Stalemate wins in K + P versus K

[Mats Winther]

"Stalemate wins if the king is blocked by an enemy pawn on the seventh rank"

Nigel Short has suggested that stalemate should count as win, as in Xiangqi. However, I am against the introduction of this rule in Western chess, since so many fine qualities are lost. I suggest instead K + P versus K as only exception.

Thus, stalemate is only a win in endgames K + P versus K. Only this little change will have an enormous impact and will reduce the number of draws. Endgames with light piece + pawn and rook + pawn will be less drawish, too.

Alternatively, the rule could be formulated as this: stalemate is only a win if the king is blocked by an enemy pawn on the seventh rank. (After all, the party with a pawn up could also be stalemated. But this shouldn't lead to a win.) Thus, the stalemated party could have additional pawns and pieces on the board, provided that they are blocked or pinned. I have implemented this variant as a Zillions program, called "Stalemate Chess":
http://www.two-paths.com/bg/stalematechess.htm

Comparatively, the rules in Xiangqi that concern draws are quite complex, especially in connection with repetition. Normally, the party who keeps repeating the position for defensive purposes will lose if position is repeated three times. However, there are exceptions. In Xiangqi they have created complex rules in order to maximize the quality of the game. So there's no reason why we can't have exceptions also in Western chess. K + P versus K could be an exception where stalemate leads to win.

Since being a pawn up must now be regarded as slightly better than before, gambit openings like the Marshall attack will probably be evaluated as not quite as good as before (although it is unlikely that the Marshall attack will lead to such a trivial endgame). This could be a valuable change, since the Marshall attack is effectively a refutation of 1.e4.

M. Winther

[Roger de Coverly]

Thus, stalemate is only a win in endgames K + P versus K. Only this little change will have an enormous impact and will reduce the number of draws.

Why do you want to reduce the number of draws? The game of chess is sufficiently balanced that a draw is a natural outcome between two players of equal skill.

That the resulting King and pawn endings are drawn is the reason many Rook and pawn ones are drawn as well.

The more serious issue is that increasing the number of wins could easily tip the current balance in favour of white from the 52 to 55 % range to 60% or higher. When that happens, you just don't want to play Black.

The Marshall cannot be regarded as a refutation of 1. e4. At best it suggess that the move 8. c3 weakening the d3 square may not be the best choice. There's plenty of lines and ideas based on avoiding this weakness, by playing d3 first.

[Mats Winther]

Roger, playing d3 first isn't regarded a theoretical way to secure an advantage. It is simply slow play, which works fine. But it usually ends in a draw.

I don't think it would lead to a significant increase of white's advantage. That would mean that most pawn endgames with pawn up have white as the stronger party. I'm not so certain that this imbalance exists.

Arguably, it isn't fair that a pawn up so often leads to a draw. It would make the game more exciting if K + P versus K is always a win (except in trivial cases).
/Mats

[Roger de Coverly]

playing d3 first isn't regarded a theoretical way to secure an advantage.

It's increasingly popular as a method amongst top GMs. In any case what is a theoretical way?

Arguably, it isn't fair that a pawn up so often leads to a draw. It would make the game more exciting if K + P versus K is always a win (except in trivial cases).

If you cannot sacrifice a pawn to secure a draw, that could make for less adventurous play.

Stalemate not a draw changes almost all the perceived wisdom of chess and chess players. Perhaps someone should modify the rules on a chess server and see whether there was any popularity to the chess variant.

[Matt Mackenzie]

Changing the stalemate rule just sucks. Period.

[Mats Winther]

Changing the stalemate rule just sucks. Period.

So they said when they introduced the queen in the fifteenth century, to replace the visir. And they probably said the same when the en passant move was introduced in 1490. In musical quarters there was an outcry when the cello ousted the viola da gamba. Today, we just love the cello and couldn't live without it.

Times are always changing, people change, and thus the rules and the instruments also change to reflect the developments in human nature and society.
/Mats

[Mike Truran]

Well, FIDE aren't going to change the laws of chess based on a few comments on this forum. In that sense any debate is something of a waste of time.

[Mats Winther]

Well, FIDE aren't going to change the laws of chess based on a few comments on this forum. In that sense any debate is something of a waste of time.

Fide decided to endorse Fischer Random chess, which cannot be regarded a success. This decision likely depended on "debates" and lobbying. The problem is that elité chess more and more revolves around Spanish with d3, and the drawish Berlin variation, etc. Chess is becoming more and more minimalistic, trying to win by small means. The scope of chess strategy is shrinking, which is a bad development.
/Mats

[Lewis Martin]

Arguably, it isn't fair that a pawn up so often leads to a draw.
/Mats

Why is this not fair? Chess is a game of skill, and being a pawn (or any piece) up is almost immaterial compared to the nature of the position. (Also with regards to a sacrifice as mentioned by RdC above)

Excellent defence should also be rewarded. For your point, you might as well change the objective of the game to: whoever is ahead in material first, wins.

[Christopher Kreuzer]

I dunno. If you fall a pawn behind, the temptation can be to cling on for a draw later. If a rule like this came in, it would mean people would have to go on the counter-attack much sooner. Might make for more open games. I wonder if some of the really nice endgame studies would carry over under such a rule change, or whether new concepts would be worked out. Not that there is any shortage of endgame studies to still be composed under the existing rules!

[Kevin Thurlow]

"Fide decided to endorse Fischer Random chess, which cannot be regarded a success. This decision likely depended on "debates" and lobbying. The problem is that elité chess more and more revolves around Spanish with d3, and the drawish Berlin variation, etc. Chess is becoming more and more minimalistic, trying to win by small means. The scope of chess strategy is shrinking, which is a bad development."

If the scope of chess strategy is shrinking, one answer is to play Chess 960 (aka Fischer Random chess).

[Fintan Hegarty]

I dunno. If you fall a pawn behind, the temptation can be to cling on for a draw later. If a rule like this came in, it would mean people would have to go on the counter-attack much sooner. Might make for more open games.

Surely any extra effort the "losing" player would supposedly have to make would be counterbalanced by the reduced effort the "winning" player would have to put in?

[Mats Winther]

No, chess players, as a rule, don't like FRC since there is a lack of control. However, it can be remedied by Relocation variants.



The above diagram shows the most simple variant, which produces 64 different positions. FRC castling is employed. Both players can relocate either the king or queen. Black begins by making the initial swap. Alternatively he can choose to leave the position as it is. The white player then has the option to relocate his king or queen, whereupon he starts the game by making the first move.



It's also possible to think of other changes such as Valiant Chess where knight pawns at the fifth rank have an additional jump move to empty square. The extra capability of the advanced knight pawn reinforces an attacking strategy.



Another idea is to introduce extended castling, as in Castle Chess and Adjutant Chess. When castling the king may jump three squares, but it can also jump two as usual. The rook ends up on its usual square. This change greatly enlarges the opening tree as many more opening variations become practicable. The longer jump of the king is useful in most variants with queenside castle. At kingside castle, it is sometimes worthwhile to place the king on a more protected square. This ought to be useful, for instance, in King's gambit variations. However, one must keep in mind that the king is removed farther from the centre, and the king no longer protects the vital f2/f7 square. Sometimes it's better to castle normal.

M. Winther

[David Pardoe]

This idea is basically rubbish....BUT

Why not give the extra credit to the player with the extra material in the form of extra grading points...

Maybe such games, where advantages of various kinds occur, could be translated into a set of bonus grading points..
Maybe 10 bonus grading points for the extra pawn, maybe 20 for a N or B...
Maybe 15 for a positional plus position...
And maybe 5 points for a clock advantage greater than 10 mins.....10 points if greater than 20 mins...
And 15 points if higher still....

[Roger de Coverly]

Why not give the extra credit to the player with the extra material in the form of extra grading points...

Grading points are meant to be used for ranking players in order of strength using results as a proxy, not rewards and prizes.