so keep an eye on that in future revisions.

Dai Chess really amps up the large Shogi feel of this series of variants, not just by using the Dai Shogi position but also by adding more promotions.  Nearly every piece promotes, and in keeping with what we expect from Chess, all promotions are an improvement in strength; I’ve avoided any demotions in the style of Dai Dai/Tai Shogi.  This game makes use of a lot of leapers, so good tactical vision is necessary to avoid early loss of strong pieces.  In the endgame, sliding pieces become dominant as they can promote much more easily, which again is reflective of the large Shogi inspiration.

A typical game of Dai Chess lasts about 400 moves, sometimes substantially more; the sample game above lasted 495 moves.  Ai Ai can actually play this game reasonably competently at 15-30s thinking time, but it can end up being a bit greedy in the opening; if you want a greater challenge, trying playing a cautious, positional opening for Ai Ai before turning the AI on.  Be careful though — for human players all these weird leaping pieces can make it hard for us to spot good moves, but the AI doesn’t have that problem!

Expanded Chess/Get Bent/Symmetric Sissa

If you’d like to experiment with those interesting bent-slider pieces some more, you can also check out my variant called Get Bent, also added to Ai Ai in this release, which includes a whole mess of them on a 10×10 board; it’s a very fast-paced tactical game.  I mainly made it to get familiar with these odd pieces, so it’s just for fun.  Pawns promote to extremely powerful Griffon/Manticore compound pieces, so watch out for that!

We have also included Expanded Chess, a nice variant by Daniel Zacharias which has Griffons and Manticores as well, and Symmetric Sissa, a showcase for the multipath Sissa piece, which on each move must take the same number of steps as a Rook and a Bishop (in either order).

None of these three variants have anything to do with Shogi, but since they’re in the new release I felt it was worth highlighting them anyway!


Next Moves

Sorry for the insanely long post, but as you can see, Stephen and I have done a whole lot of work on these games over the last few months.  As I mentioned at the beginning of this article, bugs will very likely remain in one or several of these games, but we’ve at least reached a stage where the games all function and we’ve incorporated every rule/piece option that seems plausible.

From here, we hope you Shogi fans will give these games a try, let us know how you find them, and maybe organise a tournament or two.  If/when any bugs crop up, please let us know of course!

I’m sure some folks would prefer there were less options to deal with in each game, but honestly I can’t see any other way to go; the alternative would be to decide unilaterally how these games should be played, but the information we have on the historical rules is very fragmented, so I don’t feel qualified to do that.  I also don’t want to take sides on the online debates regarding these rules, which have been going for far longer than I’ve known about these games.  In any case, my hope is that these options can be trimmed over time as the community tests them out and reaches a consensus on what pieces/rules are most enjoyable.  Similarly, as I get my hands on the historical sources I’ll revise things here and there as needed.

After this release we’re taking a little break from Shogi programming, but at some point down the line we do hope to bring in the legendary Taikyoku Shogi as well.  That will take a lot of research on my part, given there seem to be major differences in the rules in each of the three main historical sources, but I look forward to learning more about that massive game along the way.

In the meantime, go play some ancient Shogi, and have fun!


Frequently Asked Questions

“How can I choose a Shogi variant to play?”

On the File menu in Ai Ai, click ‘Choose Game’.  A new screen will open with a file browser on the left and a preview display on the right.  Open the ChessFamily folder on the left, then open the Shogi sub-folder.  Within that you will modern 9×9 Shogi, and two sub-folders, one for Historical variants and another for Modern variants.  Pick your favourite, double-click it and Ai Ai will load it up!

Alternatively, hit Ctrl+Shift+F (Cmd+Shift+F on Mac) and start typing the name of the game, and you should be able to load it that way.

“How can I play online?”

Instead of running ‘ai ai.jar’, run ‘online.jar’ which is included in your Ai Ai folder.  Once it starts up, click the button on the left to create an account, then click the link in the subsequent confirmation email (check your spam folder if you don’t see it).  Once your account is confirmed you can login and start playing!  Ai Ai supports correspondence-style online play.

“I’m playing Tenjiku Shogi, how do I use igui capture with my Heavenly Tetrarchs?”

Click and drag the Tetrarch on top of the adjacent piece you want to eat, and when you let go of the mouse button that piece will be igui captured.

“I can’t visualise these crazy moves!  Can Ai Ai show me where my pieces can go?”

It sure can!  There are three main options for this:

  1. On the ‘Settings’ menu, turn on ‘Show available moves’.  This will highlight any pieces on your side that have available moves on your turn, and if you click and hold on that piece, Ai Ai will show you what squares it can legally reach.
  2. Also on the ‘Settings’ menu, turn on ‘Highlight moves from position’.  When this option is on and you click and hold on one of your pieces, Ai Ai will dim the rest of the board and show you where your piece can move using arrows and will highlight the possible destination squares.
  3. On the ‘Game’ menu, choose ‘Available moves…’ and a window will open up that will show you a visual depiction of every possible move in the current position; you can choose your move by clicking through the different options and clicking OK.  Alternatively, simply press Ctrl+M (Cmd+M on Mac) to open that display.

There are many useful display options on the Settings menu, so do make use of them to help you get started with these games.

“How can I change the AI settings?”

On the AI menu, you can choose whether to play an AI vs AI game, human vs human, or human vs AI.  You can also use the buttons on the bottom left of the screen to select these options.  Click on ‘Set thinking time (s)…’ on the AI menu to determine how long the AI may think about each move.

“How can I see all the piece types present in each game?”

On the Reports menu, click the ‘List Pieces (Chess family only) option.  After a few seconds, Ai Ai will generate an HTML table showing all pieces present in the current game you have loaded, including their names, piece graphics, abbreviations in the notation, and Ai Ai’s estimate of their material value.

“Why are the piece abbreviations in the move notation different from the previous convention?”

In Ai Ai we need to define promoted and unpromoted forms of pieces separately.  A Lion that appears in the starting position is defined as a separate entity from a Lion that appears when another piece promotes.  In most previous implementations, promoted pieces appear in notation as ‘+[piece promoted from]’, so when a Kirin promotes in Chu Shogi, for example, the resulting Lion is notated as ‘+Kr’.  However, piece promotions often vary across this set of games, so if we wanted to maintain that we’d need to make lots of duplicate pieces to account for these various promotions.

So instead, we opted to abbreviate a piece that appears after promotion as ‘[promoted piece abbreviation]+’, since this allows for consistency across all the games, and no additional duplication of pieces in the code.  So our friend the Lion appears as ‘Ln’ when he starts on the board, and as ‘Ln+’ when he appears by promotion, regardless of what piece he promotes from.  In these historical variants this works fine, because promoted pieces never unpromote, so we don’t particularly have to care what piece they promoted from anyway.

This also means that if new large Shogi variants arise in the future, we can easily add in games with new promotions without needing to add duplicate pieces.  That in turn makes it possible for enthusiasts to make new variants with different promotions without asking Stephen or I for help writing additional code.

However, the situation changes in games with drops, like Tori Shogi, modern Shogi, or Wa Shogi.  In these games, pieces in hand will demote to their previous form, so we *do* need to know what pieces the promoted pieces used to be.  So, in these games pieces maintain the standard notation.

“How do I make those cool animated GIFs of my games?”

At the end of your game, click the Game menu and choose the ‘Review…’ option.  Once the game review screen opens, click the icon that looks like a strip of film to create an animated GIF.

“How can I tinker with the AI and try to improve it?”

On the Settings menu, choose ‘Select Role…’ and give yourself the ‘Superuser’ role.  When you next start Ai Ai, you will have a greatly expanded array of menus and options, many of which will allow you to perform various analyses on games, generate heuristics, test different AI methods and settings, etc.  There’s a lot to play with, Ai Ai has many powerful features for game analysis 🙂

“I have a great idea for a modern variant of these games!  Can you make it for me?”

If you take a look at the .mgl files that define each game, you can see that creating a new variant is pretty straightforward, so long as it can be constructed out of components (pieces and rules) that Stephen has already implemented.  Give that a try first, and if you struggle to get it working, then I can help you to fix it.

“I don’t like the default options you chose for some of these games, and I’m tired of changing them every time.  Can I change them permanently?”

Absolutely.  Just open up the appropriate .mgl file for that game in a text editor, and find the section at the end called “parameters” — this section defines the options that appear when you load the game.  Each option as a default setting, and if you change that to one of the other options listed in the “optionNames” line of that parameter, you will change that default setting permanently.

Just make sure that you save the changed .mgl file with a different name, and keep a backup copy somewhere; otherwise if you copy a new Ai Ai update into the same folder that file will be overwritten by the Ai Ai update.

“Why does the AI struggle with these large Shogi games?”

Ai Ai is a general game-playing framework, and excels at using variants of Monte Carlo Tree Search.  Chess-like games don’t suit these kinds of algorithms very well, so Stephen has added a general Chess-playing engine to make them playable.  However, because of his engine’s amazing flexibility and generality, each specific Chess or Shogi variant has to be optimised separately to get these best out of the engine, and this takes a lot of time and effort.  At this stage we are mainly concerned with getting the games functional, so we haven’t yet engaged in any game-specific optimisation (that will be my job in the months to come).

The main issue though is simply that these games are huge!  Every position has a large number of possible moves to search, and the AI is simply not able to search as deeply as it can in smaller games.  Extending thinking times does help, but even getting 4-5 moves deep requires far, far longer than in smaller games.  Eventually I hope to experiment with using other AI methods for these games, but this will take a lot of time.

So, for now at least, please be patient and give the AI as much time to think as you can stand 🙂

“Hey Eric, why didn’t you include [my preferred rule]?  It’s clearly superior because….”

I completely understand that some modern players would prefer I chose one set of definitive rules, but the problem here is there’s no consensus on what those definitive rules should actually be, and some of these arguments have been going on far longer than I’ve been involved in the Shogi variant world.  That being the case, I’ve decided to opt for a preservation approach, where any rules proposals that are plausible given the historical information we have on these games are kept alive in these implementations and given as an option.

My hope is that over time, the Shogi community will try these various options and decide on a ruleset for each game that we can consider the modern standard.  At that point, I would be happy to revise these implementations to have one ‘standard’ version for each game, and preserve old options in a secondary, customisable implementation, for posterity.

“Hey Eric, why did you include stuff nobody plays anymore, like Hodges’ moves for the Lion Hawk and Free Eagle?”

If a move or rule has been played extensively and is plausible in some way, then I’ve tried to include it, regardless of its popularity.  I initially wasn’t going to include Hodges’ moves, given the modern Tenjiku community seems to have rejected them, but then I read this old correspondence of his on that very issue:

“I well remember my and John Fairbairn’s discussions with our good friend Maruo all those years ago when we were discussing the more exotic variants in his house.  He was absolutely adamant that [the Lion Hawk] does NOT have Lion Power.  There are several points to make.  Firstly, his interpretation is in our mind in no doubt whatsoever.  Secondly, translated text … says words to the effect “…moves like a Lion…”, now this is simply a short way of describing a two-step mover in all directions.  Lion power is something quite different.”

He went on to explain why appeals to logic (“why would the Lion promote to a weaker piece?”) and symmetry don’t work, because plenty of ancient Shogi games include demotion and substantial amounts of asymmetry.

I found this argument reasonably compelling, and given that George was a scholar with connections to Shogi history researchers in Japan and had direct access to the historical sources, we can reasonably assume his opinions on these matters were well informed.  Therefore, I had to consider those moves plausible, and decided to include them as an option.

“Hey Eric, why didn’t you include my Shogi variant?”

We might do that later, but for now the focus is on the historical variants.  Modern variants, particularly modern variants that are already playable elsewhere, are of much lower priority because they don’t need our help to be preserved and made playable.  Also, speaking for myself I prefer to work on variants I have personal experience with, and I am deeply obsessed with large Shogi and much less so with smaller modern variants 🙂

Tagged , , , , , , , , , , , , , , , , ,

Shogi variants: translation notes (I)

One of the many challenges of working on the large Shogi variants is the language barrier.  Not only are the historical documents explaining these games in Japanese, they are in medieval Japanese, and medieval Japanese is written very differently from the modern form of the language.  For a start, many texts were written using only Chinese characters (kanji), and without the helpful hiragana and katakana syllabaries found in the modern language.  On top of that, the usage of kanji has evolved over time, and as with any other language, over the centuries the style of writing has evolved too.

Combine these factors with the generally terse writing style of these old documents, and a natural tendency for Shogi fans to feel quite passionate about their particular interpretations, and we end up with fairly frequent disputes over some rules and piece abilities in these games.  Just to be clear about this, for aspiring Shogi researchers: DO NOT use Google Translate on these old documents!  Google Translate understands modern Japanese (sort of), but has very little idea of what is being said in these old texts.  Even native Japanese speakers have great difficulty interpreting these documents.

Below is an example of the challenges we faced in implementing Dai Dai Shogi, a 17×17 Shogi variant with a large array of 96 pieces of 64 different types for each player.  Somewhat remarkably, most of these pieces have largely agreed-upon powers, but there are two in particular with some disputed abilities: the Furious Fiend, and the Great Elephant.  These are my notes on these pieces that I put together for Stephen, cleaned up and with a bit more explanation.  I share them here as I thought it might be interesting for some of you to see the work that goes on behind the scenes as we try to bring these games to life.

Ultimately we opted to preserve the versions of these pieces that have some historical justification, and allow the player to choose the one they prefer.  As you will see, the pieces change quite a bit depending on which source you use!

——-

Furious Fiend (promoted Lion)

The Furious Fiend is the promoted form of the Lion in Dai Dai Shogi and Maka Dai Dai Shogi.  Many of you will remember the Lion from Chu Shogi, where this unique and powerful piece dominates the board; for those new to these games, here’s a brief summary of it’s prodigious powers:

  • The Lion may move twice in one turn like a King — one step to any adjacent square
  • This double-movement power allows numerous unusual abilities:
    • The Lion may step to an adjacent square, then back to its starting square — effectively passing its turn
    • It may capture a piece on an adjacent square, then move back to its starting square, appearing to capture an adjacent piece without moving — this is known as igui, ‘stationary eating’
    • It may capture two pieces in one turn
  • The Lion may also jump to any square within the 5×5 area around its starting point (but this then is its entire move for the turn)

In the diagrams you’ll see in this article (and in Ai Ai when using the diagrammatic piece sets), orange squares represent places the piece may step to; the stars represent squares a piece may jump to; an exclamation point indicates a square where capturing without moving is possible (igui); and a red arrow indicates the piece may move unlimited squares in that direction.

Having all of these movement possibilities gives the Lion immense power and flexibility, but in the larger games, the Lion is far from the strongest piece!  One of its beefier cousins is the Lion Dog, which we also need to know about in order to understand the Furious Fiend and the Great Elephant:

  • The Lion Dog has a three-step ‘Lion-Power’ move, but restricted to one direction only.  Once the Lion Dog chooses one of the eight possible directions for its move, all of its subsequent moves must take place along that same direction of movement, and the Lion Dog may not move backward past its starting square.
  • With its three-step Lion move, the Lion Dog may:
    • Jump directly to the third square (and capture any enemy there)
    • Jump to the second square, then proceed to the third, or backward to the first, potentially capturing two pieces
    • Step outward along that direction up to three times, capturing up to three pieces
    • At any point it can stop, the Lion Dog is not required to use all three Lion-power moves.

So the Lion Dog extends the Lion’s power by allowing potential triple-captures, but its flexibility is reduced somewhat as those captures must all take place along the same line.

In Dai Dai Shogi and Maka Dai Dai Shogi, there are compound pieces that combine Lion and Lion-Dog moves with the moves of other pieces.  One of those is the Furious Fiend, the promoted form of the Lion itself.

There are two main versions of this piece:

  1. Lion + an option to slide three steps in any of the eight directions
  2. Lion + Lion Dog

The evolving consensus at this point is that (2) is the correct move.  (1) is likely a relic of a misunderstanding that pervaded the English literature on these games for a long time, where the Lion Dog was thought to have a three-step move in any direction without Lion power, but that’s considered a misapprehension now.  For those of you who may have physical Dai Dai Shogi sets from The Shogi Association, the manuals included in those games include this version of the Lion Dog; for that reason, this version is sometimes referred to as being part of the ‘TSA rules’.

So, when that non-Lion-powered move was prevalent, and the old texts said Furious Fiend moved as ‘Lion + Lion Dog’, when you combine these moves you end up with (1).  But with our corrected understanding of the Lion Dog, we get (2).  In our Ai Ai implementation, we have allowed players to choose which version of the Furious Fiend to use in Dai Dai Shogi and Maka Dai Dai Shogi, since both have been used in modern play.

——-

Great Elephant

Some Shogi scholars suspect that Maka Dai Dai Shogi had some fans back in the day — it’s the most consistently-described of the very large games in the historic sources, with few discrepancies despite the texts being written years apart by different people.  Players also developed a chant in Chinese that they’d use to remember the initial setup, so that’s an indication it was being played some amount.  This can be helpful when questions arise about other games, because if that same piece appears in Maka Dai Dai we may be able to get a more consistent picture of how that piece should work.

The Great Elephant, however, does not appear in Maka Dai Dai Shogi and instead only shows up in Dai Dai Shogi (and the 25×25 Tai Shogi), where it is the promoted form of the Lion Dog; in Maka Dai Dai Shogi, the Lion Dog is one of the strong pieces that demotes to Gold General.  The Great Elephant is described in very different ways in the available historical sources that explain the rules of Dai Dai Shogi.  These are the main options, according to the sources I’ve seen before:

  1. The piece doesn’t exist at all (!) [two sources don’t have the Lion Dog promote to anything, so this piece doesn’t exist in Dai Dai Shogi in those documents]
  2. The piece moves as claimed in a note on Japanese Wikipedia: Steps 1-2 squares diagonally forward, or unlimited slide in all other directions *and* it may jump over up to 3 pieces, friend or foe, and continue its slide on the other side
  3. The piece moves as described in (2), except that it *cannot* jump over pieces
  4. The piece moves as in Taikyoku Shogi: steps 1-3 squares diagonally forward, or unlimited slide in all other directions and it may jump over up to 3 pieces and continue sliding on the other side of them.
  5. The piece may step 1-5 squares sideways or diagonally backward, or 1-3 squares in the other directions [this move is widely discredited now, given that it’s based on the old TSA interpretation of the Lion Dog].

However, in an effort to further develop our understanding of this disputed piece, I obtained a copy of the ancient document Shogi Rokushu no Zushiki from the Japanese National Diet Library, and have gone through every scrap of info on Dai Dai Shogi in there.  This is document includes the Great Elephant as the promotion for the Lion Dog.  After deciphering the notation based on the text descriptions of various pieces’ movements, and examining the diagram of the Great Elephant’s move:

GE_SSZ_diagram

I believe this diagram from the text indicates the following move:

  • Lion-Dog-style 3-step Lion Power in the four orthogonal directions and diagonally backward
  • Or it may slide unlimited squares in the four orthogonal directions or diagonally backward
  • Or it may step 1-2 squares diagonally forward.

The key is the three slashes over the longer lines.  The text very clearly states that the long lines indicate unlimited sliding moves in those directions.  There is only one other piece that is given a text description that has the three slashes overlapping the longer lines, and that is the Teaching King in Maka Dai Dai Shogi’s section in the text, and that text says the Teaching King functions as Lion Dog + Free King.  So, we can infer from this that those three slashes indicate Lion Dog moves in those directions:

Here we have the source of some confusion.  The Lion Dog’s move is described in this text using the word ‘odoru’, which means ‘dance’, and depending who you ask, ‘odoru’ in this context either means a Lion-Power move, or a leaping move where you can jump over some number of other pieces.

So that leaves us with two plausible options, depending on how we interpret ‘odoru’:

  1. if ‘odoru’ means ‘leap over stuff’: that’s the Japanese Wikipedia move, (2) above
  2. if ‘odoru’ means ‘Lion Power move’: that’s my interpretation above

However, I consider (2) to be the more plausible interpretation, and easier to place among the rest of the Dai Dai Shogi army.  That move would make the Great Elephant a promotion above the Lion Dog, given the added mobility, but reducing the Lion-Power directions to six makes it less powerful than the Teaching King/Buddhist Spirit.  This makes some intuitive sense, as the Teaching King and Buddhist Spirit are the strongest Lion-Power pieces in Maka Dai Dai Shogi, and typically the larger games introduce additional, more powerful pieces over their smaller cousins.  Meanwhile, (1) would require a new mechanic that only appears for this single piece, and does not otherwise appear in Dai Dai Shogi at all, so it does not seem likely.  However, Dai Dai Shogi does introduce two hook-moving pieces that can perform double Bishop or Rook moves, so it’s perhaps not impossible that the creators would introduce another form of movement as well.

I should mention that Shogi Rokushu no Zushiki is known for having a lot of differences in how pieces move in all the games it covers, and that includes Dai Dai Shogi.  However, the power-pieces in each game do seem to agree with other sources from that period.  The beginning of the text opens with the author complaining about transcription errors and incorrect piece names in other documents about Shogi, so he is presenting himself as correcting the historical record, but today we consider many of his claims to be questionable.  In any case, I feel it’s worthwhile to try to understand the Great Elephant move presented in this text and see whether it fits in the game as we know it today, and give players a chance to try it for themselves*.

The English/Japanese Wikipedia seems to have simply removed the Lion Dog moves from the Shogi Rokushu no Zushiki move and given the Great Elephant the six ranging moves and diagonally-forward stepping moves only.  This is nice and simple but I don’t think this is a correct interpretation of this particular text.  However, this move has been widely used until now, and certainly is a playable variation with some historical justification.

GE_WP_diagram

The Final Move Choices

We are left with three main usable interpretations of the Great Elephant in Dai Dai Shogi/Tai Shogi, after we discount the one related to the old TSA Lion Dog move, plus one variant for use in Taikyoku Shogi only:

    1. Steps 1-2 squares diagonally forward, or slides unlimited squares in the four orthogonal directions + diagonally backward, or 3-step Lion Power move in those same six directions [my translation of SRZ]
    2. Steps 1-2 squares diagonally forward, or slides unlimited squares in the four orthogonal directions + diagonally backward [English/JP Wikipedia]
    3. Steps 1-2 squares diagonally forward, or unlimited slide in all other directions *and* it may jump over up to 3 pieces, friend or foe, and continue its slide on the other side [JP Wikipedia move in the notes]
    4. In Taikyoku Shogi it moves as (2), except it may move 1-3 squares diagonally forward instead of 1-2.  Hopping-ranging moves are commonplace in Taikyoku Shogi, so that move is uncontroversial in that game.

Given the lack of consensus, we opted to implement all three possible moves for Dai Dai Shogi and Tai Shogi (1, 2 and 3 above), and allow the player to choose.  So at the start of the game, you may choose your preferred move of the Furious Fiend and the Great Elephant before starting play.  The Taikyoku Shogi move will be relevant once we get to that game!

We also include an alternative implementation of Dai Dai Shogi which does not include the Great Elephant, since two historical sources indicate that the Lion Dog does not promote at all.

As for the diagrams, for the range-jumping move in (2) and (4), we use a numbered circle to indicate how many pieces may be jumped over in one move.  In Taikyoku Shogi this number varies between pieces, so we will continue to use this notation when we start working on that game and more range-jumping pieces appear.


Conclusions

So, after all that work, we ended up with a robust implementation of Dai Dai Shogi which we think covers all the main bases for the Great Elephant: the long-distance sliding piece espoused by English Wikipedia; the long-distance jumper suggested in the notes on Japanese Wikipedia; and the powerful Lion Dog compound suggested by the text of Shogi Rokushu no Zushiki.  Players may also choose between either version of the Furious Fiend.  Finally, we also allow players to ditch the Great Elephant entirely.

Having experimented with these variations myself, here are my impressions (note that I’m excluding the Taikyoku Shogi version, which will feel very different in that game, given the enormous 36×36 board and the presence of other range-jumping pieces):

diag_great_elephant_wp_pr

Great Elephant move from English/JP Wikipedia

This piece is highly mobile, which makes it very useful in the endgame, where mobility becomes very powerful as the board starts to empty of other pieces.  However, the lack of Lion Power does mean that players have to consider their strategic aims when deciding whether to promote this piece.  In a congested middlegame position, for example, one may want to avoid promoting the Lion Dog early so that the Lion Dog still presents a powerful capturing threat.  In a late-game situation where enemy defences have thinned out, then making a capture and promoting the Lion Dog may be well worth it, as the new-found mobility will come in handy.

diag_great_elephant_jpw_pr

Great Elephant move from the notes on Japanese Wikipedia

This piece is amazingly fun to use!  Given the congested, high-density setup of Dai Dai Shogi, having the ability to leap any distance over three pieces is hugely helpful.  Promoting this piece loses the Lion Dog’s multi-capture abilities, certainly, but in exchange you gain incredible flexibility, plus the option of threatening the enemy King even over a dense wall of protective pieces.  I’d imagine that few players would hesitate to promote their Lion Dogs if this promotion is available; the piece just gives you so many new options when on the attack, and as the only long-range jumper on the board, it’s very hard for your opponent to chase it down.

I have had so much fun with this beast that I’m getting very excited to work on Taikyoku Shogi further down the line, where lots of these range-jumping pieces will appear.

diag_great_elephant_srz_pr

Great Elephant move from Shogi Rokushu no Zushiki (my translation)

This piece ends up feeling surprisingly well-balanced, in my opinion, with the rest of the power-pieces available in Dai Dai Shogi.  When promoting to this from Lion Dog, one loses the option to perform multiple captures in all eight directions, but in return gains significant mobility.  However, that mobility is less useful on the attack, as the forward diagonal moves are only two-square steps.

The result is a piece that feels like an upgrade from the Lion Dog, but nevertheless requires finesse to use effectively.  The Great Elephant also proves to be a powerful and mobile defender, able to move backward into a defending posture quickly and mop up several opposing attackers at once.  To me it’s an interesting piece to use, and it feels like a viable addition to the game.

In my testing, the presence of the more powerful Elephants, either the SRZ version above or the range-jumping variant from Japanese Wikipedia, has not significantly influenced the length of the typical game.  Dai Dai Shogi is a big game with lots of pieces on the board, and even the most powerful pieces have to bide their time until their powers can flourish; otherwise they have to run away from constant threats of capture from the hordes of weaker pieces**.  So the stronger Elephants have a big influence in the endgame, but overall don’t feel overly unbalancing.  In my opinion, though, the SRZ version is the one I would advocate if you want a stronger piece than the English Wikipedia version; the range-jumping Elephant feels a bit out of place with its highly unusual movement abilities, whereas the SRZ variant feels more at home amongst the other Lion-Power pieces on the board.

As for our furious friend the Furious Fiend, both versions are a straight upgrade from the Lion, and I don’t think either one makes a significant difference in the overall shape of the game.  The Lion + Lion Dog version can make additional captures in some circumstances, but against a good opponent this is unlikely to come into play very often — a good player will never allow three of their strongest pieces to be in range of a Furious Fiend at the same time!

In this case the choice comes down more to personal preference.  I like the more powerful Lion + Lion Dog variant, as it feels like a more significant upgrade over the Lion, which adds incentive to move the Lion into battle in the first place.  I also feel it’s more likely to be the intended move, given that the old documents position it as Lion + Lion Dog and we now understand the Lion Dog move to have multi-capture abilities.


Next Moves

So, at this point we are close to finishing our time with Dai Dai and Maka Dai Dai Shogi, so you can look forward to a more detailed look at those games on this blog in the near future.  In the meantime we are getting started on Tenjiku Shogi, which is an exotic and exciting game with a lot of disputed rules and pieces to sort through, so I may do another ‘translator’s notes’ article on that game once we complete it.

I hope some of you out there may give these games a try in Ai Ai when you have a chance; they really are unique games in the history of the Chess family, and deserve your attention!  Though these games are many centuries old, they have unique ideas and pieces that still stand out today, even amongst the thousands of Chess variants that have been constructed since.  Give them a try, and you may find you enjoy the sheer immensity and creativity of these fascinating games.

UPDATE: I’ve been really pleased to discover that the Dai Dai Shogi page on Wikipedia has been updated with a diagram of the move I suggested from my interpretation of Shogi Rokushu no Zushiki:

wp_great_elephant_update

I’d like to give a hearty thank you to HG Muller and Wikipedia user Double Sharp for taking note of my arguments here and presenting this move as one possible interpretation on the Wiki page, I very much appreciate that my work on this was recognised.  I also agree with both of them that this piece is seriously confusing, and we’ll probably never be able to definitively decide whether it should even be in the game at all!



* I suspect some readers may be wondering: why did I pick out this one piece and elevate it to the status of a plausible move, when the rest of the source text is viewed with skepticism (and justifiably so)?  My answer is that the Great Elephant in particular is a mysterious piece, sometimes existing and sometimes not, and when it does exist, every source seems to have different opinions about its moves.  While Shogi Rokushu no Zushiki has a lot of move descriptions that don’t match other sources, it *is* a source that actually contains the Great Elephant, and so I decided to investigate it.  I found that the move looked more complex than I’d expected, and yet the positioning of the stepping and sliding moves exactly matched the diagram provided on Wikipedia.  That led me to think that it would be worth looking more deeply at this move, and what I found seemed potentially interesting to try within the current agreed-upon Dai Dai Shogi ruleset, even when taken out of the context of the SRZ account of Dai Dai Shogi.

I left the other moves from SRZ aside, as the other pieces in Dai Dai Shogi have a consensus regarding their moves and abilities, so in that context I don’t see a need to replace them with moves from a single divergent source.  That’s not to say some of them aren’t interesting; SRZ gives the Golden Bird almost the reverse of the Great Elephant move, with Lion Dog moves on the front diagonals and limited 3-step retreating moves (without Lion Power).  But that piece exists in Maka Dai Dai Shogi and the moves appear to be consistent across the other sources, so I can’t see a justification for replacing it with the SRZ move.

In short, I wanted to know more about the Great Elephant, and found the move in SRZ to be both plausible and interesting, so I wanted to give players a chance to try it!

**This phenomenon of stronger pieces sometimes being a liability is seen in all Chess-like games and was dubbed the levelling effect by great Chess variant explorer Ralph Betza.  Put simply, because different pieces have very different levels of power and influence on the board, a threat against our weaker piece from the opponent’s stronger piece is often no threat at all, because if those pieces are traded off we still end up ahead in material strength on the board.  However, a threat against a stronger piece from a weaker piece must be defused immediately, because if we allow that trade of material we hand the opponent a strength advantage.  This is why the advice in the Chess opening is never to bring the Queen out too early:  your opponent can attack it and chase it away with their weaker pieces, as you’re forced to retreat it to avoid a bad exchange, leaving you under pressure and lacking initiative.

The levelling effect is also evident in large Shogi variants, where the board tends to be thick with dozens of pieces, many of them weak.  The powerful pieces therefore have to be deployed cautiously, given the sheer number of possible threats from the opponent’s huge army.

Some large Shogi pieces are not quite so vulnerable to this however, simply because they are ridiculously powerful.  Tenjiku Shogi’s Fire Demon is a good example — the Fire Demon instantly destroys all adjacent enemy pieces, even on the opponent’s turn, so threats against it have to be made at a distance.  Similarly, the various Lion-Power pieces cannot be chased away by single-step-moving pieces, because those pieces have to be adjacent to threaten to capture, and the Lion-Power piece can simply take them at will with an igui capture.

Note that the SRZ Great Elephant does have vulnerable areas along the front diagonals, where it cannot make a Lion Dog move!  This reduces its attacking presence, as there are several weak pieces with short movement ranges (2-3 squares) that could make credible threats along those diagonals, forcing the Great Elephant to retreat or be captured.

Tagged , , , , , , ,

Shogi and some variants now available in Ai Ai — play against AI or online!

UPDATE (12 April): Stephen has updated Ai Ai to fix a bug with Chu Shogi’s Lion, and we also added several new variants including Goro Goro Shogi, Goro Goro Plus and Wa Shogi!  Grab the updated version here.

——-

For the last few weeks, Stephen Tavener and I have been collaborating to bring Shogi and several of its variants to his amazing Ai Ai software, a general game-playing program that lets you play hundreds of abstract strategy games against strong AI, or against human opposition online.

Last year Stephen added a general Chess-playing engine to Ai Ai, and has been steadily adding lots of great Chess variants to it such as Capablanca Chess, Chess960 and Grand Chess.  I had been embarking on a project to playtest some of the more promising large Chess variants — on 10×8 boards and larger — and so Stephen added a huge array of new piece types into Ai Ai so that these variants could be added in (more on these games in a future post).  Then we got started on Shogi, which turned into quite a project — Stephen had to cope with some incredible coding challenges throughout, from implementing drops to Chu Shogi’s complex Lion-trading rules.  On top of that there was lots of design work to do, because we wanted to offer multiple piece graphic sets to help people who have trouble learning Japanese kanji characters.

In the end we went with three sets of pieces: traditional pieces using a single kanji character for each piece; mnemonic pieces that combine the kanji with mnemonic diagrams designed by HG Muller; and diagrammatic pieces, where each piece is a simple square shape with a diagram of the piece’s move on it.  With these three options in place, these games should be accessible to a larger audience.

Modern Shogi

First and foremost in Ai Ai’s new Shogi assortment we have modern Shogi, played on a 9×9 board with ten different types of pieces.  For those who aren’t familiar with Shogi, it has several major differences from Western Chess:

  1. Pieces that are captured are truly captured — they become the property of the capturing side, and are placed on the side of the board under their control.  At any time, a player may ‘drop’ a captured piece to any empty square on the board in lieu of a normal move (with some restrictions).  Drops make Shogi a very dynamic and aggressive game, and drastically change the feel in comparison to Chess.  Because pieces keep coming back to life, the board stays mostly full of pieces throughout the game; endgames are often an exciting race to checkmate as both sides’ defences break down and they start launching brutal attacks back and forth.
  2. Nearly all pieces can promote — unlike in Chess, where only Pawns may promote upon reaching the opponent’s back rank, in Shogi nearly all pieces can promote when they reach the opponent’s starting area.  Pieces promote by flipping over and revealing a new piece on the opposite side, which is more powerful than the original.  Promoted pieces are demoted again once captured, though.
  3. Shogi is a bigger game than Chess, taking place on a 9×9 board with 81 squares as opposed to Chess’ 8×8 board with 64 squares.  Shogi also has ten types of pieces, significantly more than the six types present in Chess.

There’s a lot more to it than that, of course, but these three changes alone make Shogi stand out among the modern variants of Chess played around the world.  It’s a vibrant and exciting game and well worth trying if you’ve ever enjoyed a game of Chess.

Shogi games tend to take a bit longer than Chess; a typical Chess game lasts about 80 moves in total, whereas a Shogi game generally lasts around 120 moves.  Some small variants of Shogi have been designed to generate some quick-playing games that still capture the feel of the full game, and that serve as useful introductions for new players or children.  In Ai Ai we’ve added a couple of these:

In Minishogi, Shogi gets shrunken down to a tiny 5×5 board with just six pieces per player.  Surprisingly the game is still remarkably deep at this size, thanks to the complexities of drops.  Judkins Shogi is similar, but uses a 6×6 board with seven pieces per player, adding the Knight back into the mix.

Sho Shogi, Chu Shogi and Dai Shogi

As readers of this blog will know, I have a certain fascination with the ancient variations of Shogi, particularly the ambitious and gigantic ones.  I’m delighted to say that Stephen has implemented some of these variants in Ai Ai as well!

Back in the early days of Shogi, from the 13th-16th centuries or so, the game came in three sizes —  Sho Shogi on a 9×9 board (Small Shogi, which became modern Shogi), Chu Shogi (Middle Shogi) on a 12×12 board, and Dai Shogi (Large Shogi) on a 15×15 board.   At this time drops were not in the game, so pieces that are captured are removed from the game permanently.  Sho Shogi was considered a quick game, often played with children, while Chu Shogi was the most popular form and the enormous Dai Shogi was for a time the most prestigious variation.

Eventually drops entered the game sometime in the late 16th century, and this innovation suddenly catapulted Sho Shogi to the forefront of the Shogi world.  Today Chu Shogi still survives, and is considered by some to be the best large Chess game ever invented, whereas Dai Shogi is still played as well but much less frequently than Chu.

All three of Shogi’s closest ancestors are now playable in Ai Ai:

The eagle-eyed among you will notice that Sho Shogi sports an additional piece sat in front of the King.  This is the Drunk Elephant, a powerful defender that promotes to a Crown Prince, which functions as a second King!  If you manage to make a Crown Prince, your opponent must capture both your Prince and your King to win the game.  Both Chu and Dai Shogi have Drunk Elephants as well, and numerous other new pieces; Chu Shogi has 28 piece types, and Dai Shogi has 36!  As you might expect, the larger games are challenging for the AI to play well, so be sure to set the AI’s thinking times quite high if you’d like a challenge.

I highly recommend trying these historical variants, particularly Chu Shogi and Dai Shogi, which I’ve written about extensively before.  Chu and Dai are wonderful games, richly strategic and packed full of variety, and I hope some of you out there may try your hand at them now that they’re available in Ai Ai.

Tori Shogi

For those who prefer quicker and tighter gameplay, we also added a more recent historical variant of Shogi — Tori Shogi, or Bird Shogi.  Tori Shogi gets its name from the fact that all the pieces have names related to birds — even the Pawns are changed to Swallows.  Tori Shogi was invented by Toyota Genryu in 1799, and has the unique distinction of being one of only two historical variants for which we have recorded games played by professional players (the other being Chu Shogi).  Tori Shogi has gained a certain amount of popularity in the West, and there’s even a fine English-language book available on the game for those who want to learn to play well.

Tori Shogi is played on a 7×7 board with eight different types of pieces, two of which only appear by promotion.  The small 49-square board starts packed with 16 pieces for each player, making the early game quite claustrophobic!  The game uses drops as in modern Shogi, but with one major difference: in modern Shogi, you may never drop a Pawn to a file that already contains one of your Pawns, but in Tori Shogi you may have two Pawns on the same file at a time.  This small change hugely alters the game’s tactics and gives it a very different feel from standard Shogi.

Below you can see the initial position of the game, and a sample game in animated GIF form.

Next Moves

We are forging ahead with some additional variants for the next release.  First up we have Goro Goro Shogi, a modern small variant developed in 2012 as a way to help young kids in Japan to learn modern Shogi.  This game is played on a 5×6 board with a limited selection of pieces, but unlike Minishogi and Judkins Shogi, there are three Pawns per side instead of just one.  In my opinion this makes Goro Goro a much better introduction to Shogi, as the use of Pawns is essential in the full game (much as in Chess).

We have also added Goro Goro Plus, a fantastic little variant that takes Goro Goro and gives each player a Lance and Knight in hand at the start of the game, available for drops.  This addition really spices up the game and makes Goro Goro more than just a Shogi learning tool, and turns it into an exciting game in its own right.

On the historical variants side of things, we have Wa Shogi, an 11×11 game that shares with Tori Shogi a certain flair for exotic, animal-based piece names.  Unusually for a Shogi variant, Wa is playable both with and without drops, and is a great game either way!  I slightly prefer playing with drops, which gives the game an exciting pace and added tactical sharpness.  Without drops Wa Shogi becomes a delicate strategic affair, where players often try to establish coordinated invading legions that can escort the weaker pieces to the promotion zone (the weakest pieces in Wa have strong promotions).  The two faces of Wa play really differently, so it’s like having two games in one.

I highly recommend Wa Shogi for fans of modern Shogi; particularly when played with drops, it feels like a clever expansion of Shogi with a distinct feel due to its asymmetric starting position and unusual pieces.  I firmly believe that if a concerned effort were made to promote this game it could achieve a decent level of popularity!

So that’s a quick roundup of all the Shogi goodness now available in Ai Ai, and a little preview of what’s to come — please go give the games a try, and of course give me a shout if any of you out there fancy a game!

Sometime down the line I’ll be back with another roundup, in which we’ll be taking a look at the large Chess variants available in Ai Ai as well.  I’m also nearly done with an in-depth analysis of a Chu Shogi game, and an introduction to Tenjiku Shogi, so look out for those posts coming soon (-ish).

Tagged , , , , , , , , ,

Permute: Move by Move

Some time ago I introduced you all to my game, Permute — an abstract strategy game inspired by the Rubik’s Cube.  Since then, the game was implemented on Christian Freeling’s MindSports — an essential site for any fan of the genre — and in Stephen Tavener’s wonderful Ai Ai software.  Hopefully some of you may have tried the game since then, and have discovered that the game has a unique feel in play and supports some unusual strategies and tactics.

Since developing the idea I’ve played a great deal of Permute, mostly against Ai Ai’s stiff opposition, but also some games against human opponents.   In this post I’ll walk you through a game I played against Simon R on MindSports, and explain some basic tactical and strategic considerations to keep in mind while playing Permute.  This sample game contains a few instructive moments that should be enlightening for any new player.

Here’s an animation of the full game, courtesy of Ai Ai:

permute12-game2-simonR-final

Full game — Simon was Orange, I was Yellow. Final score — Orange 16, Yellow 43.

The game is also viewable on MindSports, if you want to step through it move by move yourself.

Permute Move Notation

In order to record games of Permute and replicate them later, we need some form of notation to keep track of what moves have been played.  Ed van Zon of MindSports came up with a simple notation system that is very easy to learn:

Permute 12x12 -- notation sample-01

This small sample board illustrates how the notation works.  In this case, Yellow has made a move, rotating a 2×2 face clockwise.  We can translate this move into notation easily:

  1. We start by noting the direction of the twist, with a capital C for a clockwise twist, or A for anti-clockwise.
  2. Then we note the square on the board that is in the bottom-left corner of the 2×2 face we twisted, in this case d3 (indicated by the green square in the diagram).  So our move so far is Cd3
  3. Next we add in the square marking the top-right of the twisted face, in this case e4 (the red square in the diagram).  Our move is now Cd3e4
  4. Finally, we add a hyphen for readability, then add the square we chose to bandage, which is d4 in this case (the blue square in the diagram).  So our complete move can be written as Cd3e4-d4.

That’s it!  If we note down every move in this way, in proper sequence, we can easily reconstruct a game later for analysis.  In Ai Ai the notation works basically the same, except moves are written with the word Clockwise for clockwise twists and Widdershins for anti-clockwise twists.

Preliminaries

This post is very long, so I will not recap the rules here; if you are totally new to the game, please read my previous introductory post and follow-up about the Ai Ai implementation so that you can learn how the game plays, and get yourself set up to play it.

This particular game was played on a 12×12 board, which is the standard size available at MindSports.  A board that size contains 144 pieces, which seems like a large amount, but each move in Permute directly moves a set of four pieces and also locks down a portion of the board afterward, so games on 12×12 are surprisingly fast and brutal.  Typically a full game will last between 20-25 moves per player (note I’m counting a twist + bandage as a single ‘move’).

The board size has a strong impact on play in Permute, as it does in other territory games like Go.  Playing Permute on 12×12 is similar to playing Go on a 9×9 board — fighting starts immediately, and tactics are a major focus of the game.  Controlling the centre of the board early is very important, as this offers great opportunities both for expanding your own groups and blocking your opponent’s growth.

For a deeper, more strategic battle, I recommend playing on a 16×16 board.  At this size, games will last between 40-50 moves per player, about the same length as a typical game of Chess.  Opening play on this board size will tend to focus more around the corners and edges of the board, with the centre becoming more important once territories have been settled around the sides.  If you want the territorial aspect of the game to really dominate, then you can play on a 20×20 or even 24×24 board; on 24×24 a game will last about as long as a game of 19×19 Go (generally somewhat shorter).  Ai Ai supports Permute play on all these sizes.

The Ai Ai implementation also allows numerous variants of the game to be played.  The most notable of these is the option for a 1[1]2[2]* move protocol.  This means the first player on their first turn twists and bandages once, and from then on players alternate twisting two faces, then bandaging a stone of their colour in the second face they twisted.  I highly recommend trying this variant, which allows some very clever manoeuvring, and also lengthens the game on smaller boards.  For serious strategic play however I still advocate 16×16/20×20 single-twist rules as the best option; the two-twist variant makes useful tactics substantially harder to see, because two twists can cause so many changes to the board state in a single turn.

In any case, 12×12 with single twists is the ideal introduction to the game, and that will be our focus in this article.  In a future article, I will cover strategic tips and techniques for play on the larger boards, aimed at advanced players.

Permute 12×12 — Simon R vs Eric S

The opening begins!  Simon was playing Orange, meaning he gets to move first.  He strikes out for the centre immediately, which is a sensible play on the 12×12 board.  After just that single move, both of us now have two 3-groups each — each move in Permute substantially changes the local group composition, particularly in the early game, where few pieces have been bandaged.

I respond with an anti-clockwise twist at e4f5, aiming to contain any ideas of Orange extending to the western or southern edges.  In the process I connect my two 3-groups into a single 6-group.

Orange responds by growing their own 3-group by twisting f8g9.  Expansion is never a bad thing, but this move is a bit passive; the resulting group is still smaller than my largest, and this move does not really hamper my expansion.  In response, I twist h4i5, connecting two more pieces to the 6-group by rotating a piece into the connecting corner on h5.  From there, I can extend by adding to the piece attached to the north of h5 — I like to call these one-square extensions from bandaged friendly pieces anchors, as they are good bases for expansion — or I can connect two more pieces on the eastern edge of the group in two different ways, thanks to the 3-group covering i3-i4-j4.

Orange spots the threat of more pieces connecting to my group, so he instead connects his own pieces beneath and around my group by twisting h3i4.  This is a good move, in that it serves two purposes — it instantly gives him a nice-sized group, and it actively blocks my group from expanding to the south.  I was prepared for this, though, thanks to the anchor on h6, so I connect three more pieces with a twist at g7h8.  Note that my group is quite secure; each extension has bandaged pieces one space apart, which cannot be separated by any opposing twists.  When building your groups, one-space extensions are the safest way to connect new pieces.

We’re very early in the game, but the centre is already taking shape.  Here Orange decides to secure some more pieces for their southern group by twisting e3f4.  Again this move is a bit passive; Orange has the opportunity to sandwich my centre group between their upper and lower groups, and my hopes of extension rest on me being able to protect the corner of my group at e5, or the northern extension at g8-h8.  Better would have been Ae5d6-e5 — that would have severed my corner at e5, blocking expansion to the west, while further securing Orange’s biggest group.

Having been spared a squeeze on e5 from Orange, I twist c4d5 and bandage d5 to secure that vulnerable corner and gain space on the west side, with ambitions of connecting to the western edge and cutting off the bottom of the board, so that Orange cannot connect their upper and lower groups on that side.

Orange takes action to prevent my group heading west, but in actuality this move only delays me very briefly.  My group is also still expandable to the east and north, thanks to the anchors on h8 and i7.  This move does give Orange a large group in a commanding position over the southern reaches of the board, but at this stage of the game, exerting influence and securing areas for future growth is more important.  A nice move here would have been Cg8h9-h8, which would have blocked off any northern expansion for my group.

I respond by moving forward with my plan to connect to the western edge, which restricts Orange’s growth opportunities and gives me an anchor on b4.  At this point I also notice Orange’s vulnerable corner on i8, which is an important weakness now that Orange has invested several moves in growing this southern group.

We will take a close look at Orange’s next move, which is a crucial moment in the game.  Orange elected to twist i5j6, extending their group closer to the eastern edge while interfering with my group’s potential expansion on that side.  Unfortunately this move does not protect the vulnerable corner at i4, which gives me a huge attacking opportunity.

The middle diagram shows one potential alternative — securing the corner at i5 increases Orange’s hold on the bottom of the board, but it also further cramps their space for extension, confining their largest group mostly to the bottom.  A better alternative would have been the block at h8 I recommended previously — this move is still effective now, forcing a response from Yellow and adding some tactical complications.  However, at this point Ai Ai estimates Orange’s winning chances at just 36% even if they had chosen to block at h8.

Seeing that Orange has not protected the corner at i5, I launch my attack and twist out Orange’s piece on i5 (marked with an X in the diagram).  In one stroke this slices Orange’s largest group in half.  Bandaging at j5 ensures Orange cannot easily repair this damage, and claims further space toward the lower-right corner, making it very hard for Orange to even attempt to connect around this new blockage at i5.

At this point, Ai Ai estimates my winning chances at 88%.  Orange had invested a lot of moves in this group, and now the bottom portion has been entirely cut off from the section on the right, and is unlikely to form the basis of a winning group.  I have control over the centre and plenty of expansion options on the northern and eastern edges of my largest group.

Orange can see they are in trouble, so they respond with a twist at i8j9, in an attempt to constrict my expansion options to the north.  Orange now has to defend perfectly to have a hope of victory; when forced to play moves like this which do not extend one’s own groups, the opponent is often free to build up their advantage further.

This exchange gives us perhaps the most important message to take away from this game: protect your corners!  In Permute you will often need to bend your group around to connect pieces in various directions, and any corner piece in your group left unbandaged is a potential weakness.  One twist can potentially slice your group to ribbons and leave you struggling.

Now that Orange is on the defensive, I decide to aim for connection to the northern edge.  When ahead and in command of a large group, a useful strategic sub-goal is to try to split the board.  The group doing the splitting is then unassailable, assuming you have bandaged it well, and your opponent is heavily constricted in any attempts to connect large chunks of pieces.

Orange responds by constraining my group’s northern expansion, squeezing it from the west.  This does secure the northwestern section of the board for Orange, which could allow them to build a decent-sized group, and their current group stretching down to f6 is currently pretty secure.  However, this move still gives me the chance to connect to the north or east sides.  Ai Ai recommends moves that either disrupt the unbandaged, unsecured Yellow pieces at i6-i8, or blocking Yellow’s anchor at h10 and preventing the northern connection attempt.  In either case, though, Ai Ai pegs Orange’s winning chances now at only about 3.5%.

As it happens, I elected not to complete the northern connection, and instead disrupt Orange’s newest largest group.  My Yellow piece at f9 makes connection for Orange’s group quite awkward, so my hope is that further incursion into Orange’s northwestern territory will further decrease their chances to build a competitive group in that area.

Orange hits back by disrupting my 3-group around h11, ensuring that my bandaged piece at h9 remains isolated for the time being.  However, this leaves me with an opening to cut off three Orange pieces from potential connection via a later twist at f11g12.

At this point I can see that my central group is very strong, so I decide to continue my plan to disrupt Orange’s opportunities for expansion while securing my position.  By extending down from my anchor at b4, I further constrict Orange in the south, while adding a couple more pieces to my central group.

Orange elects to start enlarging their group in the northwest, establishing a bandaged piece at e7 that also prevents further extension of the Yellow group from e6.

I estimate at this stage that Orange’s northwestern group could not grow large enough to challenge my lead, but just in case I decide to extend my group and bandage at j8, in the process securing the vulnerable piece at i8.  Orange fights back by twisting my pieces at j9-j10 into the corner and firmly blocking them off from my group.  This also builds a nice group connecting through to the northwestern section, but those Orange pieces at f11-g11 remain vulnerable.

My first impulse here was to cut off Orange’s group on the top with Cf11g12-g12, but then I spot a move that looks bigger.  The clockwise twist at c2d3 grows my own largest group by three while simultaneously blocking out three Orange pieces permanently, which seems like a favourable exchange.

Orange responds sensibly, forgetting about the unfortunate southern group for now and further building on their central group.  Orange knows that at some point I will be forced to respond if I want to restrict Orange’s growth there.

Now that Orange’s southern group is well contained, I believe that their only chance to win is to connect their central and northern groups, then extend far down the eastern edge to pick up sufficient additional pieces.  I decide to try to connect to the east to completely cut off any ideas of expansion down that edge for Orange.

Orange sees an opportunity to place a secure bandaged corner at k7, extending their northern group a bit further.

Orange was able to extend that northern group a bit, but I could immediately follow up with Ck5l6-k6, completing my goal of connecting to the eastern edge.  Now Orange has no way to connect to the bottom groups, and the bottom groups themselves are cut off from each other thanks to the bandaged piece at j3.

Orange follows up with a prudent move, securing their vulnerable corner at i10 with a bandaged piece at j10.  Unfortunately, at this point the game is fully out of each; Ai Ai sees no future for Orange, giving them a winning percentage of a flat 0%.  There is no way for Orange to make a larger group at this stage.

My goal at this point is to firmly block off any remaining threats Orange has to increase their score substantially.  Having cut off the bottom with my last two moves, now I close the door along the top as well with Cf11g12-g11, splitting the top group and ensuring it cannot connect to Orange’s central group.

Orange has seen the writing on the wall by this time, but is valiantly fighting to the end.  He extends the southern group with two more pieces, and generates a possible threat of winding around the blockage at j3, though this would require several additional moves to achieve.

I can see that Orange may be hoping to connect around j3, so I close that door as well by placing another bandaged piece at k2.  Now there is no way through on the bottom.

Orange now pursues the last viable option for building a good-sized group, which is to extend their central group as much as possible.  I have almost no presence in the northwestern part of the board, so Orange has some chances to build up a score here potentially.

Finally I respond to the threats in the northwest, and place a bandaged piece at d11, which seems a suitably annoying placement.  The idea is to blockade the 4-group extending from f10 from the Orange pieces in the top-left corner, reducing Orange’s potential for growth.  Orange is still able to grow their central group by connecting two more pieces at b10, but those four pieces around f10 are now isolated.  At best Orange could attach the piece at d10, turning the 4-group into a 5-group, but there’s no way to connect those pieces to the larger group below.

Here I get greedy, and try to connect my central group across the 2nd row along the bottom.  The plan is to construct a pair of pieces that I can then bridge between the bandaged pieces at c2 and f2, but I actually made a mistake.  By choosing the twist at e2f3, I actually ensured that I would be unable to connect my newly-formed pair.

Orange makes precisely the right reply here, placing the Yellow pair where I want it, but unbandaged.  Now I have no chance to connect across the bottom, because in order to secure those pieces I would need to bandage one of them, which means a twist must be made.  But, given that the pair is already in the right position, any twist I make will only move them out of position!

This sort of problem comes up fairly often in Permute, and I was annoyed at myself for not seeing it.  When you have set up your pieces properly, you should have them placed so that they are ready to be twisted into place, but crucially, are not already in place and unbandaged.  That means you can make a move at that location, twist the pieces into place and secure a bandaged piece in the right spot.  If you try a setup for connecting a few pieces but misjudge your plan, as I did, then your pieces may be prematurely placed, and your opponent is then able to prevent your plan at any time by twisting your pieces out of position and locking them down with a bandage.

In truth, this move sequence was overly optimistic anyway; given the two-space gap between c2 and f2 and my lack of well-placed pieces nearby, the only way my attempt to connect would work is if my opponent chose to allow me to make that final twist of the free pair.  I had no way to force that connection in the first place, and Orange had no intention of allowing it, so effectively that was a wasted move.  In a tight endgame, throwing away a move like that could cost me the game!

I cannot secure the connected pair between c2 and f2, so instead I opt to insert another bandage at g2, trapping another Orange piece on the bottom edge at the same time.  Orange then extends that bottom edge group further, but this also connects my pieces to the bottom-right corner group.

I still cannot secure the extension in the south, so instead I opt to harass Orange’s central group some more by blocking two pieces in on the western edge.  This has no effect on the size of my Yellow group, so it is a safe move that still causes some trouble for my opponent.  Orange extends the bottom edge group once more, after which I lock out two more Orange pieces from their central group, and at this point Orange resigns.

If we remove the bandaged pieces we can get a clear picture of the final position:

Permute 12x12 -- move41 -- count-01

There we have it — a final score of Orange 16 to Yellow 43.  Had we played to the bitter end, my final score would have been significantly lower of course, as Orange could disrupt my connecting pair at d2-e2 at any time.  But even then the score would still be convincingly in Yellow’s favour.

Permutation Principles

As you can see, the 12×12 game is intense right from the start, and opening moves are very consequential.  AI testing shows though that between strong players, the outcome is usually decided about 60-70% through the game (during these tests I did not allow the AI to resign).  So, while the opening is very important, most games will be decided during the middlegame.  Our goal in the early stretches of a 12×12 game should be to come out of the opening with a stable central position, with some viable options for future extensions of our largest groups; during the middlegame we want to extend our groups effectively and safely, while denying opportunities for our opponent to attack our groups or build their own.

Of course, if we want to reach that point in our games, we need some guiding principles to help us along in our efforts to become stronger players.  Here I will outline some basic principles of Permute play that will help you play more effectively.

Make multipurpose moves

This general principle applies to many games, of course, but in Permute every move is inherently multifaceted and directly affects both players.  Every twist we make moves opposing pieces as well as ours, since faces consisting of only one colour cannot be twisted.  This means that every single move has the potential to grow or shrink our opponent’s groups as well as our own.

When I am choosing a move in Permute, I tend to evaluate it on a few dimensions to see how it measures up against other candidate moves, and I aim for the moves that can accomplish more than one of these general objectives at once:

  • How much does it add to my groups?  All else being equal, a move that grows my groups more is of course preferable to one that grows them less.
  • How much does it disrupt enemy groups?  If I have a choice of two moves, both of which grow my group by 3 pieces, but one shrinks an enemy group by two while the other doesn’t, then the first move is a better choice.
  • Does this move gain space?  Gaining useful space on the board and restricting the opponent’s growth options is really important, especially in the early game.  If a move grows my groups less than another move, but allows me to connect a group to an edge and block enemy expansion for the longer term, that move may well be the more sound strategic choice.
  • Does this move have follow-up options?  A useful beginner’s metric to judge this is whether a move leaves you anchors — single pieces extending from a bandaged piece that can serve as a launchpad for further expansion.  Again this is more of a strategic consideration; sometimes a move that adds fewer pieces can still be stronger, if it produces anchors that allow for a bigger extension later.  Anchors also create threats for your opponent, as they must keep an eye on those anchors and consider which may need to be blocked, which in turn may disrupt their own expansion plans.

Like any complex strategy game, all of these guidelines will have exceptions, and there are many other ways we might judge the effectiveness of a move as we get stronger.  However, I believe that keeping these principles in mind is very valuable as you are learning the game, because each of them focusses on the longer-term consequences of our moves.

In Permute we have to fight against the impulse to play reactively, and start twisting purely in response to local threats from your opponent; each move in Permute can be so destructive that we often will feel the need to react to every attack, leaving ourselves with no chances to attack in return or to build a coherent plan for expanding our groups.  If we instead train ourselves to evaluate each move in terms of its impact on both the groups local to that move and the wider strategic situation, by considering space gains and anchors, then we are more likely to choose moves that not only protect our groups in the short term but also create longer-term challenges for the opponent.

Protect Your Corners!

This is a straightforward tip, but definitely worth emphasising!  In this game we saw the consequences of leaving a vulnerable corner on your largest group — the opponent gets an opportunity to slice the group in two, and if the group is already contained and there is no alternate path to reconstruct the connection, then that one move is enough to ensure your group cannot be saved.  However, we may not want to develop a habit of taking time to protect every corner in every group; instead, we must judge when a group has enough prospects for future expansion that we need to secure its position.

Stronger players may experiment here and there with group sacrifices, where they lure the opponent to spend moves on disrupting a vulnerable corner when their actual growth ambitions lie elsewhere on the board.  However given the high price for getting this kind of sacrifice wrong, this is a tactic that should only be used very carefully!  Generally speaking, if you spot a weak corner in one of your important groups, you should act immediately to protect it.

Bandaging is just as important as twisting

Each move we make in Permute is, to some extent, irreversible.  In Chess, if we move a Knight to the wrong place, we can potentially reverse that move in the future to recover our position, but if we make a bad twist in Permute, the bandaging ensures that at least some of the damage is permanent.  Because of this, we need to be sure that when we bandage a piece as part of an effort to grow a group, we protect as many of our pieces as possible:

Permute 12x12 -- bandaging examples-01

Here we can see some basic examples of this principle.  In the leftmost diagram, Yellow has created a secure L-shaped group — the bandage on d4 protects the corner, and ensures the pieces on d2 and c4 are fully protected as well.  In the middle diagram, Yellow’s misplaced bandage on d5 has protected none of the pieces in the middle of this group!  On the right, this T-junction formation is also safe, giving Yellow three potential avenues of expansion with the bandages on b4, d3 and d5.  Our objective when setting up our groups should be to maximise our reach, while always being careful not to leave weak points that can be twisted out of position.  One-space extensions, T-junctions, and secure corners are important ingredients in any effort to build a large group.

Conversely, if we are attacking the opponent, we should place our bandages so as to create maximum disruption.  Let’s go back to that fateful 12th move in the game, where I was able to split Orange’s group via the vulnerable corner, and zoom in a bit:

Permute 12x12 -- attacking examples-01

On the left is the move I made.  When twisting at c2d3, I placed my bandage at d2, a diagonal step away from the diagonal corner I had just disrupted.  This ensures that not only is the vulnerable corner at c3 permanently cut, but also the Orange piece at d3 is blocked in, preventing Orange from twisting around the obstruction to continue building the group.  The Orange pair at c1-c2 is still twistable, but cannot get around the bandage at d2.

In the middle diagram, we see the position if I had bandaged at c3 instead — note that Orange now has two pairs at c1-c2 and d3-e3 that can potentially be brought to bear, expanding the group around my obstructing piece.  On the right we see a potential result of this mistake: Orange twists Ad2e3-d2, completely isolating my piece at c3 and building the group southwards with a secure connection.

The lesson here is to always observe the local patterns of bandaged pieces when attacking an enemy group, and think not just about the immediate damage you can do, but also whether you can potentially block off any more pieces by setting up your bandages differently.  Make sure your attack has as much lasting impact as possible, and then you can make advances elsewhere while your opponent scrambles to repair what damage they can!

Split the board when ahead

At several points in the game above, you saw me aim to split the board into sections by connecting my main group across the whole board.  When you have a decent-sized lead and the opportunity, connecting securely can be very beneficial; by cutting across the board, you prevent the opponent making board-spanning connections of their own, and restrict their options for catching up to your lead.

Let us go back to the game for a moment and look more closely at my middlegame connection to the eastern edge of the board:

On the left is the move I played in the game, which connected my largest group to the eastern edge.  Orange is now firmly cut off from connecting their upper group to the bottom half of the board, and their attempt to move south has been contained.

On the right, we see what could have happened if I had not attempted to reach the edge, but instead had played elsewhere (in this case, bolstering my group in the bottom-right with Ch2i3-h3).  Orange is then able to generate threats by connecting significantly more pieces down that right side.

In this particular position Orange cannot make massive gains this way; on the right we see that after a couple more moves, Yellow is able to contain the advance once again.  But Orange did manage to gain two more pieces even in this constrained position, and in a close game two pieces may make all the difference.

So, when the opportunity to cut off your opponent presents itself, taking advantage of that opportunity is often a good way to consolidate your advantage and deny your opponent chances to create more threats.

Twists are local, scores are global

Finally, an important reminder — Permute is a game full of complex tactics, but the ultimate goal of the game is inherently global in nature.  When fighting for your groups to stay live, never forget that all moves should be in service of the larger goal: building the biggest group of pieces on the board.  If you get deep into a tactical battle and realise you will not be able to make any more profitable exchanges, do not be afraid to abandon your group and seek a bigger move elsewhere!

Similarly, we can easily get tied down defending only our largest group, and forget that the group scores in Permute are cascading — this means that if the two players tie for the largest group, then we compare the second-largest, then the third-largest if those are tied, and so on.  So if a game is close, remember to nurture a secondary group in addition to your largest one.  Tight games between equally-skilled opponents will most certainly have games go to the second-largest group, and occasionally the third-largest (though this is rare).

Remember too that playing strategically and planning your group-building effectively will also help you tactically; the more secure your groups are, the better-placed you are to defend them, and the more you constrict your opponent’s growth options with well-planned extensions, the more vulnerable they are to attack.

Next moves

At this point, hopefully you have a clearer picture of how a game of Permute flows, and you have picked up some useful tools in your toolbox for your next twisty battle.  Permute feels quite odd at first if you are used to something like Go or Chess, given the fact that the board starts full and every move directly affects both players’ pieces.  But once we spend some time getting acquainted with the properties of the game, these unusual aspects will start to feel natural, and you can focus on increasing your playing strength.

From here, I recommend having a bunch of games on the 12×12 board, and focussing on developing plans for each of your groups as you play.  Always keep an eye on the score, and check that each move is advancing your own groups while hindering your opponent’s.  Once you become confident navigating the sharp openings of the 12×12 board, then you can try moving up to 16×16, where you have an even more dizzying array of opening options, and the game opens out to become even more strategically interesting.

I have a large backlog of posts to work on at the moment, but at some point down the line I will come back with some more Permute tips, shown off through a fully-analysed 16×16 game.  If any of you out there are interested in a game on the larger board, let me know in the comments; I am always happy to find new opponents!

Some of you may have noticed too that Ai Ai lets you play Permute with four colours, which makes the game feel even more like a twisty puzzle.  I am still in the process of getting acquainted with the nuances of the four-colour game — which works very well with the 1[1]2[2]* move protocol — but I will cover this at some point as well.

In the meantime, I hope you will give Permute a try — Permute is my game so I naturally will have some bias towards it, but at least I can say that after many, many games of Permute so far, I have yet to get bored with it.  I hope some of you out there will have as much fun with it as I have!

Until next time, here’s a little preview of what’s to come — a tense game of 16×16 Permute that came down to a single point:

permute16-5s-tight1

Tagged , ,

A Beginner’s Guide to Hex, Part II: Sample Games

In Part I of our Beginner’s Guide to Hex, we covered some critical tactical and strategic concepts that will help you get a flying start in your journey to become a strong Hex player.  This time, we’ll look at a full game played on the 15×15 board between two strong players, so you can get a sense of how those principles manifest during actual play.  After that, we’ll take a look at another game, this time from the Game of Y, to see how Hex principles can apply in other connection games as well.

Commented Hex Game: LG #2195354, kspttw vs Arek Kulczycki

This game was played on a 15×15 board on LittleGolem.net, the most popular place to play Hex in correspondence style.   15×15 Hex is a relatively recent addition to Little Golem, but is already proving popular.  Other board sizes are also available on LG: 11×11 for quicker, more tactical games; 13×13, which is the most popular size (and is the focus of Matthew’s brilliant Hex strategy guide); and 19×19 for deeply strategic contests.

If you’d like to follow along with the game and investigate the variations Matthew recommends, you can do so via the online Hex board on MinorTriad; this site allow you to go through the game step-by-step and experiment with different moves at any point.  

diagram1_5-01

Move 1: Black opens at c2. This is a popular opening on the 13×13 board, but here I would classify it as being likely on the losing side.  All Hex openings are either winning or losing, there are no neutral ones; the best openings are right on the border between clearly winning and clearly losing openings, in that they may give some advantage but likely won’t immediately provoke a swap. On 15×15 c2 is probably weaker than on 13×13, since on larger boards we expect that winning openings will be a bit more centralised than on smaller boards. As a consequence, White elects to not swap sides.

Move 2: The 5-4 point (five rows from your own edge, four rows from your opponent’s edge) is by far the most popular acute corner opening. This move is connected by Template A-5, and it escapes second- and third-row ladders.

Move 3-4: This is a common joseki. Black’s (3) is connected by Template A-4, and escapes second-row ladders while threatening White (2). White’s response connects back to the edge with Template A-4, while restoring their ladder escapes.

Move 5: The 5-5 point is a popular opening in the obtuse corner. Lately, 4-4 has become more popular on the 13×13 board, a development spurred by the play of very strong AlphaGo-style bots. On the larger 15×15 board, it’s an open question whether 4-4 or 5-5 is stronger.

Whether on the 4th or 5th row, having a stone in your obtuse corner is always a plus. These corners are almost always taken early in the game.

diagram6-01

Move 6: White needs something on his southwest edge. You may be wondering why this of all spots was chosen. One principle is that you want stones that are “attacking,” in the sense that they can reach the edge through bridge moves to the side (see the dots in the diagram). Stones like this are harder to block. This is the closest a stone on the sixth row can get to the acute corner and still be attacking. If not for the black stone on c2, White would likely have played this move at f5. You want your stone as close to the corner as possible because it does more to disrupt your opponent.

Note also that (2) was “attacking” in this sense in the east corner. This is part of why the 5-4 point is so popular.

diagram7_9-01

Move 7: This is a popular combination with c2. Together, these stones form Template I-5, and they can escape just about any ladder from the second to sixth row along this edge. This Template is rather difficult to attack. Note also that this move invades the “attacking zone” of White’s f7 stone. 

Move 8: White plays here at the center of the board. This move is somewhat of a distant response to (7). With (7) invading the attacking zone of the f7 stone, Black could potentially block that stone now. But White doesn’t want to overcommit to this area — it’s important to spread influence around the board, especially early in the game when the potential lines of play are so fluid. This stone can support the f7 stone while also increasing White’s influence over the centre of the board and towards the northeast edge.

Move 9: Counterintuitively, in Hex it’s often stronger to play on your opponent’s edge than on your own. Having stones separated by two empty hexes on the opponent’s second row can be quite strong. Notice how White is unable to fit Template A-3 in between (7) and (9) (nor in between (7) and c2). This stone (9) is basically connected to Black’s northwest edge now, because White’s only reasonable blocks (such as at b6 or b7) allow Black to go around to the north (at c6 or c7, respectively). Ultimately, White will need to connect to the southwest edge in the area between (9) and the obtuse corner.

diagram10_19-01

Moves 10-19: White can’t break the connection of b8 to the northwest edge. Instead, White allows Black to connect. But how does this help White? This is an attack I call “undermining.” Note first that the 10-12 group is connected to White’s edge (White can connect at either A or B; Black cannot block both). Because of this, the points D and C are vulnerable for Black. If White intrudes into these bridges, they threaten to connect to the edge through 10/12, so Black is forced to reply. White can therefore invade these points for valuable territory. Additionally, towards the end of the sequence, White obtained the stone (18), gaining more territory and forcing the reply (19). Lastly, the point E is also vulnerable for Black, if they want to keep b9 connected to the edge.

diagram20-26-01

Moves 20-26: Now White begins attacking the vulnerable points, starting at (20). After Black saves the connection, White moves to the other end to attack Black’s 5-5 stone in the obtuse corner, forcing a response there. White knows that after taking the territory at the vulnerable points, they will ultimately need to connect to their southwest edge somewhere in the lower half of it, so having 22 available strengthens this area to White’s advantage. With this secured, White returns to attacking Black in the west corner with moves 24 and 26.

diagram27_30-01

Moves 27-30: Clearly, White has been playing in sente for a while, dictating the direction of play ever since the attack that began on move 10. Black seems to have had enough, and rather than respond at b7 (and handing play back to White), Black tries here to take the initiative back with a block against the h8 stone. Here the play suddenly becomes a tactical affair. White first plays (28), because having these two stones parallel to the northeast edge will virtually guarantee their connection (thanks to the help of the white stones in the east acute corner). Next White plays (30). 

Here White begins to utilise that territory they gained through undermining. Since move 30 is connected to g4, White’s approaches have expanded considerably. (30) threatens to connect to f7, and g4 threatens to connect to e5. Before we look at Black’s response, it would be instructive to see what happens if Black tries to block the former with 31.g6. White might then respond at 32.f5. From here, a few lines to consider:

33. d6 e7 d8 d7 b7 (forced) c8 b9 and then c10 is a ladder escape fork for White, connected to the edge with Template A-3, and back to the main group by either c9 or e9.

33. b7 d8 (connected back by either f6 or d7) c9 d10 (d10 + d12 make for the edge template L-4, hence d10 is connected to the edge) d9 and now f9 is connected to the edge and threatens to connect to either h8 via g9, or to h5 via f8.

diagram31_42-01

Moves 31-42: Black attempted to block at (31) instead. The situation that follows is highly tactical. White first plays 32. d8. After Black blocks at 33. g6, White links up g4 to e5 with 33. f5. Black gains some free territory with a bridge intrusion (35-36). At this point, e5/f5 are connected to d8/f7 by either d7 or f6. If Black blocks at 37. f6, White’s connection is assured by 38. d7 (White can connect by either b7 with Template A-2 or c9 with Template A-3), so Black plays 37. d7. White connects up with f6 (although e7 would have been better, offering no intrusion points). Black takes a little more territory with move 39, but after move 41 White’s responds with 42. d10. As mentioned in the second variation above, this move (along with the stone on d12) is connected to the edge via edge Template L-4. If Black blocks at 43. d9, White plays 44. f9, which threatens either g9 or f8.

diagram43_46-01

Moves 43-45: With the southwest edge lost, Black must attempt to block the northeast edge. The odds aren’t good however. Presumably Black went all in on blocking White from the southwest because Black felt the chances were better there. Black didn’t gain much in the way of territory during that sequence that could help on this side of the board, with the possible exception of the stone on h4. 

The White stone on h5 isn’t yet connected to h7, but trying to block between them will just make things worse for Black (43. h6 i5 i6 and k4 can at best be held to a fifth-row ladder, heading towards the White stones in the east corner). So Black plays 43. j6. White connects up the smart way, at 44. i5. Unlike connecting via h6, this threatens the followup k4, which would start a fifth-row ladder, as well as the potential threats from the h7/h8 group. Black is forced to attempt to block both directions at once, with 45. k6. This cleanly blocks the potential of k4, but as we shall see, can’t hold off a White attack from h8.

Move 46: White bridges away from h8. This move is based on a simple concept: note that by placing the stone out so that there’s a clear line to the edge (shown by the arrows) it can’t be stopped with simple adjacent blocks (j9 j8 k8 k7 … ). This means Black will have to block this stone to a ladder, and that’s where those two white stones on l11 and k12 will come into play.

diagram47_50-01

Moves 47-50: All that remains is for White to finish off the connection. After 47. j9, White will ultimately play j8, after which a fifth-row ladder will begin (Black could hold White to either a fifth- or fourth-row ladder; generally you want to hold a player to the higher row). Before that, though, White sets up the ladder escape with move 48, which threatens to connect back via a bridge to the stone on i9. Black is forced to block (Black’s choice to play 49 at j10 as opposed to i10, is the stronger block since it leaves White with slightly less space underneath). Now White plays 50. j8, and the game is over. Although the final sequence wasn’t played, let’s quickly look at how it might have played out.

diagram_end1-01

The naive approach is just hold White to the fifth-row ladder. White easily connects with Template A-4.

diagram_end2-01

Black might instead jump ahead with move 55 and force a bottleneck, but after move 58 White connects to the bottom with Template A-3 and back to (50) via either A or B.

diagram_end3-01

Finally, Black might try to hold White to a fourth-row ladder instead, but after move 58 White’s stones are connected in the Trapezoid template, and Black has no means of blocking White from the edge. 

 

Y Sample Game: PCM vs Matthew Seymour

Next up we have a sample game of the Game of Y.  For those of you who don’t know Y, it’s actually even easier to learn than Hex:

  1. Two players, Black and White, compete to connect all three sides of a triangular board of hexagons.
  2. Players take it in turns to place one stone of their colour on any empty square on the board.  The first player to connect all three sides of the board with a single connected group of stones wins the game.

That’s it!  In Hex, players must connect two specific sides of the board that share their colour, while in Y all three sides are relevant to both players.  As we shall see, that fact can alter some of the tactics and strategies you may have learned from Hex, but broadly speaking your Hex knowledge is a great help in Y as well.

This game was played between PCM (Black) and Matthew Seymour (White) on iggamecenter.  The board is size-14, which is relatively small for Y but still big enough for a challenging game.  Matthew has annotated the game for us below:

Game of Y -- 14 -- mv4-01

Move 1: We’re playing with the swap rule, so Black (PCM) opens along the edge.

Move 2: White (Matthew) responds with a more central move.

Move 3: Connected left via the A-5 edge template, but the difficulty will be connecting to the bottom.

Move 4: Blocking Black’s stones from the bottom edge.

Game of Y -- 14 -- mv8-01

Move 8: Connected with the B-3 template.

Game of Y -- 14 -- mv10-01

Move 10: A blunder! e9 would have been better (winning I think) than f10, with template C-5 facing the left edge and move 4 helping guarantee the 8-2 group’s connection to the south. As it stands, 2-10 is connected south with C-5 and 2 isn’t fully connected to the left.

Game of Y -- 14 -- mv15-01

Moves 11-15: This block sets up a ladder with the bottleneck formation.

Move 16: Ladder escape. The plan here is that after b7, White plays d6 d7 f7, and now White is connected to all three edges.

Move 17: Counter-threat, threatening the connection between (8) and the edge.

Move 18: 16-18 is connected to the right edge through the M-4a template, and connected to the central group (14-8) via either d7 or the ladder on the left.

Game of Y -- 14 -- mv21-01

Moves 19-21: Black first blocks the d7 route, then blocks between the ladder and the escape on move 21.

Move 22: Here I blunder the game away! I was concerned the 20-8 group might lose its connection to the right that I had through either 8 or the 16-18 group. It looked like Black had blocked off the 16-18 group, so I had to save it via (8). I missed the winning move 22. a6.  Then, if 23. i10 I could simply play a5 and I would be connected to all three edges — in other words, it would have kept the double threat alive for connecting to the right, while also connected the group to the left.

Instead, I saved the connection to the right, but now Black can now cut me off from the left at a6. I missed this “obvious” move because (I think) I’m so used to playing Hex. 21 is connected to the left via A-2, and in Hex there’s no reason to ever invade A-2 because the two empty hexes are captured. But of course, in Y, the edge is shared by both players, so these hexes are NOT captured.

Game of Y -- 14 -- m23-01

Move 23: Forced. Black blocks White from the left edge.

Game of Y -- 14 -- m30-01

Moves 24-30: Ladder, followed by a break. The black group (1-29) is connected to the left and right edges. Black needs only to connect it to the bottom to win.

Game of Y -- 14 -- m36-01

Moves 31-36: 4th-row ladder, followed by a bottleneck. White has no hope however, as the 7-5 group will help escape the ladder.

Game of Y -- 14 -- m43

Moves 37-43: There are many ways to escape the ladder, but Black elects to go with this approach. More straightforward would have been e13 f14 f13 g14 g12 and then Black can play either h13 (with A-2) or j12 (with A-3). As played, (38) is the only reasonable reply to (37) (further left on this row, Black plays e11; further right on this row, Black plays g13; for plays on row 12, Black uses (37) as a second row ladder escape). (39) and (40) accomplish nothing but there’s no harm. After (41), Black can play either h13 (with A-2) or j12 (with A-3). White blocks the former, so Black plays the latter. White resigns.


So, there we have it — a quick but well-played game of 15×15 Hex, and a tricky game of Y that shows off some of the quirks of Hex’s cousins in the connection-game world.  We hope these give you some useful ideas about how to apply the core concepts of Hex strategy to your own play.

Let us know in the comments what you think, and if there are other subtleties to Hex (or Y, for that matter) that you’d like to hear more about, perhaps we may do some more posts in the future.

In the meantime, enjoy, and good luck with your journey toward becoming a strong Hex player!

 

 

 

 

Tagged , , , ,

A Beginner’s Guide to Hex

Some time ago I talked a bit about Hex on this blog, discussing its history and how it kickstarted the entire connection game genre. Since then, a few readers have asked for a bit more detail on how to actually play Hex. So, for this post I have teamed up with Matthew Seymour, author of the brilliant Hex: A Strategy Guide, and we have put together this beginner’s guide to Hex strategy and tactics.

Below we will introduce you to basic tactics, templates, openings, and strategic considerations. Each section is very brief but will give you enough to get you started on improving your Hex play; after you tackle each section, please continue your studies with Matthew’s guide, which has much more detail and numerous examples of each critical concept.

Basic Tactics

The edges

Edge play in Hex is obviously critical, since in order to win we must connect our two edges across the entire board. Initially, connecting a chain of stones successfully to the edge seems like a baffling enterprise — an adept opponent can bend and twist your attempted connections away from the goal, and it can be difficult to understand how to set yourself up for a strong connection.

Edge templates are extremely useful tools for understanding edge play. In our previous discussion of Hex, we met the bridge, a configuration of two stones that proves to be virtually connected even though the stones aren’t adjacent.  Similarly, edge templates show us configurations of stones that, with correct play, are guaranteed to result in connect to the edge. Templates indicate patterns where, even if your opponent has the first move, you will always be able to connect successfully. Your opponent may intrude on your template, but so long as you defend your template at every step, you will succeed.

This means however that templates are always just a move away from breaking. If you have two overlapping templates, and your opponent plays a move that intrudes on both, you can only potentially save one of them.

Here are some examples of basic edge templates:

Our first edge template is the very basic A-2 template, consisting of a stone on the second row. If White tries to block at A, Black can respond at B (second diagram), and vice versa. Either way, White has no means of stopping black. This template is very similar to the bridge template.

One of the most useful templates is template A-3, consisting of a single stone on the third row. This template comes up very frequently. We’ll analyse the situation by considering Black’s possible threats. On one hand, he could play at A and connect to the edge with template A-2. Or, he could play at B, and connect by a chain of a bridge and template A-2. The important part here is that these threats don’t overlap. If White plays at A, or one of the two hexes below it, Black simply plays at B and connects by that route. And if White plays in any of the 5 hexes on the right, Black plays at A and connects by that route.

We can succinctly convey this information in what’s called a pivot diagram. The two small dots indicate Black’s alternative moves. We can see that the rightmost dot connects back to the top stone with a bridge. Both dots are connected to the edge by one of two possible moves (template A-2).

Lastly, let’s look at Template A-4, on the fourth row:

diagram5-01

Similar to A-3, Black has two threats: to play adjacent at A, connecting to the edge with A-3 (left), or to bridge to the side at B, also connecting with A-3 (right).

These alternatives can get around most possible blocks by White. But there’s one issue, the point C. Both the threats above overlap at this point. So Black needs a response to a block at this point. The solution is shown below (note that if White played (3) on the left side, Black could bridge to the right instead).

diagram8-01

There are a large number of edge templates. You can find an excellent catalogue here or here.

Interior templates: connecting chains

We’ve already encountered one interior template — a template for connections in the centre of the board, away from the edges: the bridge. As you might expect, other templates exist for more complex configurations of interior stones. Knowing these templates is very helpful indeed, as once you achieve such a configuration on the board and recognise that template, you can play elsewhere, knowing that you’ll be able to successfully defend that template against later intrusion.

Some well-known interior templates are the Wheel, the Crescent, the Span, and the Trapezoid.

Defending at a distance

We’ve talked a lot so far about connecting — positive, attacking play. But what to do if our opponent has us on the back foot, and a deadly connection is looming? How can we stop them?

First, we should remember that as satisfying as that positive, attacking play may be, Hex is what is called in combinatorial game theory a hot game. This means that it is always beneficial to make a move in Hex, and no stone of ours on the board is ever a detriment to us. On top of that, one player must always win in Hex, so if we successfully block the opponent from any possible connection between their edges, we’ve in fact won the game. So don’t neglect defensive play — in Hex, it’s precisely as powerful as attacking play, and will win you the game just as effectively! Defensive moves in Hex are also offensive moves.

Having said that, defence in Hex can be a delicate affair. As we’ve seen in numerous examples, stones can be connected even if they aren’t adjacent, and skilled players can move across the board at high speed, staying connected the whole way. Therefore, if we attempt to block simply with adjacent blocks — playing right next to opposing stones — a skilled opponent can easily bend around us. Likewise, if we block at a distance but misjudge the situation, the opponent may still get round us by using bridges to cover enough ground to do an end-run around our defences.

In practice we may need to combine adjacent blocks with more distant blocks in many cases. The adjacent blocks restrict the opponent’s options for bridging forward, while the distant blocks contain those advances:

The classic block short-circuits the opponent from afar, allowing the defender to respond whether the opponent advances directly forward or takes a more oblique approach:

In either case, in order to defend we need to evaluate the opponent’s options for onward connection, and place our stones in anticipation of those options. If we defend reactively, and follow the opponent around right at their heels, then we’ve no hope of survival. If we instead constrict their choices and contain their subsequent advances, then we may just keep them at bay.

Ladders

Often, when approaching an edge of the board, you’ll end up in a situation like this:

ladder1-alt-01

White wants to connect and Black can’t allow it, so Black blocks at (3). White can’t connect right away but can force Black to carry on blocking all along the edge. This series of back-and-forth adjacent plays is called a ladder. White is completely in control here; Black has to respond to every ladder stone White plays, otherwise White’s connection is assured.

Go players will be familiar with sente — the concept of maintaining the initiative, by making moves that force an immediate response from the opponent. When you have sente you are in control of the game; you are making profitable moves, and all your opponent can do is match you, stone for stone, unable to direct play to their advantage. Sente is just as important in Hex as it is in Go, and ladders are one common manifestation of it.

However, you have probably noticed that if White continues playing the ladder here, it’s Black that ends up connecting across the board:

ladder1-01

To make the ladder profitable for them, White needs to incorporate some additional tactical plays. If White had an additional stone in place to form a ladder escape, then when the ladder reaches that stone, they can connect to the edge with ease.

ladder1-escape-01

Of course, when a ladder is already forming, taking a turn to place a ladder escape stone simply dooms the ladder. So players will often place ladder escape stones during the opening phase of the game, to allow for profitable ladder play later.

Another option is to place a stone that is both a ladder escape and a threat to connect by another route. Your opponent will be forced to block either the threat or the ladder, and then you can connect by the other means. In the diagram below, Black plays a stone at 1. This can escape the ladder, but it also threatens to connect via A (with two bridges). White can’t block both approaches, so Black will connect. This is called a ladder escape fork.

diagram14-01

If you have no ladder escapes or forks available, you’ll have no choice but to “break the ladder”, as Black does with move 7 in the diagram below (note that 7 is connected with Template A-3). In the acute corner, this tends to reverse the roles: notice how now it’s White who’s the attacking player with a ladder.

diagram13-01

Strategic Considerations

Openings

The opening in Hex is an interesting moment, as the first player needs to consider not just what is the best move to play, but what is the best move to play that won’t get swapped. In Go or Chess, you can play your opening move without fear of that move suddenly becoming your opponent’s opening, but not so with Hex!

As the second player, you need to do the opposite calculation: has my opponent played a move that, with perfect play, would give them a winning advantage? If so, I should swap; if not, I can safely play on.

On smaller boards, Hex has been solved, meaning that we know the precise outcome of any given opening move. That leads to diagrams like these:

Above are three diagrams showing the ultimate winner, with perfect play, of opening moves played at every cell on the board. There’s an important trend to notice here — the winning openings for any given board size are not straightforwardly extendable to larger boards! While we can see a general theme that opening moves in the centre are stronger than those on the edges, the specific outcomes of those edge cells change as we change the board. That means that on the boards we humans play Hex on — from 11×11 upward — not only do we not have these convenient maps of what moves win or lose, but we cannot use the opening maps from smaller boards as a definitive indication of the outcome of any opening on the bigger ones.

Matthew’s guide focusses on the 13×13 game, and for openings on that board, he’s produced a swap map that can help guide you in the opening. The cells with black dots are Matthew recommendations for good opening moves for Black. When you are the second player, if your opponent opens anywhere in the shaded area, you should swap — those moves have the potential to give a winning advantage, so you’re better off taking that stone for yourself. If your opponent plays outside that shaded zone, let them carry on — you can possibly do better by playing your own opening.

swap-map-13x13-01

For larger board sizes, like 15×15 and 19×19, we don’t yet have enough games played at a high level to put together a reasonable swap map. However, we can make some reasonable inferences about good opening moves; in particular, opening in the obtuse corners seems a good way to go on all board sizes.

These are very simple principles, but should be enough to get you started. One thing to bear in mind is that we humans are far from perfect play, even on 11×11, so both sides are likely to make mistakes, not just in the opening but throughout the game. So our goal at this stage should be simply to ensure that our opening doesn’t obviously disadvantage us; we don’t need to fret too much about whether a particular move is 100% winning or losing.

Playing in the corners

The Hex board has two types of corners — acute and obtuse — that have different properties. Corners are the only parts of the board where your stones can both strengthen your own position and weaken your opponent’s, and for that reason, players tend to play stones in the corners early in the game. Typically you will want to play stones in at least one corner on each of your edges during the opening.

The corners being so important often leads to pitched battles to establish control over them, and so strong players may study corner patterns (think joseki in Go) to navigate these tactical scuffles. If you don’t have a presence in a corner and your opponent does, invading is useful in order to reduce their influence there, and corner patterns will help you to reduce that influence. Conversely, if your opponent invades your corner, you can use these patterns to settle the fight and maintain as much of your initial influence as possible. The challenge in these situations is judging when there is no more profit to be gained, and thus when it’s time to move on from the corner battle and establish yourself elsewhere.

There is a lot to discover in these corner patterns, but don’t worry too much about these early in your Hex journey; as you start to face stronger opposition and find your corner play is letting you down, refer to Matthew’s guide for detailed examples of how to fight for the corners.

Influence

We’ve alluded to this concept in the previous section, so now let’s expand on what influence means in Hex strategy. Stones in Hex are not just localised points — they have impact on the board around them and on other nearby stones. Every stone has the potential to connect to something or to block something else, and when placing our stones we need to consider the influence of the stones around our planned placement.

In the early stages of a Hex game, gaining influence is important. We would do well to place our stones around the board, to spread them out; this maximises the potential influence of each stone. Conversely, if we don’t spread our stones out, we may have a strong influence in a particular area but will be weak elsewhere. If we are struggling to find an effective place to play, we can look at our relative influence on different areas of the board; if we find some areas where we have low influence, those might be good places to play our next moves.

As you might expect, stones in the corners have a high degree of influence — the proximity to two edges means those stones are better able to restrict your opponent’s activity in that area and force them to work around you in a limited space. Placements in the corners also are tougher to block, and provide you with ladder escape stones for later in the game.

The edges are somewhat less intuitive. We might feel secure playing near the centre of our own edges, as this seems a useful way to block the opponent, but in practice these kinds of placements do not provide strong influence. Instead, we should play near our opponent’s edges — this forces them to work around you and makes it harder for them to connect.

Beginning Go players often play in a style referred to as Puppy Go, where they continually play very close to every one of their opponent’s moves, following them around the board like an excitable puppy. We can easily be tempted to play Puppy Hex in a very similar way. Unfortunately this is an adorable, but poor strategy; in a Puppy Hex scenario your opponent is dictating play completely, and since you are always one stone behind they will have free choice of where to establish influence and you will always be playing catch-up. Always keep an eye on the broader board situation, and try to take the initiative when the situation allows it — don’t let your opponent drag you around by the nose!

As you become more comfortable playing in an influence-oriented style, you can start to focus on making moves that serve multiple purposes. Gaining influence is good, but gaining influence and blocking the opponent is even better! This is a challenging step, requiring you to have both tactical and strategic vision, but as you gain more experience and become able to recognise common tactical motifs, you’ll be better able to keep these in mind as you seek to expand your presence across the board as well.

As a final note, we should remember that Hex is fundamentally a scalable game — we can play Hex on any size board we like without changing the rules, but the feel of play will change. Hex on larger boards is a challenging and rewarding affair, but specific tips on those epic battles is beyond the scope of this article. However, we encourage you to try larger boards, as they by necessity will make you play in an influence-oriented style. With so much additional empty space on the board, you’ll need to learn to anticipate where battles for influence and territory will rage, long before they actually happen. That experience can help you on the smaller boards too, training you to think globally more consistently.

Territory

Territory is a critical concept to understand in Hex strategy. Think of territory as the potential your stones create for future connection; the more territory you control, the more tactical options you have for later attempts to form connections between your stones.

As a starting point, we might say that each stone creates territory in the area immediately around itself; in other words, the empty hexes immediately adjacent to it. However, as we see below, this definition falls apart fairly quickly:

useless-stones-01

These intrusions by Black gain no useful territory. In both cases, White simply blocks any onward connections, so the ‘territory’ gained (the shaded cells) offers nothing that Black didn’t already have!

If we believe that stones create territory regardless of their disposition, then we will run into situations like the above, where our stone is effectively a wasted move, as it will never actually be able to connect to anything. Instead we should restrict the definition a bit more: the territory around our stones consists of the adjacent hexes that could in theory participate in a connection. If we want to invade somewhere and gain influence from that play, we need to be certain that the placement provides useful territory; if the stone does not gain territory, then we have simply placed a stone for no real purpose. Without territory we cannot claim influence, as the enemy can simply work around us at no real cost.

Taking the initiative

As in many other abstract games, in Hex gaining the initiative is of huge importance. Recall the Puppy Hex discussion earlier — imagine if we could force the opponent to play Puppy Hex. If we can place stones with aplomb while our opponent can do nothing but respond, we can dictate the pace of play and dominate the board at our leisure.

Here we will go in-depth into some Go terms we mentioned earlier: sente and gote. In Go, when we play a stone that forces the opponent to respond — because a group is threatened with capture, for example — we say that is sente, meaning we are gaining the initiative. Our next move after the sente move is essentially free; the opponent’s response is mandatory, so our next placement can be anywhere we like, and we can use that to gain influence or territory. Conversely, the forced response the sente move creates is gote — we are forced to be the puppy for that move and play where the opponent demands.

In Hex we also have sente and gote moves. For example, we may recognise that our opponent has an edge template in play, so we may choose to intrude on that template and gain some influence. That move is sente because it demands a response; the opponent must play to save the template, otherwise that connection is lost. At that moment our opponent’s move is gote, lending us the initiative.

As we gain more experience of Hex strategy, we will be better able to identify opportunities to gain sente. At the same time, we must be mindful of our opponent’s threats, and remember that playing gote moves to save a critical connection is vital too! We should try to avoid being the Hex puppy whenever possible, but sometimes there’s no escaping it.

Tenuki

Let’s look at another situation:

tenuki1-01

Here Black is threatening to cut the stone A off from the top-right edge, and the straightforward response would be for us to save the connection and take gote, such as by responding at L3. After all, by not playing there we lose the connection.

However, in this situation we can see that White has an opportunity to make an intrusion of their own, on the other end of the board. Black’s threat depends on using the stone B to connect to the bottom-right edge of the board. White’s board situation will allow them to make other connections, even if they sacrifice the connection under attack by Black, but Black’s situation is just as fragile. In cases like this we may elect to tenuki — to play away from the threat and allow our opponent to break the connection. Instead of defending against the threat we attack elsewhere, and now they must make a choice: either save their own template, or finish ours off. If they finish ours off, they must make a second move, giving us influence elsewhere; if they take gote to save their own connection, then we have regained the initiative.

In this game, White elected to attack Black’s B stone with move (2), rather than save the connection of A to the edge. Black elected to save the connection, playing out a standard joseki sequence, leaving White with the initiative.

tenuki2-01

Tenuki is an advanced concept, and often difficult to judge. In general, you will have more opportunities for playing away from threats in the early- and middlegame, when the board is less full and there will be opportunities for other connections. In the late game, typically both players will have committed many stones to particular connections, and there is inherently less flexibility; if we ignore a threat, we are more likely to hand the win to the opponent.

The Joy of Hex

We’ve covered a lot of ground in this post — over the course of these few sections we’ve gone from the basics of the board geometry through to advanced strategic play. Yet for all that, we’ve barely scratched the surface. From here, you can move on to Matthew’s detailed guide to Hex, and dig deeper into all of these concepts. While you’re there, be sure to try out his fantastic collection of 500 Hex puzzles (also available in PDF, in Hex style and Go style) to sharpen your tactical vision.  If you need help with openings, he also used over 6,000 online games on 13×13 to generate a very useful opening database.

Having said that, resist the temptation to power through all this material. Take some time with these concepts, apply them to your games, and move on only when you feel comfortable and confident. Remember too that Hex is perhaps the most famous modern abstract strategy game, but it is still very new in the grand scheme of things. Traditional games like Go, Chess and Shogi have had centuries for strategies to be developed, whereas in Hex we are all still beginners in some sense! So there is always more to discover and more to learn.

If nothing else, we hope this brief introduction will give you an appreciation for Hex’s incredible depth and nuance. Hex is a disarmingly simple game, so much so that a brand-new player may be tempted to ask ‘…that’s it?!’ when told the rules for the first time. But within that sparse framework lies a world of intricate tactical and strategic variety. This simplicity means Hex also has amazing flexibility — we can play lightning-fast blitz games on 11×11 boards, strategic masterclasses on 19×19, or mind-bending, baffling escapades as long as a game of Go on 26×26. Each one of these configurations is rich with possibility. Learning Hex also benefits you in other connection games — the tactics you learn here can transfer to other games, like the Game of Y (more on that in our next post).

Above all, we hope you have fun with the game! Go spend some time testing out your strategies online, entering tournaments, analysing games and writing about them. But alongside that, teach your friends and family (when Covid restrictions allow!), help them learn some basic tactical and strategic concepts, and show them why you love it. Every new player we bring to the game makes Hex’s future ever brighter, so the more we help others to see what the game can offer, the more enjoyment we’ll all have in the years to come.

Next moves

In the second and final part of our Hex mini-series, we will analyse a complete game of Hex in detail, and show how the concepts we’ve introduced here play out in a game between strong players.

Then we will analyse a brief game of Y, as well, to demonstrate how Hex concepts transfer to other, related games — and we’ll point out how some concepts change when we move to a different game.

Extra nerd stuff

Check out these papers if you’d like to know more about how the small-board swap maps above were generated:

SOLVING 7×7 HEX: VIRTUAL CONNECTIONS. AND GAME-STATE REDUCTION. R. Hayward, Y. Bjomsson, M. Johanson, M. Kan, N. Po, J. van Rijswijck. Department of Computing Science, University of Alberta, Edmonton, Alberta, Canada.

Solving 8×8 Hex.  Henderson, Arneson and Hayward, IJCAI 2009.

9×9 Hex: Scalable Parallel Depth-First Proof Number Search.  Paulewicz and Hayward, Proc. Computers and Games CG2013, Springer LNCS 8427 (2014) 138-150.

Tagged , , , ,

Quick picks: interesting abstract games in brief

As some of you will be aware, as a way of keeping myself occupied during the pandemic I’ve learned how to use Adobe Illustrator to design stuff.  A particularly enjoyable, if slightly odd, area of design I’ve gotten into is designing game boards for abstract strategy board games.  I’ve had a good time getting to know the software and experimenting with many different designs, and now that nice neoprene game mats can be custom-printed for affordable prices, I’ve actually gone ahead and had some of my designs printed out as well.  Hopefully, in some theoretical future where the pandemic is over, I can use these boards to introduce friends and colleagues to some of my favourite games.

I’ve made a lot of boards over the last year, so rather than wait until I can find the time and energy to write detailed blog posts on all of the games that go with them, I thought I’d share a few abstract strategy gems with you with just a few sentences about why they’re interesting.  Each brief review includes links to full-size images of the boards I’ve made for each game, which you can print if you wish.  Some of these games will get covered in detail in the future; for now, hopefully these short descriptions will entice some of you to give these games a try.

As a side note, I can output these designs in a huge number of formats — PDF, PNG, JPG, SVG, whatever — so if any of these strike your fancy but you need a different format, just let me know in the comments and I’ll upload it for you.

Catchup

Catchup is a wonderful game by Nick Bentley that I’ve mentioned briefly before, because the scoring system inspired my choice of scoring system for Permute.  This is a game I’ll definitely cover in the future, as it’s incredibly easy to learn, yet within moments of starting to play you’ll realise the core strategic dilemmas at the heart of the game.  Catchup is a really dynamic and exciting game, and personally I think Catchup is Nick’s best design by far.

Why it’s great: Catchup’s unique feel stems from its unusual movement protocol: each turn, you place two stones anywhere on the board, unless your opponent equalled or exceeded your score after their last move, and then you can place three stones.  The winner is the player who forms the largest group of connected stones at the end of the game, so the result is a tense back-and-forth where you absolutely must connect your stones to win, but each time your biggest group becomes equal to or larger than your opponent’s, they get a much more powerful move with which to fight back.

About the boards: The board on the top left above is a standard hexhex board, seven hexes on a side, with a scoring track where players can place a stone on the number representing the size of their current largest group.  The other five are variant boards with uneven sides, which an experienced Catchup player has suggested may generate more interesting play.

Chess: Supersized

These are simply enlarged chessboards — 10×10 squares and 12×12 — that I plan to print on mats and use to play large variants of Chess.  Many Chess fans over the years have attempted to transport the magic of the Royal Game to larger boards, and thankfully a number of them succeeded in creating some very enjoyable variants that feel like Chess, but still have a unique personality.  I’m planning to write an article in the future that will cover a bunch of large Chess variants and give you some detailed recommendations; for now, here’s a few worth checking out on both board sizes, should you fancy giving them a go.

Some recommended 10×10 Chess variants: Caissa Brittania (checkmate the Queen instead of the King!), Decimaka (hybrid of Chess and Maka Dai Dai Shogi), Elven Chess (hybrid of Chess and Chu Shogi), Grand Chess (Christian Freeling’s most famous Chess variant), Grand Shatranj (ancient Persian Chess brought to 10×10), Omega Chess (commercial variant with Wizards and Champions), Opulent Chess (Grand Chess but more my style — higher piece density, less wild tactically), Shako (Chess with Cannons and Elephants).

Some recommended 12×12 Chess variants: Chu Shogi (the best 12×12 Chess-type game, period), Gross Chess (mix of Grand Chess, Omega and Asian variants, very playable), Metamachy (fast-paced Pawns and crazy historical pieces give it a unique and fun feel), Zanzibar-XL (dense and diverse piece selection with a variable setup).

Exo-Hex and Iris