One of the many great attractions of the game is the large number of appealing variant rules one can play. Many of the variants can compete with the strategic and tactical richness of chess, and some even surpass it. I believe there are two primary reasons why none of these variants has caught on as a serious "second version" of chess. First of all, for that small portion of the population that is already attracted to the game, chess is so rich that it already provides more than enough fascination for a lifetime. Second, there are so many reasonable kinds of variants that no single one of them has caught on as an official "alternative version." This latter explanation is a plausible theory suggested to me by IM Larry Kaufman.

Chess players should remember that the rules have not always been the way they are. Some of the more recent of the "standard rules" include, unsurprisingly, the "artificial" castling and en passant rules, and also the incredible power of the queen. All these changes helped to speed up the action in the game and to make it more lively. Chess must have evolved considerably since its early days, when it was a gambling game of some sort, involving dice, and as such was actually forbidden by Arab religious law. We should also remember that three of the greatest world champions - Capablanca, Fischer, and Kasparov - have been responsible for some of the most significant ideas for changing the game. Some have suggested that after one masters the game, one wants to "destroy" it, but perhaps it is more accurate to say: if there is no more ambition to increase in strength to defeat stronger players, one might more quickly turn one's attention to contemplating the game as a whole.

It seems to me that the best motivations for changing the rules are just fun and variety. How, then, might one vary the game of chess? We can make it "bigger" by changing the size or shape of the board, and possibly adding new pieces. Keeping the traditional board, we could modify some of the movements of the pieces or some of the rules under which they operate, or change the objective of the game. We could change or randomize the setup position, or disallow castling. We could change the number of players. We can change the definition of a move. Finally, we can play the same game under the same rules, and change the circumstances under which it is played.

Table of Chess Variants

Unusual conditions of play

Time odds

One of the motivations for chess "variants" is to create an imbalanced game for unequally matched players. One way to do this is simply to give one player more time for his moves than the other. It is common to find hustlers on the street taking on all comers at the odds of, say, 10 minutes to 5, or 5 minutes to 2. (I suspect these odds would generally account for a difference of about 150 and 300 rating points, respectively. It's surprising that serious statistical research has never been done on this kind of thing.) With digital clocks and all their features, the possibilities here are endless: e.g., one player gets time added or a delay while the other doesn't.

Back to the top

Blindfold chess

Another ancient "handicap" is to have the stronger player play without sight of the board. This is a mental feat that never ceases to amaze people, and indeed, not without reason. To handle just one game well in this way takes a lot of concentration even for an expert. The really remarkable feats are players such as Alekhine, Koltanowski and Najdorf, who could manage simultaneous blindfold exhibitions of 35 to 45 opponents, winning the bulk of the games. Of course, with equally matched opponents it is always possible for both to play blindfolded at once; this is a good way to pass time at train stations and on bus rides. Unfortunately, our only opportunity to witness today's strongest masters in blindfold competition is the Melody Amber tournament, but we can be glad for that. It also has its share of entertaining bloopers.

Back to the top

Kriegspiel

Back to the top

Advanced chess

Gary Kasparov, inspired by his encounters with the topflight chess programs, came up with this idea of having two opponents playing a relatively quicker kind of game but with the help of a computer program. He imagined that the level of play would be extraordinarily high, and that spectators would appreciate it because they could follow a player's thought process, and because the game wouldn't drag on interminably. He did have a point, and Advanced Chess, though very new, seems to be gaining in popularity. However, it seems to me that this sort of game takes away the element of calculation from the human player, and this is practically the essence of the game. Also, there is no reason to memorize any openings, but a lot of reason to prepare them deeply. I'm not sure I like that either. I'm not a fan of Advanced Chess, but it does have its place.

Back to the top

Correspondence chess

Postal chess is an ancient way to play the game: send a postcard in the mail with a move, get one back a couple days later. There are even time limits here, of sorts: usually two or three days from receipt to return postmark. Obviously the desirable feature of this format is that depth and precision of analysis will be rewarded, rather than speed or alertness or the setting of crude traps. I find it very valuable as a way to improve one's understanding of the game, because it forces a player to use the analytical method in a rigorous way, and that is exactly what we need to do over the board. It is also a useful format for personal opening analysis and for mastering theory.

Time and technology have dramatically changed the nature of correspondence chess. On the positive side, it is now possible to play by email, which eliminates the dependence on the postal system. However, the negative side is probably far more important: there is no way to know you aren't playing against someone's computer. Of course, for people who do this, it's their loss. However, it's a frustrating prospect for an honest player, and I fear that it renders postal chess pretty much obsolete as a competitive game. Maybe the next thing will be "advanced" email chess, in which two players play a correspondence game with the help of computer programs. Now THAT should be a high level of play, and maybe not such a bad idea either!

Back to the top

Changes to the initial position

Odds games

One of the oldest, and possibly most obvious kinds of ways to change the opening position also solves the problem of "equalizing" the game for unmatched opponents: material odds, or x amount of material plus y free moves at the start. It is interesting to speculate what the value of the extra material could be in rating terms. I think an extra pawn is worth 150 points or so; pawn and move 200. A knight, at least 600 points, but this might depend on which 600! Anyway, there is a strong collection of ancient offhand games, played with these kinds of odds.

Back to the top

Shuffle chess

Changing the initial piece configuration is one of the most logical ways to keep the character of the game while eliminating a possibly undesirable characteristic: the explosion of opening theory. A popular version is "shuffle chess", in which the set of eight pieces is aligned randomly on the back rank, and the black set of pieces will mirror the white one. One of the many miracles of modern chess clocks is that they can be programmed to generate shuffle chess positions according to vartious restrictions. Various restrictions of the alignments are possible, for example that the bishops must start on different colored squares, or that the king must stand somewhere between the two rooks. Rules of castling can be designated according to the variant chosen.

Back to the top

Shuffle chess, and FischeRandom

With both the above restrictions (which I like) and castling to the traditional locations, we have the game "FischeRandom." It is an excellent alternate version of chess, in my opinion, in that it keeps the chess thought-processes alive, but puts both players on their own devices from the very start. My only doubt about this game is that the castling rules seem to be really artificial. Why not castle according to the positions of the pieces? Other kinds of shuffle chess exist, too. For example, the king plus a set of seven random other pieces are placed on the back rank. This allows for bizarre combinations, like three queens or three knights, but also allows for some dull situations like four bishops. Another game, which I once thought of, involved setting up the pawns first. (It turns out this game is called "Pre-Chess" and was written up in Chess Life, November 1978, by Pal Benko and Burt Hochberg.) Then the clock was started and White's first move was to place his king, and Black would follow suit. Then, the queen, the rooks, and the four minors. The advantage of this game is that Black can choose to imbalance the game from the very outset. The disadvantage is that the "good" structures may be fairly well defined. Nevertheless, I find all these kinds of games to be healthy variety. (A recent game invented along these lines is Jim Callan's Go-chess which involves placing all the pieces first on a large and non-conventional board, before commencing "chess" play.)

Back to the top

Comprehensive rule changes

A given piece must mate

One of the strangest kinds of odds to give is to announce from the outset the piece that will deliver checkmate. Thus, for the opponent to win the game it would suffice to capture that one piece. For this kind of game, the magnitude of the odds depends on what piece is designated. If it is a rook, there shouldn't be much of a problem. If it's the a-pawn, it will be difficult enough to prevent the opponent from surrounding it! This is a fun sort of odds game but can't be taken too seriously as a form of competition.

Back to the top

Stalemate wins

A popular suggestion to change the game and reduce the number of draws is to consider stalemate a loss for the side that can't move. Many times people have written in to "Larry Evans on Chess" to complain about the number of draws in grandmaster play and to suggest this as an alternative. If stalemate were a win then all the endgame theory we ever learned could be thrown out the window. It is true that draws would diminish substantially, since any protectable extra minor piece or pawn would win by force. No matter how many times it is suggested, this rule change never has caught on; it isn't offered as a version on the internet chess servers I know, and I hope it never will be.

I find this rule change suggestion completely ludicrous for many reasons. First of all, the endgame theory would be quickly rewritten, but obviously by computers. Is this what we want to promote? Usually it is the same people who suggest this sort of change who also complain about the computers. Secondly, the proportion of draws at GM play is not really all that high (about 50% if I recall) and many of these draws are actually quite interesting. This rule wouldn't necessarily change the situation with those other kinds: draws by agreement, boredom or mutual uncertainty. Finally, from an aesthetic standpoint, I find stalemate - and those situations where an extra piece can't win - to be a "charming irrationality" that is healthy for a logical game like chess. So many remarkable charming endgame compositions have been created using this theme.

Back to the top

Suicide chess

A completely different "animal" is suicide chess. This is a fascinating and worthy game on its own right. It is played by "checkers" rules: if one can take something, one has to do it, but may choose any of the available capturing moves. The objective is to lose all your pieces. There are beautiful situations in suicide endgames in which one side gradually forces the opponent to take the remaining pieces. It is a comical thing when a piece reluctantly invades the undeveloped enemy camp, and is gradually forced to capture reluctantly, one piece after another, until suddenly it manages to sacrifice itself or to lure a powerful enemy piece into its own camp.

I respect suicide as an independent game, but I have to admit I'm not a great fan of it, for a number of reasons. In general, suicide seems to exaggerate what may be the "less desirable features" of chess. First of all, the game is extraordinarily tactical in nature; calculation, which already dominates sophisticated chess thinking, is almost all there is to suicide. There is strategy, to be sure, but let us say that the "scale of catastrophe" is much greater. Overlooking a tiny detail in one distant branch of one's search tree will result in a titanic blunder that will immediately damage one's game beyond repair. Since, beyond a certain point, under time constraints, it is impossible to see everything, a certain "luck" can be said to play a role. In fact, it seems that suicide games are determined mostly by blunders, and I don't care for this aspect of chess. The loser gets upset and the winner doesn't feel any sense of accomplishment either. Games like this are the stomping field of the computer. Another example of how this characteristic plays a negative role in suicide is opening theory. A knowledgable player will crush the neophyte just by reciting opening theory. The majority of opening moves lose almost trivially by force. I don't like this sort of thing. Therefore, I doubt that suicide chess could ever find a serious following.

Back to the top

Neanderthal chess (my coinage)

Speaking of promotions, and zany things, there is even a version of chess which starts from the "opening position," except upside down! As long as neither player succumbs to the famous, crude smothered mate trap, there will soon be many queens on the board. Again, many possible tactical scenarios but not too much variety with all the same pieces. I don't care for this game, frankly, even as a diversion. Also, there is so-called "power chess," in which each side has a king, 8 pawns, and 7 queens. Basically this is just a way to take all the beauty and variety out of the game all at once.

Back to the top

More complicated versions of the game

Progressive chess variants

From the zany side of things, we have games in which more than one move is allowed at a time by the same player. For example, White moves once, then Black twice, then White three times, etc. In general in this game, one is allowed only one of those moves to extricate himself from checkmate. It is an ancient question, whose side of this game is to be preferred! It is also possible to play a simple game with two moves at a time from each side. These games are harmless little diversions but can't be taken seriously. They are rather "unnatural" and as in suicide, they are almost entirely games of nitpicky tactical detail. Thirty or so moves at a time from the opponent are quite enough to consider, aren't they? A curious version of these "multi-move games" is the one where White has a king and four pawns against Black's entire position, but White gets to move twice at a time (and yes, a king on e5 may take a king on f7, walking through check and all). I believe this is the official version of "Échecs marseillais." It is very treacherous to defend the Black position, so dangerous is this extra move. However, I think Black is winning in this game without much trouble, and there is another version in which White has eight pawns. Here I don't know the result, but I think I like White, thanks to promotion. Get me one queen, and it's all over!

Back to the top

Grand Chess

In the 1930's Capablanca pronounced the game of chess to be played out, and concluded that the board needed to be increased and new pieces added. I think now we can and should laugh at his arrogance, and I predict and hope that in another 50 years we will laugh at people who are so worried about it today. Nevertheless, for those of you possessing Capablanca's natural talent and self-assurance, his game is worth a try. The board is 10 by 10 and new pieces are added, the marshall and the cardinal, placed to the right of the kings. Since the chess queen already combines the features of rook and bishop, it seems logical to construct other fearsome pieces that combine the features of rook and knight, and bishop and knight, respectively. The pawns are set up on the 3rd and 8th ranks, the bulk of the pieces get the 2nd and 9th ranks, and the rooks are placed on the corners, with more freedom of movement than usual. I have my usual objections: it makes the game too complicated and tactical, and we humans won't have a chance against a computer. Nevertheless, this form of chess has its particular charms and was written up by Burt Hochberg in Chess Life, in August of 1997. The game was refined by Christian Freeling, who found an excellent way for rooks to achieve a mobility they are often denied in regular chess; click here for an excellent source of information on this game.

Back to the top

Crazyhouse

Certainly this is by far the best of the games in this rather dubious section. It is a special type of bughouse game, played one against one. Whenever you take off an opponent's piece, you "convert" it to your own type of piece, and hold it as a possession. On any future move, instead of playing a normal chess move, you may place that piece on the board instead. Obviously there is nonstop activity and games are played with constant mating attacks. Although I'm certainly better equipped to play chess, I am a fan of the crazyhouse concept. It takes away the hypercritical microsecondal timing importance away from bughouse, and also forced people to be careful when they give up material, because they have to give it away "double." It is possible and worthwhile to analyze crazyhouse games, and to play them at a serious time control.

Back to the top

Team games

3- or 4-way chess

I understand that there is a new kind of chess game which is played on a Chinese checkers type board. In my opinion, the problem with these kinds of games - where there are multiple players and an individual winner but where everyone interacts with everyone else's gamepieces - is that the game is much a political contest as a test of skill.

Back to the top

Consultation and Alternate move games

These are friendly ways to consolidate a number of players onto one board. In the consultation game, the "allies" will discuss their moves before playing them. This can be done within earshot of the opponent, or not, as desired. Again, this is a very old form of competition, and old game anthologies are full of examples of great players playing against teams of players. Occasionally, even groups of great players play against one another. Naturally, in consultation games, assuming reasonably equal strength in the participants, the playing level should be somewhat higher than either one player. Maybe if they are both equal, the ability to consult should add 50 to 100 points.

On the other hand, in the "alternate move" format, each player has his own thoughts about the game, which the partner might not follow. Therefore the strength of an equally matched alternate-move team must be at least 100 points lower than their individual rating. It is always a risk for one player to choose a line of play requiring risk, or a precisely planned followup of any kind! And there are many entertaining stories from these kinds of games, especially when the teammates' styles aren't well matched. One player will impetuously thrust with his knight, and on the next move his cautious partner will put it back "where it belongs." These kinds of games are a lot of fun, on the "exhibition" or "casual" level, but it's hard to imagine them forming any kind of serious competition.

Back to the top

Bughouse

I have saved one of the most popular and compelling chess variants for last. It is a rare player who has tried bughouse and not become addicted to it at some point. For the uninitiated, bughouse is a two-board team chess game. When a player takes off his opponent's piece, he gives it to his partner (who is playing that color) and the partner may then place it on the board instead of making a traditional move. Ferocious sacrificial attacks, king hunts, stalling, partner communication and arguments, and pieces flying in the air are all commonplace.

There is - unfortunately - a very serious flaw in bughouse, in my opinion. That is the critical dependence on every microsecond of moving time. Why? In many situations player X1 simply needs to capture one piece for his partner X2, for X2 to wrap up the game. Suppose X1's opponent Y1 realizes that that piece is surrounded, and in one more move he must lose it. Then Y1's only recourse is to stall. If he has more time on his clock than X2, then eventually X2 will have to move normally, giving Y2 an opportunity to correct the situation. This scenario is extremely common in bughouse; it also happens with impending mates. (Obviously a digital clock is a necessity to play this game at all meaningfully.) Thus, the critical time relationships in bughouse are generally between a player and his partner's opponent (partopp). Clearly, if one assumes properly set clocks (and no internet "lag"), TimeX1-TimeY2 will always equal TimeX2-TimeY1. Either of these quantities can be used to determine which team is ahead on time. Whoever is ahead, even by as little as a second, completely controls the flow of material. Experience shows that it takes a substantial advantage on the board to overcome a time deficit. Therefore bughouse ends up being a real circus. Of course, that is the point - to be a circus - and it really is a blast if you're in the right mood, but it can quickly become annoying that the game is substantially a contest is reflexes and agility, and that fast moves are on the whole often better than good ones.

Back to the top

Back to chess page