[Next] [Up] [Previous]

Some Comments on Chess

Here, I have made two comments on matters relating to the game of Chess, and now add three additional ones.

Styles of Chessmen

On the right of this page, a diagram illustrates the Staunton pattern of chessmen, along with a few others.

Above them is the French Régence pattern, which, to those of us used to the Staunton pattern, are hard to tell apart.

Below them are three patterns used in England, the St. George pattern, the Barleycorn pattern, and the Edinburgh Upright pattern.

There were variations within these patterns; not all sets had the luxury of a horse's head for the Knight; some English pattern sets had a St. George's cross atop the King just as is found in the Staunton pattern.

The Edinburgh Upright pattern, also known as the Northern Upright pattern, is sometimes noted (along with the Dublin pattern, not shown) as a predecessor of the Staunton pattern. While that is true to a limited extent, since it shares with the Staunton pattern having an unadorned, and hence less distracting, stem, as is clear from the diagram, this pattern, which did precede the Staunton pattern, cannot be used as a basis for failing to fully recognize the originality of the Staunton pattern.

The distinguishing features by which the different Chessmen were told apart one from another were basically identical between the St. George pattern, the much criticized Barleycorn pattern, and the Northern Upright pattern. And since the great strength of the Staunton pattern is that the different Chessmen are easily distinguished in it, it really doesn't owe all that much to the Northern Upright pattern.

However, the Barleycorn and St. George patterns as well as the Northern Upright pattern are an improvement on the Régence pattern, since in them the Bishop is now easy to distinguish; and then the Staunton pattern is an improvement on all of those three, equally, including the (rare and valuable) Northern Upright pattern, by making the Queen easy to distinguish as well. Still, in fairness, making the stems unadorned was taking an additional, if small, step in the direction of the Staunton pattern.

Explaining the Rules of Chess

When I was first instructed in the rules of Chess at the age of 10, one thing about them was confusing. Since it was so strongly emphasized that the King cannot be captured, I didn't think there was any problem with moving the King into check!

One modern textbook of Chess today has gone to the other extreme, and flatly stated that the objective of Chess is to capture the opponent's King. Of course, that isn't true. But if a qualification were added to that statement, it could be changed to match the truth, and be a more understandable form of the truth.

Here is how I would explain the objective of Chess:

The move where the King is captured does not actually occur in a chess game, since it would happen two moves (one for each player) after the move that decides the winner. Although this means that the capture of the King is not actually played as a move, it is still effectively the object of the game. However, this has one important qualification. A position in which the next move cannot prevent the King being captured on the move after can be a win, or it can be a draw, depending on whether the position is checkmate or stalemate. For checkmate, the King also has to be in check at the start of the move; that is, if neither the King nor any other piece of the losing player moved on his turn, the King could still be captured on the next move.

In practice, the stalemate rule is intended to make Chess a more demanding game of skill, requiring care to win even by the player who has developed a lead. But the rule can be made easier to remember by thinking of it in symbolic terms. If Chess is a game of war, then when victory is inevitable, if the enemy King is not under the watchful eye of one of your soldiers, perhaps by committing suicide, or by destroying papers, or a symbol of his office, he can still cheat you of your victory.

If we keep firmly in mind that the object is still, in effect, to capture the King, then it is clear that a move into check, or a move leaving the King in check, would lose the game; then it can be added that such moves are forbidden in order that a game will not be decided by a single lapse.

Another somewhat confusing part of Chess are its special rules, the two-space first move of Pawns and the associated possibility of en passant capture, and castling, and the rule that the King cannot, when castling, move across a square that is attacked.

A famous Grandmaster (one from the Soviet Union who wrote a book on endgames) once forgot that the Rook could move across a square that is attacked during the castling move.

But there is a way to make this particular rule easier to remember as well.

There are only two kinds of chess pieces that move only one step at a time. One is the lowly Pawn, the other the King. Both are allowed to move two steps for their first move, but in different ways.

A pawn can move two steps forwards for its first move. But this is intended only to allow it to move faster, not to skip past a point where it may be captured. Thus, if, when it moved two steps forwards, it could have been captured by an enemy Pawn had it made a regular one-step move, then that Pawn can still capture it by advancing diagonally into the square it would have been in had it made that regular one-step move.

Castling is also always a two-step move for the King, and it must be the King's first move. Thus, the reason why the King cannot move across a threatened square is now obvious: doing so, it would be moving into check. Because it would be liable to the form of en passant capture applicable to Kings; since a King can move in any direction, not just forwards, it can be taken en passant by any piece, not just by a Pawn or another King.

Some History of the Special Rules

Today, the rules of Chess are well-known and well-defined, as they are set by FIDE (the Fédération Internationale des Échecs) and also by the chess federations of individual countries, but in conformity with those set internationally.

At one time, the situation was much less tidy.

Thus, in 1843, Jaenisch, in his Analyse Nouvelle des Ouvertures du Jeu des Échecs notes the following about three aspects of the rules of Chess:

En Passant Capture

Other sources note that at about the same time as the moves of the Queen and Bishop changed, the Pawn gained the power of moving two steps on its first move as well.

According to Jaenisch, there was, at least for a short time, some question about whether or not the Pawn should be subject to en passant capture, and, as well, one possibility that was considered was that any piece that might capture the Pawn, had it advanced by only one square, should be allowed to capture that Pawn en passant.

The modern rule was supported by most amateur Chess players, and by some noted players in Spain and Portugal, he says, and this led to it being adopted in his time in every country but one: Italy, in which the Pawn may advance two squares on its first move, and is not subject to en passant capture.


Other sources note that the older version of Castling, where the King is allowed to jump two squares as his first move, so that the same position as is obtained through Castling can be obtained by first moving the Rook next to the King, and then, on the next or a later move, jumping over the Rook with the King, existed prior to the changes to the movement of the Bishop and the Queen. This special move of the King was called the King's Leap.

Here, Jaenisch notes that again it is Italy that is different. The King's Leap was gone in his day, but in Italy, free Castling is practiced, in which the King may leap over the Rook to any square beyond, up to the edge of the board. This is noted as being first described by Alessandro Salvio. Greco advocated the form of Castling we use today, identifying it as the Calabrese style of Castling.

Pawn Promotion

Here, the rule we are familiar with today, that a Pawn must be promoted immediately upon reaching the back rank, and it may be promoted to a Queen, Rook, Knight or Bishop (of the same color!) at the player's option, no matter how many of these pieces are on the board, was the general practice only in Britain and France: and even there, only in the serious Chess clubs.

Elsewhere, both in other countries and in most people's casual play even in Britain and France themselves, the rule that a Pawn can only promote to a piece that had already been captured, with the Pawn sitting idly on the last rank waiting to promote if its owner had not suffered any captured, was followed. This is referred to by Jaenisch as the German rule for promotion.

I know that even in my own youth I personally encountered people playing Chess by this alternate rule for Pawn promotion, and so I can attest that the current official rule for Pawn promotion took a long time to become well established and generally known.

Also, at one point, Staunton drafted a set of rules for Chess in which the obvious qualification that one can't promote a Pawn of one's own to a piece of the opponent's color was omitted - which prompted several chess problem composers to compose problems in which doing exactly that was required to win! The obvious way this could be the case is where it seems that the position must inevitably lead to a stalemate.

Changing the Game of Chess

Three problems with Chess have often been noted as interfering with its popularity, particularly in the sense of its popularity as a spectator sport. There is a fourth problem, less often discussed, but more fundamental, that I don't think anything can solve: one can enjoy a game of football without being as strong as a quarterback, just as one can enjoy ballet or gymnastics without being capable of the feats one is watching. But to enjoy watching a game of chess, one has to be able to understand what is going on, what threats and possibilities are present in the chessboard as it stands before the players. The fact that one doesn't have to be quite as smart as the players themselves to enjoy the game, since of course the point of the grandmaster's move is often only obvious after he makes it, as with many discoveries and inventions, or the deductions of Sherlock Holmes, ameliorates this, but only partly.

The third problem, although often recognized as a problem, is also difficult to solve; it is closely related to the second problem noted below, and also somewhat related to the fourth problem noted above.

That is the increasing prevalence, ever since the Modern era of Chess was inaugurated by Steinitz, of defensive and positional play, in contrast to the exciting tactical fireworks found in games like La Bourdonnais - McDonnell or Anderssen - Kieseritzky. Solutions to the first and second problems noted below are often hoped to address this issue at least partially,

However, I categorize this as a difficult problem to solve because it is fundamentally the result of a deeper and more profound understanding of the game being achieved. One cannot expect players, even with an awareness of the larger problems affecting the game in general, to which an individual game may slightly contribute, to play in a way other than that which, to the best of their knowledge and understanding, is best suited to produce the most favorable likely result of the game they are playing.

But the two problems that people keep trying to solve are these:

Among the solutions proposed for these problems are making the board bigger by one or two squares, either just horizontally or in both directions, so that some new pieces can be added, or having the arrangement of pieces on the first row selected at random for each game.

But while such solutions have led to the invention of many interesting variations of Chess, they have not caught on. Replacing Chess with another, although similar, game is just too drastic.

Later on, in the section on Random Variant Chess, I suggest my own variation of Chess to address the first problem. Rather than spoiling the symmetry of the chessboard by arranging the pieces randomly, or expecting one enlarged version of chess to do more than temporarily delay the emergence of a new body of opening theory, I take a modestly enlarged chess board, and by selecting chess pieces from a set of pieces with different powers, I create a family of thousands of variants of Chess, one of which is chosen at random for a pair of games between two players. The mechanics of this are envisaged to be similar to the choice of a predetermined opening at random in Checkers through the three-move restriction or the earlier two-move restriction.

Of course, an enlarged chess board, and a set of many possible pieces, involves considerable additional equipment. Perhaps another possiblity would be to randomize Chess slightly. In the early days of Chess, players rolled a die each turn to decide which piece they would be allowed to move. No one would want to return to that. But perhaps players could roll a die, and it would give them the option of making one special move in addition to the regular moves. For example:

1) Pawn may capture by moving one square forwards.
2) Bishop may move one space orthogonally.
3) Knight may move one space diagonally.
4) Rook may move one space diagonally.
5) Queen may make a Knight's move.
6) King may make a Knight's move.

Presumably, the option of making such moves, coming at unpredictable times, would be enough to prevent the use of fixed openings. By advancing the die along the side of the board, one could limit its use to the first eight moves of the game. Rules limiting the use of a bonus move to give check or to escape from mate would also be needed.

Another possibility would be to roll the die once every three moves (two moves for one player, with one move for the other player in between). That would balance out the advance knowledge of what special move is available, and yet reduce the frequency of dice rolls. Again, advancing the die between uses against the edge of the chessboard would limit its use to the first 24 moves, twelve for each player.

In that case, a fair rule for the use of the special move during its lifetime might be:

In the first turn, the special move is novel, and too much would hinge on chance; in the second turn, the move is known, and part of the game; in the third turn, the defender will not have the use of the special move in the next turn to escape from a check given in that turn. But such a rule, doubtless, is just too complicated.

I would like to suggest a solution to the second of these problems which involves, in my opinion, only a very slight change to Chess.

I propose that, in addition to scoring a Chess competition so that, when a game is won, the winner recieves 1 point, and the loser recieves 0 points, and when a game is drawn, each player recieves 1/2 of a point, that a player who can force stalemate should recieve 3/5 of a point, with the other player recieving only 2/5 of a point.

In this way, a result that gives a benefit to one player can be obtained even when play has led to a smaller advantage to one player than can lead to checkmate. Yet, that result is significantly less valuable than that given by checkmate:

Checkmate      1      0       1
Stalemate      0.6    0.4     0.2
Draw           0.5    0.5     0

a checkmate is worth five stalemates by that rule.

Thus, opening theory is not altered at all. The play of the middle game is not changed.

The endgame is altered, but even here, nothing in existing endgame theory becomes obsolete. The large difference in the value of a checkmate and a stalemate means that it remains just about as important as ever for the winning player to ensure a potential checkmate is not reduced to a stalemate.

The change, then, is only that endgame theory is augmented, because now it is enriched with the science of forcing stalemates in positions that before were merely abandoned as drawn.

It is possible that this won't be an unmixed blessing: the ability to put smaller advantages on the scoreboard might make players even more cautious than they already are. This could mean that draws would be still almost as common as they are now, after even less eventful play. Or, even worse, perhaps the advantage that White has by virtue of the first move might allow the player with the white pieces to force stalemate, in game after game. Even without these unfortunate results, Chess is already a grueling game at high levels: this rule might force players to work harder, playing many more moves, in order to win a victory that is only worth 1/5th as much as a real won game, and thus this innovation may be resisted for that reason.

But I think that it is quite likely to have beneficial results in making Chess matches more exciting, rather than the possible negative results that I've noted.

Since the first move gives the player with the white pieces an advantage, and since a change in the game which allows smaller advantages to count might increase the importance of that, I should note a possible rule change to reduce this.

In Chess, since turns alternate, half the time White has moved his pieces one more time than Black, and the other half of the time, White and Black have moved their pieces an equal number of times.

If White were to make one move on his first turn, and for each turn thereafter, each player made two moves, then half the time White would have made one extra move, and the other half Black would have made one extra move. But allowing two moves in each turn would be a major change to Chess, transforming it into something completely different.

But the same result could be obtained if each player made one move per turn, except that the very first move in the game, always made by the player with the white pieces were only half a move. This could be roughly approximated by not allowing White to make a two-step Pawn move as his first move.

This, unlike the change previously proposed, would radically change opening theory, by eliminating many of the openings considered strongest for White. It might make the openings where White begins with a Knight move considerably more popular, as they might be seen as the strongest of those remaining.

A more radical proposal for a modification of Chess, which gives a larger partial credit for stalemate, and which also allows partial wins to be obtained through bare king and even perpetual check, is now described on a subsequent page.

In one USENET post, instead of giving something like Chancellor Chess or Capablanca Chess, or the version of Hexagonal Chess invented by Wladyslaw Glinski, as possibilities when I wanted to give examples of how the rules of Chess might change in the distant future, I gave simpler examples based on changes that had already taken place in board games. Since Chess had changed from Shatranj by making existing pieces more powerful, one example was that the Queen might be replaced by an Amazon, which has the move of the Knight in addition to the existing Queen's move. Another example I gave was analogous to the change from the jeu plaisant to the jeu forcé in Checkers, that a rule might be adopted making it compulsory to capture if a Pawn capture is available.

These two examples were noted as illustrations of how history could repeat itself, with a regional variant of Chess eventually being recognized as an improvement, and supplanting the previous game. I noted, however, that the chance of such a development seemed extremely remote.

The Crisis in Chess

As I write these words in May of 2008, the United States Chess Federation is considering whether or not to make its flagship publication, Chess Life, optionally available to members by means of a web based subscription instead of on paper, allowing them to save money on their membership dues if they choose.

This is controversial, for two reasons.

If this were simply a conservation measure, proposed to save America's declining forests, that would be one thing. When it is an economy measure to respond to a declining membership, then concern becomes more intense.

Radio, television, video games, and the Internet have given people many new options for spending their spare time. But this hasn't made it hard to find a chess set in stores.

What, then, is the problem?

I think that one can speak of a crisis existing if we recognize that it is the organizational face of Chess that we are talking about. How many people take Chess seriously enough to join a Chess club? How much coverage do daily newspapers give to Chess tournaments and matches? This can vary, even if Chess as a pastime for individuals remains as popular as ever.

Also, in order to say that the situation, as it exists now, falls short of expectations, it is necessary to have a baseline of comparison.

When Wilhelm Steinitz switched from tactical play to positional play, and starting at the Vienna tournament of 1873 showed that this was a more effective way to play Chess, the crisis began.

A superior understanding of the factors which determine why one attack is successful, and another is not, is in itself a good thing; if Chess players aren't open to learning more about the game, and they refuse to play as well as they can, then they would be less likely to be taken seriously. But the increased understanding of Chess that Steinitz gave us had the unfortunate result that because Chess players understood what they had to do in order to mount a winning tactical assault, they played in a less flamboyant fashion, and, thus, chess games between higher-level players became less exciting on a surface level.

Individual players, such as Capablanca, Tal, Petrosian, or Fischer, by their ability and style of play, did still manage to squeeze out some room for exciting play at the higher reaches of Chess. The institutional support given to Chess under Communism in the Union of Soviet Socialist Republics, while it led to a discouragingly lop-sided Chess world, contributed to the continued institutional visibility of the game.

If we compare the visible world of Chess among experts, masters, and Grandmasters today, after Steinitz, after the collapse of the Soviet Union, with the game's popularity in the days of Labourdonnais and McDonnell, Anderssen and Kieseritzky, that we may be disappointed should not come as a surprise.

Given that the Soviet Union is not something we would wish to resurrect, and given that there is no way to put the knowledge of positional play revealed by Steinitz back in the bottle, it seems hard to avoid the conclusion that using a baseline of comparison which excludes these two major and irreversible factors is unrealistic.

Despite this, there are still two ways to respond. One can work through conventional measures, such as public education and outreach, to achieve realistic goals for maintaining and increasing the popularity of Chess. One can also view the past glories of Chess as showing the existence of an opportunity; if there were a way to reverse the impact of the transition from the Romantic Era to the Modern Era of Chess, then a large potential for growth would be unleashed.

It is this latter dream that has prompted so much effort being made in hopes that something such as the widespread adoption of Capablanca Chess, or Randomized Chess, particularly in the form developed by Fischer, or Seirawan Chess, a development of Pocket Knight Chess where the additional pocketed pieces are instead the extra pieces of Capablanca Chess might achieve this result.

While I think that there are sound practical reasons for fearing that this hope is illusory, it has not stopped inventors from proposing new chess variations. I have tried to take a hard look at the question, and with Dynamic Scoring and Random Variant Chess, I have tried to investigate if alternatives could be found that were at least slightly superior to previous proposals.

A Simpler Idea

Dynamic Scoring may be objected to on the basis of being too complicated, or on the basis that while it may encourage Black to play more boldly, White, instead of responding by playing more boldly still, could instead respond by playing more defensively.

I have come up with a simpler idea to reduce the number of draws in Chess, and a simpler idea to reduce the advantage to White of the opening move. While White's initial move advantage is not, in itself, a particularly great problem for Chess, that is because of the prevalence of draws, and so I believe it is an issue that does have to be addressed before one can address the draw issue aggressively.

Reducing White's Advantage

If White could only make half a move as his first move, then there would be no first move advantage. While this isn't really possible in Chess, the ability of a Pawn to move two steps forward on its first move suggests an obvious way to approximate this. One could suppress this privilege on White's first move.

But this doesn't seem to be that good an idea. All it would do would be to encourage the White player to restrict himself to playing the strong openings that still are not handicapped, by moving out one of his Knights on the first move.

So, how could one amend this particular way of equalizing the game so that it would promote a more diverse repertoire of openings instead of encouraging a more narrow one?

Here is what I suggest:

On White's first move, the possible choices of first move would have the following results:

In this way, the weaker Pawn openings are an option for avoiding the single-step penalty, but the strong Knight opening does not avoid it. Still, possibly shifting the penalty to an edge pawn might, under some circumstances, reduce its impact.

While this increases opening diversity by making the remaining first moves for White more equal in their value, it still reduces the total number of possible first moves from twenty to sixteen. If this is considered to be a serious problem, one could allow White yet another option: to advance a center Pawn two steps as his first move, in exchange for giving up the right to Castle. I don't really think that this is workable, though.

Since the urgent need to eliminate White's advantage, as noted above, comes from the fact that we're also going to take a radical measure to reduce draws in conjunction with this change, a milder penalty suggests itself: instead of giving up the ability to Castle, White could, in this case, lose the Pawn exchange privilege to be described below. Essentially, this makes a more conventional Chess game an option, although not a fully conventional one, since Black would still retain that new option.

The thing I dislike about this scheme is simply that instead of White's decision as to which course to take depending on his planned opening, it seems to me that a decision which makes itself felt so much later in the game is more likely to be made on White's estimate of the relative skills of himself and his opponent. To me, that seems to warp the game in an unfortunate manner. So I personally think it better to live with the fact that White now has 16 opening moves instead of 20.

On the other hand, if one takes away both Pawn exchange and Castling in return for the unrestricted double-step first move, that should at least be harmless, even if it's unclear that it will then make a difference in practice, since we now have an option that seems as though it would never be worthwhile. Or things could be even more finely graded, taking away only Pawn exchange for a first move of advancing a Knight's pawn a double step (thus further encouraging a maximum diversity of opening moves) while still taking away both Pawn exchange and Castling for an initial double-step move of the King's pawn or the Queen's pawn.

Reducing Draws

Above, on this page, it has been suggested that inflicting stalemate be allowed to put a plus score of 3/5 - 2/5 on the scoresheet. While this will reduce the number of draws somewhat, I feel that the effect of this measure alone will be slight.

Here, however, is a change to the rules that I think will have a drastic effect on the incidence of draws, since it will allow even a slight material advantage to be turned into a win:

Each player has the option, once during a game, to move a piece (other than the King or a Pawn) to his own back rank, and replace it by a Pawn.

But there is a price for this.

The player who uses this option now does not receive as much credit for winning.

If he checkmates his opponent, the score is now 3/5 - 2/5 instead of 1-0.

If he stalemates his opponent, instead of scoring 3/5 - 2/5 (as that rule change is to be combined with this proposal), he scores 0-1; that is, inflicting stalemate after taking advantage of the Pawn exchange privilege now loses the game.

This would significantly reduce the incidence of draws due to insufficient force, but again players would be strongly encouraged to win in a conventional manner instead, so that play which misses an opportunity to checkmate normally is still penalized.

A Further Modification

While the privilege of Pawn exchange is a very drastic measure to reduce the number of draws, making stalemate a loss for players who avail themselves of that option may work against that. As well, in Dynamic Scoring, in addition to allowing a player to put a smaller positive amount on the scoresheet through forcing stalemate, bare king and perpetual check are also allowable victories.

Thus, perhaps this would be a more appropriate scoring schedule:

Game Outcome Normal Division of Points Points if Win After Pawn Exchange
Checkmate. 1000:0 600:400
Stalemate. 600:400 520:480
Bare King 550:450 510:490
Perpetual Check 510:490 502:498
Draw 500:500 500:500

Instead of one point per game, with fractions, the gradations required make it convenient to use 1000 points for the game to be split between the players.

Note that the score for stalemate after pawn exchange is different from that for bare king without pawn exchange; instead of pushing the score down one level, the rule is that the benefit to the winner is always divided by five after pawn exchange, while bare king is worth half as much, not one fifth as much, as stalemate.

[Next] [Up] [Previous]