An idea to kill draws in computer chess by different rules

[Ras]

It seems we'd want a game where we have more frequent gambits and sacrifices that are spectacular

Successful sacrifices don't come out of nowhere - you need a positional advantage. These days, such an advantage doesn't fall from the sky, you have to work for it with... positional play.

[mvanthoor]

I agree that a perfectly played game should be a draw. The problem is that with current rules, one player must make either several mistakes or one very big mistake to lose. Ideally any mistake should lose, a draw should result only from ideal play which no one should ever be able to achieve. Or at least draws should result from only nearly perfect play. That's what this is all about.

The top humans already play as perfectly as humanly possible. Engines are even more perfect, especially if the opening books and endgame databases are big. Thus, you'll have a lot of draws. That's just the nature of the game.

In professional Go, at 9th dan level, only two things happen:
- A player resigns when he can see that he's clearly going to end up with less territory (taking Komi into account) if the game continues to its end.
- The game is played out to counting because its very close and neither player has the time to keep counting for each move.

The number of games decided by a 0.5 to 2.5 point margin is staggering. I even think (but don't know for sure) that current-day 9-dan pro's rarely win or lose a game with more than the difference of Komi.

To get wins or losses as close as this in chess is almost impossible. It'll probably mean that we chuck a "draw" position into an engine such as Stockfish or Komodo, have it calculate for X minutes / up to move Y, and call the player with the plus evaluation the winner. (We add 0.25 - 0.5 for Black, because of White's first move advantage; i.e., "Komi".)

If the engine evaluates to 0.00, Black wins, because White started with initiative. That''d be the 0.5 point in Komi.

[Ovyron]

It'll probably mean that we chuck a "draw" position into an engine such as Stockfish or Komodo, have it calculate for X minutes / up to move Y, and call the player with the plus evaluation the winner. (We add 0.25 - 0.5 for Black, because of White's first move advantage; i.e., "Komi".)

The position is actually 0.00, the farther away it is evaluated from 0 by the engines, the worse they misevaluate the position. Giving wins based on flawed evals wouldn't be satisfactory.

[corres]

...

You want to kill the classical chess.
That is the chess play, because the draw belongs to the essence of chess play.
Who is tired of draw playing classical chess, that person should not play classical chess.
If you want to find out a derivative of classical chess you ought to call it other than chess.

[corres]

I don't even like the underlying idea. In chess, one player can only win if the other player makes a mistake. In absence of mistakes, i.e. with perfect play, draw is the natural result - and not a problem to be solved.

I agree that a perfectly played game should be a draw. The problem is that with current rules, one player must make either several mistakes or one very big mistake to lose. Ideally any mistake should lose, a draw should result only from ideal play which no one should ever be able to achieve. Or at least draws should result from only nearly perfect play. That's what this is all about.

For repairing the mistake it also needs knowledge what should reward - this is the draw, without perfect play.