Moderators: hgm, Rebel, chrisw
So 0.15 would be a reasonable approximation of the probability of making a mistake (as defined in the thread's first post) using current technology.Dann Corbit wrote:Even the best chess engines make mistakes all the time. I guess that if two strong chess engines play a game of 100 moves, there are at least 30 mistakes in it.
As a SWAGtowforce wrote:So 0.15 would be a reasonable approximation of the probability of making a mistake (as defined in the thread's first post) using current technology.Dann Corbit wrote:Even the best chess engines make mistakes all the time. I guess that if two strong chess engines play a game of 100 moves, there are at least 30 mistakes in it.
Thanks Dann - this is very interesting!
it may be that near-perfect play can be achieved with a relatively short ply depth
What is ply depth? [...]
If you're unhappy with "ply depth", then feel free to think in terms of either full-width ply depth, or, if you wish to focus on "modern engines", substitute "positions evaluated per second" for "ply depth".
No. My example of endgame knowledge about wrong bishops when rook pawns are involved could usefully be applied during middlegame decisions about which pieces to exchange. Positions are "encountered" in internal analysis as well as on the board.I also believe, though, that with the higher standard of play that will result from bigger searches, positions where particular pieces of knowledge are important are just simply less likely to be encountered.
They make mistakes but I do not believe that they make mistakes so often.Dann Corbit wrote: Even the best chess engines make mistakes all the time. I guess that if two strong chess engines play a game of 100 moves, there are at least 30 mistakes in it.
I am very skeptical of this claim. Of course in part it will depend on what you mean by "mistakes". If by mistake you mean a move that (with perfect play) changes a won position to drawn or lost, or a drawn position to lost, then the estimate of 30 mistakes in 100 moves is clearly way too high. On the other hand other definitions of mistake will be unavoidably subjective.Dann Corbit wrote:Even the best chess engines make mistakes all the time. I guess that if two strong chess engines play a game of 100 moves, there are at least 30 mistakes in it.
If an item of knowledge can make a difference between making a mistake (as defined in the thread's opening post) or not in the endgame, and it is important to understand it in the middlegame, then it could be said to be useful in the early middlegame - or possibly even the opening.Marek Soszynski wrote:No. My example of endgame knowledge about wrong bishops when rook pawns are involved could usefully be applied during middlegame decisions about which pieces to exchange. Positions are "encountered" in internal analysis as well as on the board.I also believe, though, that with the higher standard of play that will result from bigger searches, positions where particular pieces of knowledge are important are just simply less likely to be encountered.
Please take this subject here if you want any meaningful answers.towforce wrote:If an item of knowledge can make a difference between making a mistake (as defined in the thread's opening post) or not in the endgame, and it is important to understand it in the middlegame, then it could be said to be useful in the early middlegame - or possibly even the opening.Marek Soszynski wrote:No. My example of endgame knowledge about wrong bishops when rook pawns are involved could usefully be applied during middlegame decisions about which pieces to exchange. Positions are "encountered" in internal analysis as well as on the board.I also believe, though, that with the higher standard of play that will result from bigger searches, positions where particular pieces of knowledge are important are just simply less likely to be encountered.
For many items of knowledge, it may well be that they increase in importance (relative to other pieces of knowledge) the closer you get to a "key" move (one where possesion of a piece of knowledge could make the difference between making a mistake or not).
In any case, I have already acknowledged that greater knowledge will reduce the search depth required to play near-perfect chess - in the thread's opening post, I wrote, "it may be that near-perfect play can be achieved with a relatively short ply depth (the exact depth required will, of course, depend on the quality of the eval - but I would be surprised if the difference between poor eval and good eval was more than the equivalent of an extra 5-10 ply of full-width search)".
maybe it even means engine making a move that gives somewhat +2 score advantage in its eval but missing the better move in that certain position that would have given it the considerable advantage..smirobth wrote:I am very skeptical of this claim. Of course in part it will depend on what you mean by "mistakes". If by mistake you mean a move that (with perfect play) changes a won position to drawn or lost, or a drawn position to lost, then the estimate of 30 mistakes in 100 moves is clearly way too high. On the other hand other definitions of mistake will be unavoidably subjective.Dann Corbit wrote:Even the best chess engines make mistakes all the time. I guess that if two strong chess engines play a game of 100 moves, there are at least 30 mistakes in it.
Chess is a game when one mistake may be enough to lose so you do not need many mistakes of the opponent to be significantly stronger than the opponent by hundreds of elo.Dann Corbit wrote:I think that when the board is cluttered, the best moves are often missed.
So that if you had a 64 CPU version of Rybka, it will find moves that single CPU Rybka missed and beat it handily. So the choices Rybka makes are not perfect or even nearly perfect. I also think that chess programs can improve by hundreds of Elo still.
I think that perfect play might be as difficult to achieve as proving the game. Note that I think that these are two completely different problems.