yes
A perfect game of chess is always a draw. Discuss.
Sort:
Mar 27, 2024
0
#63
tygxc wrote:
A perfect game is a game where both players play optimally.
Optimal play is play without errors.
An error is a move that worsens the game state won/drawn/lost.
The linked 97 perfect games were played by ICCF (grand)masters with engines at a pace of 50 days per 10 moves.
The point is that whatever white tries, black has several ways to draw against it.
Fact is you have absolutely zero reason to assume they're perfect, most were agreed draws and they're not even the same rules.
Fact is you have absolutely zero reason to doubt these games are perfect. They are perfect.
Most are hard fought and end in 3-fold repetition or a 7-men endgame table base draw.
The rules are slightly more decisive, that is even more of an argument that chess is indeed a draw and does even more indicate these are perfect games.
Mar 27, 2024
0
#65
I'm solidly in the "chess is a draw with perfect play" camp.
In fact, I consider the margin of the draw to be fairly wide... in my experience, it usually takes MORE than one error to lose. One error will give you a bad game and a long, miserable struggle for the draw. Two errors will lose. IMO.
@65
That depends on how you define an error.
If error = move that worsens the game state won/drawn/lost, then 1 error suffices to lose.
However consensus is that black can afford to lose a tempo and still draw and white can afford to lose 2 tempi and still draw.
Mar 28, 2024
0
#67
tygxc wrote:
Fact is you have absolutely zero reason to doubt these games are perfect. They are perfect.
...
Ah but I do. I can check the performance of the strongest players against the perfect tablebases and they don't do too well. Those positions are much much simpler. The idea that the top players play perfectly in complicated positions where you can't check and only fluff in simple positions where you can is laughable.
@67
ICCF (grand)masters with engines @ 5 days/move average are much much stronger than the strongest human players.
It is not 2 3000+ engines playing at 5 s/move or 5 min/love or 5 h/move, but 5 day/move jockeyed by 2 ICCF (grand)masters. That is an estimated 4000+ strength.
Over the years the number of decisive games in the ICCF WC Finals has dwindled. Last year I needed a Poisson distribution to estimate the number of errors from the sparse decisive games.
Now it is 97 draws in 97 games, except 10 cases of exceeded time limit in otherwise drawn positions. Let us await how the 39 ongoing games end.
Each game represents 3 engine years, or about 10^15 considered positions.
As previously calculated weakly solving Chess requires 10^17 positions.
So 100 such ICCF WC draws together represent about a weak solution to Chess and certainly an ultra-weak solution.
For the initial position the strategy to achieve the game-theoretic value of a draw against any opposition is to follow an ICCF WC draw for as long as possible and then switch to an engine at 5 days/move jockeyed by an ICCF (grand)master until a 7-men endgame table base draw or a prior 3-fold repetition is reached.
Mar 28, 2024
0
#69
You don't half talk a load of BS. I already invited you to try your "method" on a whole bunch of games that SF drew against itself from a winning position and (naturally) you ignored it.
@69
"SF drew against itself" ++ At 5 days/move?
Mar 28, 2024
0
#71
It wouldn't get too much deeper in 5 days than it did in the 37 min/move a games anyway. Still nowhere near down to its ceiling. But I also showed it's blunder rate was generally increasing with time per move if you remember, at least in KNNKP.
"Do we have to go through all this again?"
++ No, your bunch of games is completely irrelevant. ICCF players use 7-men endgame table bases, so anything with 7 men or less is irrelevant: strongly solved already.
For weakly solving chess only drawn positions are relevant: the point is to prove that at least one path exists from the drawn initial position to other drawn positions and finally to a 7-men endgame table base draw or a prior 3-fold repetition. That is exactly what the 97 ICCF WC draws have done: from the initial position at 5 days/move to position, position and finally a 7-men endgame table base draw or a prior 3-fold repetition.
Interesting is that whatever white tries, black has not one but several paths to the draw.
Mar 28, 2024
0
#74
tygxc wrote:
"Do we have to go through all this again?"
++ No, your bunch of games is completely irrelevant. ICCF players use 7-men endgame table bases, so anything with 7 men or less is irrelevant: strongly solved already.
Not remotely strongly solved already under FIDE competition rules, nor under ICCF rules. Go back and read the earlier discussions.
And that's what I meant by not even the same rules. Your games are irrelevant not just because they're different rules but you have absolutely no way of estimating how far adrift from perfect play they are.
SF without tablebase will draw all or almost games from both drawing or winning positions when it's out of its depth. You wouldn't be able to point out the mistakes it makes in 7 man positions nor distinguish between the winning and drawn positions without consulting a tablebase, but you expect us to believe you can recognise perfection just so long as there's no way of checking.
For weakly solving chess only drawn positions are relevant: the point is to prove that at least one path exists from the drawn initial position to other drawn positions and finally to a 7-men endgame table base draw or a prior 3-fold repetition. That is exactly what the 97 ICCF WC draws have done:
No it's not the point if the starting position is winning. You have no valid reason to assume it's drawn. You don't know if the players are getting any closer to perfection or just emulating SF when it gets out of its depth in KNNKP.
from the initial position at 5 days/move to position, position and finally a 7-men endgame table base draw or a prior 3-fold repetition.
And they're mostly agreed draws anyway in positions we'd still have no chance of proving drawn.
And we have been through all this before. You're as bad as dandelions in the garden.
Why not read and take in what's already been discussed.
@74
"Not remotely strongly solved already under FIDE competition rules, nor under ICCF rules."
++ The 7 men endgame table base has strongly solved all 7-men positions. ICCF rules are more decisive than FIDE rules, as they allow 7-men endgame table base win claims that exceed 50 moves, but such claims never occur.
"Go back and read the earlier discussions." ++ That is advice for yourself.
"And that's what I meant by not even the same rules." ++ ICCF rules are more decisive than FIDE rules, so if 100% of games draw under ICCF rules, then they draw under FIDE rules too.
"you have absolutely no way of estimating how far adrift from perfect play they are"
++ I have. If some errors were present, then some decisive games would occur.
"you can recognise perfection just so long as there's no way of checking"
++ If the games were imperfect, then some decisive games would occur.
"if the starting position is winning" ++ It is not, otherwise some decisive games would occur.
"You have no valid reason to assume it's drawn" ++ I have 97 valid reasons.
"they're mostly agreed draws" ++ No that is a lie. Take e.g. the most recently finished game:
https://www.iccf.com/game?id=1360178
It ends in a draw by 3-fold repetition.
"Why not read and take in what's already been discussed." ++ Take your own advice.
A perfect game of chess that ends in a draw is often seen as a symbol of equal competition and the high level of skill of both players. In such a game, both participants demonstrate deep analysis, strategic thinking and the ability to predict the opponent’s moves. A draw emphasizes balance and mutual respect between opponents, while at the same time leaving room for further study of the game and improvement of skills. However, despite this, for some players, winning or losing may be a more preferable and stimulating outcome than a draw.
Mar 28, 2024
0
#77
tygxc wrote:
@74
"Not remotely strongly solved already under FIDE competition rules, nor under ICCF rules."
++ The 7 men endgame table base has strongly solved all 7-men positions.
No such thing as the 7 men endgame table base.
No tablebases have solved all 7 man positions strongly or otherwise for any set of rules. Go back and read previous posts, but this time try to take it in.
ICCF rules are more decisive than FIDE rules, as they allow 7-men endgame table base win claims that exceed 50 moves, but such claims never occur.
ICCF rules are probably not what OP was asking about. He is obviously asking about theoretically perfect play in any case, so whether such claims have occured in practice is immaterial.
"Go back and read the earlier discussions." ++ That is advice for yourself.
"And that's what I meant by not even the same rules." ++ ICCF rules are more decisive than FIDE rules, so if 100% of games draw under ICCF rules, then they draw under FIDE rules too.
No. ICCF rules are more decisive than FIDE competition rules but less decisive than FIDE basic rules. You proved that you were totally confused by the difference though it was many times explained to you.
Obviously 100% of theoretical games don't draw under ICCF rules. Whether they would with perfect play is not known, nor for any version of the FIDE rules (which pretty much answers OP's question).
"you have absolutely no way of estimating how far adrift from perfect play they are"
++ I have. If some errors were present, then some decisive games would occur.
Nonsense, SF draws against itself pretty much 100% when it's out of its depth from winning or drawn positions that can be checked by tablebase. It's generally SF playing, just there's no tablebases to check most of the time.
According to your logic the games posted here contain no errors, but Syzygy doesn't agree with you. They all contain blunders.
"you can recognise perfection just so long as there's no way of checking"
++ If the games were imperfect, then some decisive games would occur.
Or maybe that should be some draws would occur. You don't know.
"if the starting position is winning" ++ It is not, otherwise some decisive games would occur.
Actually they do, but it's a complete non sequitur anyway.
"You have no valid reason to assume it's drawn" ++ I have 97 valid reasons.
97 irrelevant red herrings. Totally nothing to do with the subject
"they're mostly agreed draws" ++ No that is a lie. Take e.g. the most recently finished game:
https://www.iccf.com/game?id=1360178
It ends in a draw by 3-fold repetition.
Oh sorry. That proves it then! Two players finish in a 3-fold repetition and chess is solved. Amazing!
"Why not read and take in what's already been discussed." ++ Take your own advice.
I take it that means you're not going to. I expected nothing less.
I try to explain once more.
Observed fact: 97 games all draws in the ICCF World Championship Finals, by ICCF (grand)masters with engines.
10 exceeded the 50 days / 10 move time limit in otherwise drawn positions.
Hypothesis: chess is not a draw.
Then all 97 games must contain an odd number of errors (?).
An error (?) is a move that changes the game state from draw to loss or from won to drawn,
and a blunder (??) changes the game state from won to lost and counts as a double error.
So all 97 games would contain 1, 3, 5 ... errors and none would contain 0, 2, 4... errors.
It is absurd that some games would contain 1 or 3 errors and none would contain 2.
It is absurd that some games would contain 1 error and none would contain 0 or 2.
The hypothesis was false.
Thus chess is a draw.
Hypothesis: chess is a draw and the 97 games are not perfect.
Thus all 97 games contain an even number of errors: 0, 2, 4...
It is absurd that games would contain 0, 2, 4 errors and none would contain 1 error.
Thus all 97 games must contain 0 errors.
The hypothesis was false.
Thus all 97 games are perfect games.
Mar 29, 2024
0
#79
And by exactly the same logic all 12 games posted here (on the first page) are perfect and the position they start from is a draw.
Except none of them are perfect and the position they start from is a win for Black. The Syzygy tablebases (which are perfect) tell us that beyond doubt.
Does that not suggest to you that there is a flaw in your logic?
The flaw has been comprehensively explained to you in past posts. I suggest again you read them and take in what they say.
@60
White has the first move, is a tempo up and has the initiative. White tries to win. Black tries to draw. White fails, black succeeds and even has several ways to draw against whatever white tries.