Sorry. GMs lose won and drawn games ALL the time. Their propensity for making mistakes will not decrease over the coming years, while the computers skills will continue to climb.
The point is once you're at knight odds, it's the skill of the side with the material advantage that is primarily being tested - not the odds giver. The game is beyond-hope lost and extreme measures need to be taken to muddy the waters.
The way the computer is currently winning odds matches is to:
1) stabilise the position
2) accumulate advantages til nearly equalised
3) crush the human on a tactical blunder
With an extra piece, step 1 is frankly just not possible, it's too much mobile power you're missing. The side with a material advantage is going to be able to find somewhere on the board to just overpower the odds giver if it just tries to maximise resistance.
To win when giving piece odds assuming your opponent is not going to spontaneously hang a whole piece you have to resort to coffeehouse chess - something top engines are not currently designed to do and potentially not something that would work against a top GM even if it was done well (almost certainly a program could be designed along those veins to beat a 2000 rating human though which i would expect to be able to beat a normal engine when receiving knight odds).
I am not convinced it is beyond hope. For example, care to challenge me in a KRN vs KR ending?
bob wrote:You think a GM can play tactically perfect and make ZERO mistakes, even though they can't do that today?
In an easily won position a top GM should certainly be able to play perfectly. I believe that knight odds is sufficient for a position to be "easily won" in this sense.
As I said above, you cannot compare perfect play from the normal start position with perfect play from an easily won position. As Adam noted, knight odds is not real chess. It is a different game. Elo differences in normal chess do not carry over to a different game.
As Kevin put it, knight odds primarily tests the skill of the side with the material advantage. Giving the disadvantaged side 1000x as much time will not help one bit if the side having the advantage is able to win the game against any play.
Btw, Nakamura actually seems to allow for the occasional human blunder on the GM's side, as he only argued that knight odds will always favor the GM.
Someone should take a set of 1,000,000 GM games, and first locate positions where they are a piece ahead, and then analyze to see if they "played perfectly". I remain unconvinced after watching many of the old zonal and inter-zonal events were I watched GMs drop pieces and even queens on occasion. That's not perfect chess. I'm not drinking this "perfect play when a piece ahead" cool-aid without some sort of supporting evidence. "I believe" or "I think" is not evidence. If they can play perfectly when a piece ahead, they can play perfectly when material is even.
syzygy wrote:
In an easily won position a top GM should certainly be able to play perfectly. I believe that knight odds is sufficient for a position to be "easily won" in this sense.
if a computer playing a top gm at knight odds can get the score down from 3 to 2 before the super gm can get it up from 3 to 4, the computer will win. currently a computer cannot do that. how do you know it will never be able to do that. ?
Engine scores are not physical measurements that can be forced up or down simply by throwing enough computer cycles against them.
If a position is won, you don't even have to play any moves to get the score up (from you point of view). Just let the opponent computer think long enough and its score will eventually drop to mate. Making the computer stronger will only make it realise more quickly how hopelessly lost it is.
So if in a particular position the computer today cannot get the score down from 3 to 2, then a future computer in that same position will not be able to get the score down from 23 to 22.
bob wrote:Someone should take a set of 1,000,000 GM games, and first locate positions where they are a piece ahead, and then analyze to see if they "played perfectly". I remain unconvinced after watching many of the old zonal and inter-zonal events were I watched GMs drop pieces and even queens on occasion. That's not perfect chess. I'm not drinking this "perfect play when a piece ahead" cool-aid without some sort of supporting evidence. "I believe" or "I think" is not evidence. If they can play perfectly when a piece ahead, they can play perfectly when material is even.
I'm fairly certain even you think a GM can win against a king and pawn with a full set of pieces, so somewhere there is a hard limit on what kind of odds a computer can give and still have a chance. Hikaru, and many of us, think that knight odds is beyond the limit. You apparently think it may be substantially more. Let's agree to disagree.
syzygy wrote:
Engine scores are not physical measurements that can be forced up or down simply by throwing enough computer cycles against them.
If a position is won, you don't even have to play any moves to get the score up (from you point of view). Just let the opponent computer think long enough and its score will eventually drop to mate. Making the computer stronger will only make it realise more quickly how hopelessly lost it is.
So if in a particular position the computer today cannot get the score down from 3 to 2, then a future computer in that same position will not be able to get the score down from 23 to 22.
a computer is not told to evaluate the truth of a chess position and resign when hopeless. It is or can be told to equalise or at least minimise the difference where possible by relying on human error. the computer of the future will find many more human errors than the computer of today. hence many more opportunities to equalise. these many more opportunities many not be enough to overcome a knight in the hands of a super gm, but otoh it may be sufficient.
bob wrote:I'll say it again. In 1970, A human could beat a computer giving the COMPUTER knight odds, trivially. That has changed. Just because they can't do it today doesn't mean they can't do it next year, or next decade, or next century, or next millennium. Just means they can't today. We won't have 32 piece EGTBs ever, as far as that goes.
I'll say it again as well. A 1000+ Elo advantage does not allow you to find a winning or drawing move in a position that is lost. Not today, not tomorrow, not ever.
The only real question is whether, say, Carlsen is good enough at chess that he is able to convert, against any play, an easily won position into an actual checkmate position. Since we don't actually know the answer in case of knight odds, this is a matter of opinion. But your analogy with predictions from 1970 is flawed.
Btw, 32 piece EGTBs are of no help here. They'll just show that a lost position is a lost position and that all moves from a lost position lead to equally lost positions.
Sorry. GMs lose won and drawn games ALL the time. Their propensity for making mistakes will not decrease over the coming years, while the computers skills will continue to climb.
Only your understanding of the analogy is flawed. In 1970 computers couldn't convert easily won positions to wins. Not so today. To believe there is some max handicap beyond which a GM will never lose suggests there is some max Elo difference (between GM and best computers) that can't be exceeded. Do you have ANY empirical evidence to support such a claim? I've not seen any. I don't believe a human will be able to win ANY games should 32 piece EGTBs ever exist, assuming the ultimate game result is a draw. I don't believe they will be able to draw a single game against such a player, even though theory says the game is drawn. You think a GM can play tactically perfect and make ZERO mistakes, even though they can't do that today?
I think that is a stretch. I won't begin to guess when this might happen, but I believe it is a certainty that it will happen.
You do not need to play zero mistakes to win against the computer with knight odds.
zero mistakes means that you always choose the fastest mate but you do not need to choose the fastest mate in every move in order to win.
There are humans who know how to win KR v K but do not do it in the fastest way.
They will be able to beat chess engines always from KR vs K inspite of not playing perfect chess.
syzygy wrote:
Engine scores are not physical measurements that can be forced up or down simply by throwing enough computer cycles against them.
If a position is won, you don't even have to play any moves to get the score up (from you point of view). Just let the opponent computer think long enough and its score will eventually drop to mate. Making the computer stronger will only make it realise more quickly how hopelessly lost it is.
So if in a particular position the computer today cannot get the score down from 3 to 2, then a future computer in that same position will not be able to get the score down from 23 to 22.
a computer is not told to evaluate the truth of a chess position and resign when hopeless. It is or can be told to equalise or at least minimise the difference where possible by relying on human error. the computer of the future will find many more human errors than the computer of today. hence many more opportunities to equalise. these many more opportunities many not be enough to overcome a knight in the hands of a super gm, but otoh it may be sufficient.
I'm not sure about the use of "many" above. When a human player fails to play the best move (which itself is hard to define when there is more than one optimal move), I think the odds are pretty good that today's Komodo will identify the error. There is room for improvement, but maybe not so much.
I would look at the problem this way. Let's say that there is some handicap that would produce an even score in a serious match between Komodo and Magnus Carlsen. I would estimate that this handicap would be in the 1 to 1.5 pawn range based on Komodo's eval after a long think. Let's say 1.25. Now both Komodo and Carlsen make errors of some average magnitude. We'll call Carlsen's error rate C, and Komodo's K. I think it's pretty obvious that K is much less than C, let's say K = .4xC. If a future engine drops the error rate to zero, then C - K increaases to 5/3 of it's former value, so the proper handicap should also increase in that ratio. That would put it at 208, a bit over two pawns but way below the roughly 3.5 value of knight odds. Of course there is a lot of uncertainty in the above, but I don't think the estimate would be way off. In order for the estimate to reach knight odds, the K/C ratio would have to be about 64%, which does not seem plausible to me. Probably this ratio can be estimated more precisely by comparing the error rates of Komodo with some engine about the level of Carlsen (maybe the pocket Hiarcs version that earned a 2900+ performance in 2009) as measured by the moves chosen by a "referee" such as Stockfish with 100s as much time. But that's a big project.
Uri Blass wrote:You do not need to play zero mistakes to win against the computer with knight odds.
zero mistakes means that you always choose the fastest mate but you do not need to choose the fastest mate in every move in order to win.
There are humans who know how to win KR v K but do not do it in the fastest way.
They will be able to beat chess engines always from KR vs K inspite of not playing perfect chess.
I don't think "fastest mate" is very important to playing chess. So I would define "perfect chess" simply as playing moves that preserve the theoretical outcome of the position.
With this definition, playing KRK perfectly is trivial and within the reach of any chess player.
Playing knight odds perfectly is far more difficult, but might well be within the reach of top GMs. Obviously any human will make some errors in the long run, but in a serious match a top GM seems rather likely to come out on top when given knight odds.
Playing the opening position perfectly with any consistency is much and much harder.
Laskos wrote:Actually Naka seems right to me for GMs above ELO 2700, at least against simple 32 men tablebases and aside occasional obvious large blunders by GM.
It probably depends on the time control. At a one minute game even Hikaru would probably lose today. At game in two hours plus increment it will probably never happen.
Yes, I am talking here always only about 2 hour normal time control.
It may be possible that the GM extremely does not feel well during the game so in a single game everything is possible but I believe that no computer will get even a fide rating of 2400 in the future against humans if fide allow computers to play without knight b1 for rating(time control should be at least 90+30).
