Based on what games? Bullet and blitz games don't count as humans can easily make very stupid mistakes when short of time. We need a minimum of one minute per move to avoid the worst thoughts we can imagine. A minus knight removes so many resources for the handicapper. It frees the human mind of troubles.bob wrote: But MY guess, based on looking at lots of GM games, is that a knight won't be enough against perfect play assuming that a GM is playing the stronger side. Perfect play might well show a knight wins. But can a human complete the game with zero mistakes? Seems beyond implausible based on existing games...
which would take more moves to win
Moderators: hgm, Rebel, chrisw
-
- Posts: 133
- Joined: Fri Apr 09, 2010 3:26 am
Re: which would take more moves to win
-
- Posts: 20943
- Joined: Mon Feb 27, 2006 7:30 pm
- Location: Birmingham, AL
Re: which would take more moves to win
You have an easy method to detect a quark? AND store information using it? And addressing it in a 3d space?duncan wrote:if there are more quarks in a proton than atoms in the universe and they can be manipulated and energy supply is not an issue, then maybe chess can be solved as early as this or the next century.bob wrote: I don't believe chess will EVER be solved. Just look at the size of the necessary EGTB and figure out a rational way to store it in a device the size of the moon, and then have a way to access any atomic particle quickly and accurately...
Such is not even in physics discussions anywhere at all, no matter how far one is extrapolating into the future (at the moment).
How many atoms do we use to store a bit today? How many atoms for the pathway from that storage unit to the detector?
-
- Posts: 20943
- Joined: Mon Feb 27, 2006 7:30 pm
- Location: Birmingham, AL
Re: which would take more moves to win
At TOURNAMENT games. I watched Topolov, for example, blunder a queen to a move I saw instantly. After making the blunder, the next move he had to give up his queen or else be mated on the next move. Use a computer to go through GM games, give them an hour per move, which is hardly up to the standard we call "perfect" and see how many mistakes are made. Usually mistakes that the opponent fail to take advantage of because they didn't see the punch line either...EroSennin wrote:Based on what games? Bullet and blitz games don't count as humans can easily make very stupid mistakes when short of time. We need a minimum of one minute per move to avoid the worst thoughts we can imagine. A minus knight removes so many resources for the handicapper. It frees the human mind of troubles.bob wrote: But MY guess, based on looking at lots of GM games, is that a knight won't be enough against perfect play assuming that a GM is playing the stronger side. Perfect play might well show a knight wins. But can a human complete the game with zero mistakes? Seems beyond implausible based on existing games...
As it stands, I have no idea whether a knight is a forced win or not given perfect play by both sides. But even without knowing that, I know how prone humans are to make mistakes so that I consider the chances of a GM winning with a knight handicap to be far worse than for a perfect player winning. How much worse is "far worse"? No way to know at the moment since there is nothing even remotely approaching perfect play if we go beyond 7 pieces left on the board, total.
-
- Posts: 5566
- Joined: Tue Feb 28, 2012 11:56 pm
Re: which would take more moves to win
The correct answer is definitely a). The extra knight does not come out only at the end of the game but is there from the start and that advantage can be wielded from the start to gain further advantages. Regarding b), no defense will ever save a lost position against perfect play.S.Taylor wrote:This is very interersting. And the question is:
a).Would it be a win for the knight up (with perfect play), because it can bring more power to bear down on the opponent and can always use to win.
OR
b).Since it will be perfectly defended, it will be a draw, and all the other activity aroumd it during the game, will be defendable, even with the extra knight bearing down on it.
And it is.But with humans, we have always understood that a piece up, if before the very end, is only a matter of technique to win.
No, it will not be different if the losing side plays "perfectly".Would this be different with perfect play?
We already see it today with top computer games (e.g. in TCEC), that it is often often a draw.
Obviously blunders by the GM will be punished harder, but if the GM plays carefully and avoids unnecessary risks, he should be able to avoid blunders. (And just forget what Bob will respond to this.)
In a lost position, all moves are losing. So in a lost position, perfect play won't magically help the engine find a drawing move.OR is this because computers still have a long way to go, and then, when perfect one day, they would always find that way to win.
-
- Posts: 5566
- Joined: Tue Feb 28, 2012 11:56 pm
Re: which would take more moves to win
Oh my... I hope you are not serious...bob wrote:The ORIGINAL discussion was GM + knight vs perfect computer, if you care to look. _I_ didn't change the subject.
This thread:
http://talkchess.com/forum/viewtopic.ph ... 227#659227
Title: "which would take more moves to win"
Opening post:
(Emphasis added)duncan wrote:a top gm playing komodo with knight handicap
or 2 computers one with a knight handicap but both with access to a 32 piece tablebase
and how many moves is the former ?
-
- Posts: 20943
- Joined: Mon Feb 27, 2006 7:30 pm
- Location: Birmingham, AL
Re: which would take more moves to win
Of course, NOTHING has proven that a knight handicap is losing. That's merely an assumption.syzygy wrote:The correct answer is definitely a). The extra knight does not come out only at the end of the game but is there from the start and that advantage can be wielded from the start to gain further advantages. Regarding b), no defense will ever save a lost position against perfect play.S.Taylor wrote:This is very interersting. And the question is:
a).Would it be a win for the knight up (with perfect play), because it can bring more power to bear down on the opponent and can always use to win.
OR
b).Since it will be perfectly defended, it will be a draw, and all the other activity aroumd it during the game, will be defendable, even with the extra knight bearing down on it.
And it is.But with humans, we have always understood that a piece up, if before the very end, is only a matter of technique to win.
No, it will not be different if the losing side plays "perfectly".Would this be different with perfect play?
We already see it today with top computer games (e.g. in TCEC), that it is often often a draw.
Obviously blunders by the GM will be punished harder, but if the GM plays carefully and avoids unnecessary risks, he should be able to avoid blunders. (And just forget what Bob will respond to this.)
In a lost position, all moves are losing. So in a lost position, perfect play won't magically help the engine find a drawing move.OR is this because computers still have a long way to go, and then, when perfect one day, they would always find that way to win.
So much better to forget what Ronald is saying since it is 100% speculation even though he proclaims it to be an actual fact. However, conjecture != fact.
-
- Posts: 20943
- Joined: Mon Feb 27, 2006 7:30 pm
- Location: Birmingham, AL
Re: which would take more moves to win
Please learn to read. (a) and then he asked a specific question about (a), not (b). However, BOTH "answers" are unknown and only guessed at. That has not changed regardless of your nonsense claims with no evidence to support them.syzygy wrote:Oh my... I hope you are not serious...bob wrote:The ORIGINAL discussion was GM + knight vs perfect computer, if you care to look. _I_ didn't change the subject.
This thread:
http://talkchess.com/forum/viewtopic.ph ... 227#659227
Title: "which would take more moves to win"
Opening post:(Emphasis added)duncan wrote:a top gm playing komodo with knight handicap
or 2 computers one with a knight handicap but both with access to a 32 piece tablebase
and how many moves is the former ?
-
- Posts: 133
- Joined: Fri Apr 09, 2010 3:26 am
Re: which would take more moves to win
I guess humans would blunder now and then but it wouldn't change the score completely. The more difficult the position is, the easier it is to make a mistake. A knight off helps us an extreme amount. From the start humans could play in a totally different way than normal. No need to fight for the initiative or to defend accurately. Give a pawn to trade some pieces and maybe another one too to get an endgame. Then just play for a win with a basically guaranteed draw in the worst case.bob wrote:At TOURNAMENT games. I watched Topolov, for example, blunder a queen to a move I saw instantly. After making the blunder, the next move he had to give up his queen or else be mated on the next move. Use a computer to go through GM games, give them an hour per move, which is hardly up to the standard we call "perfect" and see how many mistakes are made. Usually mistakes that the opponent fail to take advantage of because they didn't see the punch line either...EroSennin wrote:Based on what games? Bullet and blitz games don't count as humans can easily make very stupid mistakes when short of time. We need a minimum of one minute per move to avoid the worst thoughts we can imagine. A minus knight removes so many resources for the handicapper. It frees the human mind of troubles.bob wrote: But MY guess, based on looking at lots of GM games, is that a knight won't be enough against perfect play assuming that a GM is playing the stronger side. Perfect play might well show a knight wins. But can a human complete the game with zero mistakes? Seems beyond implausible based on existing games...
As it stands, I have no idea whether a knight is a forced win or not given perfect play by both sides. But even without knowing that, I know how prone humans are to make mistakes so that I consider the chances of a GM winning with a knight handicap to be far worse than for a perfect player winning. How much worse is "far worse"? No way to know at the moment since there is nothing even remotely approaching perfect play if we go beyond 7 pieces left on the board, total.
-
- Posts: 5566
- Joined: Tue Feb 28, 2012 11:56 pm
Re: which would take more moves to win
I understand your difficulties admitting you lost track of the discussion. Or for that matter, admitting anything.bob wrote:Please learn to read. (a) and then he asked a specific question about (a), not (b). However, BOTH "answers" are unknown and only guessed at. That has not changed regardless of your nonsense claims with no evidence to support them.syzygy wrote:Oh my... I hope you are not serious...bob wrote:The ORIGINAL discussion was GM + knight vs perfect computer, if you care to look. _I_ didn't change the subject.
This thread:
http://talkchess.com/forum/viewtopic.ph ... 227#659227
Title: "which would take more moves to win"
Opening post:(Emphasis added)duncan wrote:a top gm playing komodo with knight handicap
or 2 computers one with a knight handicap but both with access to a 32 piece tablebase
and how many moves is the former ?
Well actually, no I do not understand it. That part of your psychology is certainly not part of mine. I have a sense of honour.
-
- Posts: 20943
- Joined: Mon Feb 27, 2006 7:30 pm
- Location: Birmingham, AL
Re: which would take more moves to win
But not any "common sense" to see the difference between conjecture and fact. I'd think that is an absolute prerequisite for being a lawyer.syzygy wrote:I understand your difficulties admitting you lost track of the discussion. Or for that matter, admitting anything.bob wrote:Please learn to read. (a) and then he asked a specific question about (a), not (b). However, BOTH "answers" are unknown and only guessed at. That has not changed regardless of your nonsense claims with no evidence to support them.syzygy wrote:Oh my... I hope you are not serious...bob wrote:The ORIGINAL discussion was GM + knight vs perfect computer, if you care to look. _I_ didn't change the subject.
This thread:
http://talkchess.com/forum/viewtopic.ph ... 227#659227
Title: "which would take more moves to win"
Opening post:(Emphasis added)duncan wrote:a top gm playing komodo with knight handicap
or 2 computers one with a knight handicap but both with access to a 32 piece tablebase
and how many moves is the former ?
Well actually, no I do not understand it. That part of your psychology is certainly not part of mine. I have a sense of honour.