Check out Atomic Chess, our featured variant for November, 2024.

On Designing Good Chess Variants

This article is an investigation into what it takes to design a good Chess variant. It covers both design goals and design principles. Design goals refer to top-level abstract qualities that any good Chess variant should have. Design principles refer to various means of meeting design goals.

What is a Chess Variant?

To understand the qualities of a good Chess variant, we first need to understand what a Chess variant is in general. A Chess variant is an abstract turn-based board game that is similar to, related to, or based on Chess. Two of the most widespread and important similarities are that each player controls an army of pieces that have different powers of movement, and the game can be won by checkmating (or sometimes capturing) an opposing royal piece, usually the King. Although a Chess variant may sometimes dispense with one of these, I don't know of any game that dispenses with both and is still called a Chess variant. As proof that a Chess variant must have at least one of these qualities, Checkers is very similar to Chess except that it lacks both of these qualities, and no one rightly regards Checkers as a Chess variant. Like Chess, it is an abstract turn-based board game played on the same board, in which each player controls an army of pieces. But the pieces are all the same, and the goal is elimination, not checkmate.

Qualities of a Good Chess Variant

Before setting out principles for designing a good Chess variant, it is important to have in mind what qualities a good Chess variant has. These describe the goal, whereas design principles will describe means of reaching the goal. Here are important qualities of a good Chess variant. This list includes both top-level qualities and subordinate qualities. The top-level qualities are most important, and the subordinate qualities are common components of these qualities.

Besides the qualities listed above, there are some overarching qualities that incorporate some of these together. Tom Braunlich's "Designing a Variant" article, which appears in Pritchard's Encyclopedia of Chess Variants, recommends elegance. It says, "The main attribute to look for in any good chess variant is elegance. An elegant game combines minimum rules with maximum strategy." Elegance is essentially a combination of simplicity with depth and challenge. Another overarching quality is beauty. Beauty is known mainly by intuition, but it is affected by the various qualities that make a variant good, with elegance and balance being two of the most important.

The Classics: Chess, Xiang Qi, and Shogi

The three most popular and enduring Chess variants in the world are Chess, Xiang Qi, and Shogi. These are the predominant forms of Chess that evolved in Europe, China, and Japan. Let's consider how well each one meets the qualities listed above. I have played all three games several times, and I can attest from personal experience that all three are very enjoyable. Each one is exciting, satisfying, of suitable duration, and not overly drawish. Of these three, I find Shogi the most exciting, because the ability to drop pieces gives it greater dynamism than the other two games have, but I also find its opening game a bit more tedious, because it has more slow moving pieces. Besides my personal experience, the large followings of these games further testifies to the enjoyment people find in them. I can also attest from personal experience that these three games are very playable. Their rules are simple enough to easily learn, remember, and apply. None have weird pieces whose powers of movement are difficult to remember, or arbitrary complicated rules, or too many pieces to easily remember. Also, it isn't too difficult to think ahead several moves in these games. The literature on these three games attests to their depth, and the existence of many strong players for these games attests to how challenging they remain for experienced players. All three games are fair enough that the player who makes better moves normally wins. In all three games, both players start out with equal forces, and the only opening advantage is the ability to move first. This slight advantage will make a difference mainly between strong, equally skilled players. When both players are weak, either side is liable to make a mistake, and that can nullify the previous advantage of moving first. Also, aside from who moves first, none of these games have any chance factors. So, to sum up, the three classics score very highly on all the qualities listed above.

Given this, it is appropriate to recognize these games as examples of good game design. Using these games as models is a useful shortcut to creating a good variant. I have frequently done this myself, and so have many others. The advantage to this is that it can help you avoid a lot of errors. Some Chess variant inventors are very critical of the classics. Without proper regard for what makes the classics good, they try to create something original, and this usually results in bad games. In fact, I have yet to see any outspoken critic of the classics turn out a good game. The classics may not be perfect games, but they are a lot more error free than what most inventors would come up with by striking out on their own without a sound understanding of good game design. If you start out with something that already works and just change it slightly, odds are that you will have a good game -- even if your modifications are all for the worse. But this is not to say that using the classics as models will always guarantee a good game. Even if you start out with one of them, your changes could muck things up so much that you end up with a terrible game. You improve your chances of creating a good game by understanding good game design. The classics are useful not just as models to copy but as examples that illustrate the various goals and principles of good game design. When you understand these, it will be easier to create good games that are more original than the usual lot of obvious derivatives. This is not to discount derivatives. Some of my personal favorites are derivatives of classics, such as Vernon Parton's Alice Chess, Peter Aronson's Anti-King Chess II, or my own and Roberto Lavieri's Kamikaze Mortal Shogi. But a sound understanding of good game design does open the doors to a greater variety of good Chess variants, and that is all to the good.

Why Create Chess Variants?

If the classics are so good, why create more Chess variants? Simple. Although they're good, they're not everything. Many more good Chess variants have been created, and many more can be. While some people like to devote themselves to mastering one game, some people like to sample a smorgasbord of various games. There is nothing wrong with this, and it is not exclusive to games. Some people like to listen to many kinds of music, read many kinds of books, eat many kinds of food, etc. You don't have to be dissatisfied with the classics to consider trying variants. You might simply be excited about experiencing the wide variety of possible Chess variants. The field of Chess variants is abundant and rich in various possibilities, and it is rewarding enough for multiple lifetimes. Some of my own chief reasons for creating Chess variants are exploration and creativity. I like to explore the various possible kinds of Chess variants, and I enjoy the act of creativity.

General Principles

These principle are organized around the top-level qualities listed above. They focus on various ways for attaining the qualities of a good variant. When appropriate, I have included examples from Chess, Xiang Qi, and Shogi, because these are well-known examples of good games. In some instances, I have included examples of other well-known games. I have also included various examples from my own games, because these are the games I am most familiar with.

Enjoyment

Include elements for building and releasing tension.

A good example of this is Shogi. Its rules for dropping captured pieces create a dynamic that is not found in Chess. Tension is created by the need to keep the initiative, and it is released when the initiative is lost. This can go back and forth in Shogi, which helps make the game exciting and enjoyable. The dynamic tension found in Shogi can be created by other means. In Thunder Chess and in Clockwork Orange Chess, I found other means for creating dynamic tension.

Don't make your game too small or too large.

Small games can finish too quickly, and large games can last too long. Note that the three classics are on moderate sized boards, ranging in size from 8x8 for Chess to 9x10 for Xiang Qi. 10x10 has proven a good size for many games, though 12x12 and up might be too large. I have recently (November 2009) created a 12x12 variant called Gross Chess. To some extent, this is an experiment with a board of this size. The only pieces it adds have been tried and tested in other variants, which allows the game to be a test mainly of the increased board size. Games against Zillions of Games suggest that the game is enjoyable but the larger size makes it harder to keep track of everything. I haven't yet won a game against Zillions without taking back moves. I have played a couple games on Game Courier, drawing one game and losing the other.

Let piece power fit board size. Reserve the most powerful pieces for large boards.

In too confined a space, powerful pieces can quickly capture each other, which will nullify the point of including them in the game. On a large board, slow pieces will make a game too tedious. Chess has a good fit between piece power and board size. Xiang Qi and Shogi have weaker pieces on a larger board, but each compensates for this. Xiang Qi makes the royal piece weaker (see the next principle) and confines it to a nine square area, and Shogi let's players drop pieces, allowing them to get around more freely than their powers of movement would normally allow. Regarding my own games, I think Grand Cavalier Chess is an improvement over Cavalier Chess. Both games give players more powerful pieces than in Chess. In Cavalier Chess, a piece's freedom of movement is often constrained by all the spaces attacked by enemy pieces, but in Grand Cavalier Chess, the pieces have more space to move around in and more freedom of movement.

Let the power of the royal piece be a guide to the power of the army in general. Use more powerful armies with more powerful royal pieces, or less with less.

Cavalier Chess uses more powerful pieces with a more powerful royal piece. Xiang Qi uses weaker pieces with a weaker royal piece. When the pieces are much more powerful than the royal piece, the game can end too quickly. When the pieces are much weaker than the royal piece, the game can last too long or become too drawish.

Make offense stronger than defense.

An overly strong defense is the major cause of a game being too drawish. It can stop either player from launching a sufficiently strong attack to win the game. This will result in a drawn out and less enjoyable game. The difference between offensive and defensive capabilities is most clear cut in a game that isolates certain pieces to certain parts of the board, such as Xiang Qi. This game has specific pieces dedicated to defense. Councellors may not leave the palace, and Elephants may not cross the river. If this game had more pieces dedicated to defense, it would become very drawish. One common way to favor offense over defense is to have pieces that cannot move backwards. In Chess, Xiang Qi, and Shogi, Pawns may not move backwards. As they move forward, they gain offensive power but lose defensive power. Furthermore, none may ever protect pieces on the first rank, which limits their defensive ability. In addition to Pawns, Shogi does not let Knights and Lances move backwards, and its primary defensive pieces, the Gold and Silver Generals, are more mobile forward than backward. This compensates for the advantage that dropping pieces can give to defense. So does the inability of Pawns to protect each other in Shogi. In Xiang Qi, Pawns may protect each other after crossing the river, but it compensates for this by initially separating Pawns from each other. In Ultima, the Pawns can move backwards, and its creator, Robert Abbott, has complained that it favors defense over offense. See his article What's Wrong With Ultima? for more details.

Don't make defense too weak or offense too strong.

An enjoyable Chess variant has a good balance between offense and defense. Although a stronger offense is important for making a game decisive, defense should not be neglected. Rifle Chess seems to be a good example of a game that is too imbalanced in favor of offense. In this game, pieces capture from a distance by shooting each other. This makes defense practically negligible, because a piece can't be expected to move to the space it has just captured a piece from.

Playtest games.

Don't rely solely on the abstract application of principles to tell whether your game is enjoyable. Try it out. Play it with a friend or play it on Zillions of Games. Playtesting can help you find flaws in your game before the general public sees it. Show the public your best stuff, not untested stuff that could be junk.

Playability

Avoid complicated rules. Let any complexity to the game be the result of small, simple rules working together.

Simple rules are the best. The difficulty of a game should be in mastering it, not in learning it. As an example of how simple rules can create complexity, consider Metamorphin' Fusion Chess. The rules of this game allow for reproduction, but no rule concerns reproduction. Instead, it has simple rules for promotion and for splitting a piece apart. Reproduction happens when a promoted piece splits into its components.

Don't overcomplicate your game.

The rules of Ultima may not be overcomplicated, but their application has become overcomplicated. Robert Abbott, Ultima's inventor, has complained that the game lacks clarity. (See What is wrong with Ultima?). Clarity is a measure of how easily players can look ahead and foresee potential outcomes. Ultima lacks clarity, because some pieces have powers of capture that depend not only on themselves but on other pieces. The Ultima Pawn captures with assistance from other Ultima Pawns. The Coordinator captures by coordinating with the King. Worst of all, the Chameleon captures another piece by imitating its powers of capture, and it may use multiple means to capture several pieces at once.

Avoid arbitrary rules.

Arbitrary rules can be harder to remember, because they don't make much sense and don't fit well with the other rules.

Avoid kludgy rules when you can.

Chess is not without kludgy rules. A kludgy rule is a rule that inelegantly fixes a problem with the game so that is better. In Chess, the rules concerning double moves and en passant are kludgy rules. So kludgy rules aren't to be avoided all the time. But it is generally better to design your game so that kludgy rules will not be needed. In an early version of Storm the Ivory Tower, a hybrid of Smess and Xiang Qi, some pieces could get trapped in the Ivory Tower. I fixed this with a kludgy rule that certain pieces moved against the arrows in their own towers. This didn't fit well with the idea that pieces are supposed to move in the directions given by arrows, but it at least made the game less drawish. I eventually came up with a more elegant solution and redesigned the game.

Favor pieces whose moves are easy to visualize, such as Knights and line riders.

An early version of Storm the Ivory Tower included a piece that made a two-step move according to the directions given on its square and the square it passed over. Although based on the Chinese Chess Horse, it could sometimes move to spaces a Horse would not be able to reach, and which spaces these were would depend on the arrows on its square and the surrounding squares. In a new version, I simplified the piece by making it move as a Horse or a Moa, beginning in a direction given by an arrow on its square, but being unaffected by arrows on neighboring squares. This simplification of the piece has improved the game.

Don't litter your game with many new and unfamiliar pieces.

This will make your game harder to learn, because players won't be able to fall back on what they know from other games, and they will have more to remember. Note that it is easier to remember new pieces when they are similar to the pieces they replace. For example, Storm the Ivory Tower introduces several new pieces, but they are all Smess adaptations of Xiang Qi pieces, and Cavalier Chess introduces several new pieces that are more powerful counterparts of the usual Chess pieces. Furthermore, commonly used fairy pieces will be more familiar to many players than others. For example, all the pieces introduced in Gross Chess are commonly found in other games. The main thing to avoid is the introduction of several entirely novel pieces.

Interest

Use slight left/right asymmetry to distinguish all opening moves from each other.

Although it is good to retain left/right symmetry between most pieces, a couple different pieces on each side is good for distinguishing moves on each side from each other. Chess does this by having the King and Queen on different sides, and Shogi does it by having the Rook and Bishop on different sides. Most Chess variants, usually in imitation of Chess, have done something like this.

Pawn structure is the soul of Chess. Do not get rid of Pawns unless you can effectively compensate for their loss.

Philidor, a great Chess player of antiquity, said, "Pawns are the soul of Chess." Indeed, much of Chess involves hiding behind and moving around Pawns. Pawns serve as the shields for the more powerful pieces, and working around them makes the game more interesting. When you remove Pawns and don't replace them with something equivalent, your game can devolve into a senseless slugfest between powerful pieces. Among my own games, Cavalier Chess replaced Pawns with Cavaliers, which provided their own kind of structure, and Interdependent Chess replaced Pawns with Guardians and Stewards, and without any powerful riders, such as Rooks and Bishops, the Pawns weren't missed as much.

Include pieces with differing powers of movement. Each can attack the other without being attacked back.

This is one of the reasons I like Chess better than Checkers. The variety of pieces in Chess allows for a variety of tactics, such as forking, skewering, and pinning. It is hard for a piece to fork, pin, or skewer a piece that moves just as itself. This is why the pieces should represent a variety of movement types. With its Cannon piece, Xiang Qi includes a tactic not found in Chess or Shogi, the double pin. The Cannon attacks pieces behind a screen that it must hop over to make a capture. When two pieces stand between a Cannon and the General or some other valuable piece, both pieces are pinned, because moving either one away would leave the other piece behind as a screen for the Cannon to hop over.

Include both weak and strong pieces, including both major and minor pieces.

Doing this will make exchanges more interesting, and weak pieces sometimes have important jobs to do. One of the great things about Pawns is that your opponent is unlikely to capture them with more valuable pieces as long as they're protected. This makes Pawns better for defending pieces than more valuable pieces. This is one good reason for including both weak and strong pieces. Another is that it gives a player more to consider when comtemplating an exchange. And it gives players the opportunity for sacrifices and gambits. When pieces are more or less equal in value, exchanges are not quite as significant.

Avoid any piece that can force checkmate by itself, such as the Amazon or the Cavalry Chess Knight.

The Chess army should work as a team, not sit on the sidelines while one star piece wins the game on its own. Too powerful pieces throw a game off balance. When I designed Cavalier Chess, I initially replaced the Queen with an Amazon, which moves as a Knight or a Queen. This piece proved too powerful. If a player made a mistake in defending against the Amazon, it could proceed to win the game on its own. The same goes for the Knight in Frank Maus's Cavalry Chess, which moves as a Knight, Camel, or Zebra. This piece could easily put the King on the run and checkmate it before long, even though the Cavalry Chess King was more powerful than the usual King, and it was surrounded by many friendly pieces. Any piece that can easily win the game on its own should be excluded from a game. Chess is a team game. Although played by individuals, each individual controls a team, and it should be mastery of the team, not mastery of one powerful piece, that makes someone good at a Chess variant. The most powerful piece in a Chess variant should generally be one that can't, or at least has difficulty, forcing checkmate by itself even when all defenders are gone, but which can force checkmate with the help of one's own royal piece (usually a King). The Queen is such a piece in Chess. It cannot force checkmate without assistance from another piece. The strongest pieces in Xiang Qi and Shogi are even weaker than a Queen. Capablanca's Chess and related games, including several of my own from Cavalier Chess to Gross Chess, include a piece that can checkmate a King on its own. This piece is a Knight/Bishop compound. It can checkmate a King that happens to move to a corner, but it is unable to force a King to go where it can checkmate it. Thus, it is unable to force checkmate. This is a subtle but important difference. It is alright to include a piece that can checkmate a King by itself so long as it is still unable to force checkmate by itself.

When a game has special rules or terrain, include a supernumerary piece that can take advantage of this.

This will make the game more interesting by offering something that just won't be found in other games. Examples include the Voidrider in Voidrider Chess and the Universalist in Interdependent Chess.

Choose pieces that work best with special rules or terrain.

Wormhole Chess and Chesire Cat Chess are based on the same idea, but Wormhole Chess plays to this idea by using all leapers, making it a more interesting game.

Fairness

Avoid weaknesses, such as an unprotected Pawn, in the opening array.

Any structural weakness in the array could magnify White's opening advantage, since White would have the first opportunity to exploit it. Capablanca's Chess has this problem, but some variations on it, including my own Grotesque Chess, have fixed this.

Make the Black and White armies the same, positioning them symmetrically to each other. This makes the game balanced.

Three main types of symmetry are possible. Chess has mirror symmetry, which means that one side looks like a mirror image of the other. Shogi has rotational symmetry, which means that each side looks like the other when the board is rotated 180 degrees. Xiang Qi has perfect symmetry, which is simply a combination of mirror symmetry and rotational symmetry. Both mirror symmetry and rotational symmetry are important, but it is not so important to have both completely. It should be sufficient to have one fully and the other partially. Except for the d and e files, Chess has rotational symmetry, and except for the 2 and 8 files, Shogi has mirror symmetry. To give figures, Chess has 100% mirror symmetry with 75% rotational symmetry, and Shogi has 100% rotational symmetry with approximately 78% mirror symmetry. The fairness of these games seems unaffected by their lack of perfect symmetry.

If you intentionally design an unequal armies variant, work hard at making them equal in power.

I won a game of Chess with Different Armies because the army I played, the Nutty Knights, had an advantage over the other army, the Fabulous FIDEs. After a series of exchanges of my pieces for their FIDE counterparts, I got it down to a Charging Knight and a couple Pawns on my side verses a Bishop and a couple Pawns on the other. The Charging Knight was supposed to be roughly equivalent to a Bishop, but since it wasn't colorbound and was technically a major piece, though a relatively weak one, I ended up with the advantage.

Include elements that reduce the importance of material advantage.

Shogi has an advantage in fairness over Chess and Xiang Qi. The ability to drop pieces makes the game more dynamic. You sometimes have to sacrifice many pieces to keep the initiative, and your opponent may then have to do the same. This tips the balance of power back and forth, which helps to reduce the advantage of presently having more material. In contrast to this, Chess can easily reach a point where the difference of a Pawn will determine the outcome of a game.

If you do create a game with a chance element, do what you can to still make the outcome depend on skill.

There are various games with elements of chance that are still largely matters of skill. These include various card games, such as Poker, and various board games with dice or spinners, such as Monopoly, Risk, or Snit's Revenge. It is sometimes worthwhile to add a chance element, because it can make a game more dynamic and exciting. One really good Chess variant with a chance element is Knightmare Chess, which allows players to play cards that change rules, pieces, terrain, and other elements of the game. Despite the chance element of the game, it still takes judgement to decide which cards to play. A good player knows when and how to use cards to best effect.

In some circumstances, chance elements can make a variant more fair.

Although games with chance elements can reduce fairness between equally matched players, they can put unequal players on a more equal footing. For example, if an experienced Chess player plays a randomized variant, such as Heraldic Chess or Vegas Fun Chess, with an inexperienced child, the child has a better chance of winning than he would have at Chess. My experience with playing Knightmare Chess several times is that the advantage shifts back and forth a lot, making small advantages less critical, and making the ability to take advantage of opportunities more important.

Written by Fergus Duniho.
WWW page created: December 10, 2005.