perfect play = draw??

So, the prevailing logic is that even though white is slightly better at the beginning, he isn't slightly better enough to actually win, therefore it's a draw?

Not exactly rock-solid logic, ha ha.

these are just raw data about some endings I don't care about. Irrelevant for assesing the starting position of the game as equal. To make use of these stats, you need to find a way for white or black (some have claimed black may well have a forced win from the starting position too!) to get into these endings. Good luck with that. Otherwise they're just random information.

If you can't really make sense of that... Well good luck with life.

If I thought it was rock solid I would have used a stronger word than "suspect".

You're trying to use logic, but your argument has no credibility because you obviously rarely play chess! It's obvious to all strong players that white's advantage is going nowhere, even if we can't play 90 moves of perfect play to prove it's drawn. Obviously we can't prove it and there is a possibility, albeit an extremely strong one, that somehow by force white will be able to win a pawn or something and win the game. Anything close to that has never really happened in any high level chess game, if one loses they may not even make any real mistakes, just a bad strategy perhaps. But if just about every move they play is best with the best strategies then it's always a draw in those games.

Are you seriously saying that no high level chess game has ever been won straight out of the opening?

No, but any game won out of the opening even in top level chess involves an outright blunder or big mistake, and when is that ever forced? If anything, the wins will have to be grinded out.

Or a strong opening novelty, against which it isn't immediately clear what the best course is.

For instance, last year suddenly 10.e5 against the Najdorf Poisoned Pawn turned out to be extremely dangerous after all after a very strong novelty; it took months before an antidote was found. Quite a few GM level games were lost right out of the opening when it was unclear for months what could be done.

Now that seems to be what always happens, eventually a defence is found, until the next dangerous novelty. But not always -- it could just as well have happened that there was no defence, and the Poisoned Pawn would have been refuted. It's not given in the rules of chess that the Najdorf PP is sound, after all.

And it's conceivable (though of course very unlikely) that there are enough such refutations still unfound that one side can force a win. There are always plenty of opening lines where Black is in trouble; who knows.

I think you make it sound far, far too simplistic -- paraphrasing "high level chess games are always more or less equal after the opening, so chess is a draw".

Well it is very unlikely that that kind of refutation (and that assesment often changes) will stand forever in every single opening black plays. Also that "refutation" also has to force a win, which it often doesn't. Lines are abandoned most of the time because white can get a big advantage and black must suffer, but it doesn't mean even those positions couldn't be defended if one played perfectly.

If such a forcing move exists without a refutation it seems likely the Black's blunder lies a number of moves back and that the proper prophylaxis is to avoid that line altogether.  The real question at hand is whether such a forcing move exists on move one.

"You cannot assume that chess is a draw with perfect play just because of your "feelings" or "intuition", or even because of the insubstantial, elementary, barely-scratching-the-surface analysis that has been done in chess so far using computers."

Actually, it's quite reasonable to assume that. Of course the way you describe it all of our studies seem useless. Yes, there are tons and tons and tons of chess positions out there, but the vast majority of them would never happen in games with logical players (because all positions would mean openings like 1 a4 f6 2 Ra2 Na6 3 Ra1). It's not correct just to use math in this case.

Silly me, trying to use logic to describe chess. Everyone knows that chess is more a game of feelings and emotions than logic.

Whatever.

The reason math is so difficult is that all it takes is ONE counter example to completely blow a theorem out of the water. For example, in 1993 Fermat's Last Theorem had been proven for all primes less than 4 million [1]. Why didn't mathematicians just declare it solved and go on their merry way? Because the first 4 million primes are only 0% of all the primes out there (there are infinity). Fermat's Last Theorem still had to be PROVEN. There might have been a prime number in the hundreds of trillions that would have proved it false.

It's the same thing with chess. Even if 99.999999% of all chess games lead to a draw, if there's even a single forced win in that extra 0.000001% it blows your entire draw theory out of the water. There are something like 10^120 possible chess games, and even if we limit that to only the "reasonable" games and ignore all the stupid openings that are obviously crap, we have still barely scratched the surface with our analysis.

Don't you think it's just a little bit arrogant for you to declare that you KNOW the outcome of chess based on a study of 0.0000000001% of the game?

[1] http://en.wikipedia.org/wiki/Fermat%27s_Last_Theorem#Computational_studies

the 2012 thing is pure garbage...

i'm saying that although wehite holds an unquestionable advantage, it is very slight and black should be able to draw.

i tried chessmaster vs. itself and black drawed with a rook behind passed pawn..

if it is like that with chessmaster, it should give the same results with perfect play on both sides..(also should be the same result if we have patzer vs. himself.)

also, if this gets figured out and we realize every game should end in a draw with perfect play, whic master would want o play anymoore? so it's probably best if this remains undiscussed

Ooh!  I was hoping someone would bring up the subject of infinity!

Just because the number of valid moves in a game of Chess is really huge compared to our understanding of known games, and that the number of moves approaches a number that can be compared to the age of the universe, that doesn't mean it even comes close to being compared to infinity.  Infinity is something that swallows up even the biggest numbers we can imagine.

No matter what the sequence of moves in a game might be, as long as move sequences can be regressed to a finite form, there are no real outcomes that result in the game not regressing to one of 3 forms, i.e. winning, losing, or a draw.  And these outcomes, no matter how huge the number of possibilities there are, can be eventually broken down into one of these choices.

Essentially, although we suspect that a logical understanding of the analysis would lead to the possibility of finding a choice where the outcome remains undefined, we know that the bounds of the game are fixed and are not infinite.  Therefore, there exists a solvable solution for every position in the game, which, based on the amount of advantage given to either side at the beginning of the game, as long as the initial position does not result in an instant loss, should guarantee that the side in question doesn't lose the game.

Really, what we're trying to prove on the converse is that the starting position doesn't lead to white winning every game, which is what perfect play refers to (black forces the draw).

No, simply pairing two equally matched opponents together, no matter how strong, gives you no information about a perfect game.

Hey, I love logic, but I hate how you stopped the quote right there, because after that I said that understanding the actual game is much more reliable than coming up with a huge amount of chess positions.

Arrogant? Not really. I mean it takes some understanding of chess to be confident that it is a draw, but you don't have to be a GM or anything. I didn't say I knew, I implied that I feel very strongly that it is a draw however. That .00000000001 % of the game may well be all of the logical games possible, everything else being worthless. Again, math (and I love math by the way) gets you nowhere here because although there are a huge amount of possible positions, that doesn't mean anything about there actually being a position in EVERY SINGLE opening where by force white gets a decisive advantage. In fact there aren't any reasonable openings EVEN CLOSE to being refuted, at all. Even 1...h5 might be close to refuted, but not necessarily refuted. When it comes down to it, those opening positions are all that matter. From our understanding of chess, we know that when the GM's say a position is roughly even, we know it's pretty damn close to even.

WHAT IT WILL REALLY TAKE TO SHOW BLACK IS LOST IS FOR EVERY BLACK OPENING FOR WHITE TO HAVE A WAY TO GET A DECISIVE ADVANTAGE, AND MAYBE IF YOU PLAYED CHESS MORE YOU WOULD SEE HOW EXTREMELY LOW OF A CHANCE THAT IS. BUT ALL YOU'RE DOING IS GUESSING, AND ASSUMING THIS IS THE SAME THING AS ASTRONOMY. ASTORNOMY IS MUCH HARDER BECAUSE UNLIKE A CHESS BOARD YOU DON'T HAVE THE SOLAR SYSTEM SITTING RIGHT IN FRONT OF YOUR FACE.

And I'm not saying chess is solved (because that would be required to be 100% sure), but there is an extremely reliable approximation that white can't force a winning position in all openings black might try. It doesn't take just ".000000000001%", it requires that out of nowhere in every single opening, somehow a move or two from white give him a decisive advantage. Yeah, maybe in an awful opening or even 1 bad move. You know, once you start playing chess more, I think you'll realize how astronomically unlikely it is that white is winning. I don't think you can understand why this is until you actually play chess. You're just trying to use math with no understanding of the reasoning the great players have when saying that "chess is a draw".

I think you just insulted astronomers.