Monday, June 25, 2007

Metaphysics of Chess (revised and extended)

I.The game of chess is a draw. From the draw of the starting position, for every move by White, there is at least one equivalent move by Black that maintains the draw.

II.The only possible way a player can win is if the other player commits an error, usually many errors. Taking advantage of errors is not always easy, and so, even in the presence of errors, wins are not always obtained.

III.Chess is a game of incremental improvement or degradation of a position. From a player’s perspective, the increment may be small or large, positive or negative, depending on the quality of the move, and which side makes the move.

IV.In a perfectly played game of chess, the incremental change for each and every move is zero.

V.The incremental change away from the draw of the starting position is the result of the move of the players. In the case of blunders, the increment may be large. In the case of excellent moves, the increment is zero.

VI.There are two possible results of a player’s move: 1) the position for the player making the move is of the same quality as before the move, 2) the position for the player making the move is worse than before the move. A player by his move cannot make his position better. Thus, degradation of one’s position comes as the result of one’s own moves, while improvement of one’s position comes as the result of errors of one’s opponent.

VII.Fears that the opponent's previous move has degraded one's position are unfounded.

VIII.Chess is a game of virtually an infinite number of positions. Thus, for the human mind, calculation in non-forcing positions is filled with error. However, in any given game, most of the positions can be played very well without deep calculation.

IX.No matter the rating difference between player 'A' and 'B', the level of understanding of some positions and the skill in playing them may be 1) exactly the same, 2) greater for player 'A', 3) greater for player 'B'. Understanding and skill in some position does not guarantee understanding and skill in all, or even other, positions.

X.Most players have an average ability to visualize and calculate future chess positions and moves. A more skillful player is the result of having learned how to play more positions than the weaker player, not in having greater calculation ability.

You can only win if your opponent loses. By "loses," is meant "makes bad moves."
You cannot wrest the win from your opponent by force. Your opponent must, essentially, give it to you.

Minimizing the errors in your own play is the first step towards chess mastery, for, when you remove your errors, your opponent cannot win.

Identifying the errors in your opponent’s play is the second step towards chess mastery, for you cannot win without taking advantage of your opponent’s errors.

Question: Is it possible to make your opponent commit an error? Answer: No, however, by your move you can attempt to present him with problems that are difficult to answer. This is the general idea of gambits.

Question: I’m a beginner; must I play in such a way as to present problems to my opponents? Answer: yes and no. First the no: chess is of sufficient complexity that at the beginner level, your opponent will make errors of his own accord. Whether or not you can identify them and take advantage of them is another question. Now to the yes: in a sense every good move presents a problem. For example, 1. e4 occupies the center, and this is a problem for Black. There are a number of ways to suitably address this problem: e5, d5, c5, Nf6, etc. However, h5, as far as we can determine, is not one of them and probably degrades Blacks position (although it probably does not degrade it enough to guarantee a win for White).

Question: I’m rated 1500 and I am playing a guy rated 1900. Should I be worried? Answer: Not overly so. 1) His calculation ability is similar to yours. 2) He cannot win unless you make errors. 3) He probably knows how to play more positions, but there are plenty of positions that he doesn’t know how to play any better than you. Therefore, don’t worry; minimize your errors and take advantage of the opportunities that he will undoubtedly present to you, and above all, have fun.

Question: My opponent just played a great move. Am I lost? Answer: If you were not lost before your opponent played the move, then you cannot be lost after he played the move. If the position was balanced before his move, then there must be resources in the position to meet his challenge. Whether you can find them or not is a measure of your skill as a chess player.

Question: I cannot calculate 10 moves ahead in this position. Should I be worried? Answer: No. In many non-forcing positions, Grandmasters cannot calculate future moves with any certainty. First in non-forcing positions, you have no idea what move your opponent will play, and second from non-forcing positions there are too many positions to keep straight. In such, situations it is futile to calculate deep and lofty plans. However, from forcing positions it is relatively easy to calculate deeply. In most cases, reliable calculation is the result of 1) familiarity with the position, or similar positions, and 2) chunking. As applied to chess, in this instance, chunking is the grouping together of individual moves into one unit or chunk. Known positions, patterns and tactics allow the player to chunk many moves into one unit and achieve greater tactical acumen. Finally, many positions can be played sufficiently well without deep calculation.

Question: You say that a player cannot improve his position by his move, but what about the so called, "accumulation of small advantages," that some of the greatest players have practiced? Answer: This expression simply means taking advantage of the errors introduced by one’s opponent. But the move itself doesn’t create something in the position that was not there before; that would violate the law of non-contradiction.

Question: Aren’t you advocating very passive chess? Answer: Not at all. Players have different styles. Some like to play cutting edge, aggressive moves that seek to force one’s opponent to find a difficult best continuation and thereby increase the likelihood of error. Others like slower, more positional, games, just letting the fruit ripen on the vine, as it were. Subjectively, one approach may be better for you than the other, and subjectively, one approach may be more difficult for your opponent than the other. But this does not change the objective evaluation of the position. This presentation has tried to emphasize two things: 1) You should not unduly pressurize yourself to play great, incredible, moves that will win, for you cannot win, unless your opponent makes an error, usually more than one. 2) While you should respect your opponent’s move, you should not fear his move, for he cannot create something in the position that is not already there. If the position was good before his move, then it must be good after his move, and there must be resources available to meet his move.


Blue Devil Knight said...

Good stuff. Heisman discusses some similar ideas in his fun book 'Elements of positional evaluation.' I think he is working on an updated version.

HardDaysKnight said...

BDK: Thanks, I'll look for Heisman’s book. It's been fun to put things down in this way, and hopefully it will help some aspects of my game.

Now, back to tactics!

Temposchlucker said...

I hope you don't mind I try to show a different viewpoint.
I see some problems with this kind of reasoning.

It is not proven at all that chess is a draw. Quite a few facts even indicate different:
Most games between humans DON'T end in a draw.
Most games between computers don't end in a draw.
The game isn't symmetrical because someone has the initiative. This initiative causes black to lose if he insists to play symmetrical.
So white has an initial advantage. How do you know that this advantage isn't decisive by perfect play?

Let's for the sake of reasoning assume that chess is a draw. And let we call perfect play = 100%
How far are the two best humans away from perfect chess?
98% vs 96%?
45% vs 40%?
7% vs 6.85%?
We don't know, but since most games don't end in a draw it must be far away from 100%

I'm rated 1743, I often win from someone who has a rating of 1954 while I always lose from somebody who has 1820. Which means that I am able to exploit the bad moves of a 1954-player while I'm not able to make use of the bad moves of the 1820-player, however he makes more bad moves.

In this light, how useful can your conclusions be?

HardDaysKnight said...

Tempo:"So white has an initial advantage. How do you know that this advantage isn't decisive by perfect play?"

How do I know? I don’t. It’s my presupposition. Just like you don’t know that with perfect play White has an advantage sufficient to win. But that can be your presupposition (you’re allowed to have, ‘em), and you can develop a theory of play based on it, if you want. The presupposition allows me to make sense of the game, to the extent that it leads me to the conclusion that, assuming that any given position is a draw before my opponent’s move, then it must be a draw after his move. He does not have the power to create something that is not there in the first place---sort of a law of conservation of winning energy in chess; thus, after his move, there must be resources in the resulting position to cope with, to balance, his move. Really, it’s this law that is the point. And, it was something that I needed to think about, because I have a tendency to see dragons where there aren’t any. Anyway, even with a presupposition that from the start White has a sufficient advantage to win, the same ideas apply.

"We don't know, but since most games don't end in a draw it must be far away from 100%"

I totally agree with you; therefore, if I minimize my errors, and take advantage of his, I will win.

"In this light, how useful can your conclusions be?"

That you are able to exploit the positions presented to you by the 1954 player and not the 1850 player makes perfect sense. Just because a player is higher rated doesn’t mean he understands all positions better than a lower rated player; I would theorize that you understand the set of positions presented to you by the 1954 better than you understand those presented by the 1850. (That, or may be the 1850 is a beautiful blonde in a low-cut dress! ;)

Beyond that, I’ve always sort regretted that chess is a closed system into which no new energy is presented. In games I guess this is usually accomplished by a randomizing factor. But now, that fact has made it clear that the other player can’t … magically… take the game away from me, nor I from them. Any win must flow from previous error. My function as a chess player is to minimize my own errors and exploit those of my opponent. This can be approached in either a Kasparov-like or a Petrosian-like manner. I find this line of thinking very useful. But your mileage may vary.

Regards, HDK

The retired pawn said...

HDK: This is a very good list and I have thought about it often during my losses after I tried it. It is easier thought about than employed. However, it's tenements are sound and should lead to more wins in the long run.

BTW, I have tried to send BDK an e-mail to no avail...I also sent you one...hope you don't mind.

Blue Devil Knight said...

I would be very surprised if the game isn't a theoretical draw with best play from both sides. Regardless of what happens in practice (though, in practice the better two opponents, the more likely that they will draw: take the limit out to infinite playing strength and you get 100% draws :)).

Quantum computers will tell us...if they are ever built.