@7541
"1. We have no way of every knowing how long a chess game last we perfect play."
++ We have. We have perfectly played ICCF WC draws and on average they end after 42 moves.
2. We have no way of every knowing is chess is a win, loss, or draw with perfect play.
++ We have. We have perfectly played ICCF WC draws.
A tempo in the initial position is not enough to win.
A tempo cannot queen, a pawn can.
A pawn in the initial position is worth 3 pawns as we know from gambits.
@7541
"1. We have no way of every knowing how long a chess game last we perfect play."
++ We have. We have perfectly played ICCF WC draws and on average they end after 42 moves.
If ICCF players have already solved chess then there's no need to wait 5 years with a supercomputer and Sveshnikov's ghost.
A pawn in the initial position is worth 3 [temi] as we know from gambits.
That's a useful rule of thumb for practical play in opening positions. It's obviously not a rigorous game theoretic value.
Some people try to use forced twins in 549 moves from such positions as an indication that chess isn't drawn with best play; and of course they don't take into account that the position shown is completely unbalanced even though roughly equal in force.
Do they?
I just use it to comment on @tygxc's "proof" that no chess game can last longer than 100 moves.
Doesn't advance a solution of chess much, I admit.
Hoping to cut down on @tygxc's manure rate. Don't know what we can do about yours.
This is absurd.
1. We have no way of every knowing how long a chess game last with perfect play.
2. We have no way of every knowing is chess a win, loss, or draw with perfect play.
We can not use short cuts, guesses, or assumptions.
When you start your argument that chess is a draw or a win. You have already fallen down the rabbit hole.
S'cuse me. I didn't start my argument with any assumption about the theoretical result.
1. We do know how long a chess game can last with legal play under some sets of rules.
Under FIDE basic rules since 2017 or FIDE rules prior to 2017 there is no finite limit, ℵ₀ can be taken as the minimal (unachievable) limit. Under FIDE competition rules since 2017 it's exactly 8849 White moves.
2. We have no way at the moment of knowing if chess is a win, loss, or draw with perfect play. Indeed you'ld need to alter the FIDE laws for the question to even make sense.
That doesn't mean there is no way of ever knowing (but the answer may be different depending on which set of rules is settled on).
The question doesn't arise in my post, because I'm not talking about perfect play, only legal play.
...
Stockfish error rate at 1 second PER GAME on my Threadripper.
GM Magnus Carlsen error rated at Fide Standard Time controls.
Even at 1 second per game. Stockfish is much superior to any human chess players.
Stockfish dev-20221221: 7/4/10/14/1/4/2/2/41/6/3/12/1/17/2/4/3/4/17/5/0/11/1/1/3/26/6/2/2/10/14/2/4/28/9/3/2/3/5/8/2/4/6/5/9/8/4/7/2/0 => Average=0.07
Carlsen Magnus: 2/10/10/21/14/23/23/4/30/22/49/8/4/15/18/35/6/8/16/5/11/4/9/13/14/7/5/12/7/20/12/27/7/14/1/13/8/2/18/7/11/10/10/6/4/10/2/4/13/7 => Average=0.12
If you can tell what the error rates are you must have solved chess.
Congratulations - beat @tygxc by five years + the time he's going to take to con someone out of $3M (which could be about the same as a genuine solution might take).
If you want an idea of how SF15 compares with Rybka/Nalimov - which is not just perfect, but also perfectly accurate from the position below - you could always ask Coach, because SF15 is what he uses to decide.
...
A score of 0.00 does would not mean perfect play, but only perfect compared to the best play we have today from the best chess engine we have today.
Not sure how you're scoring, but I assume it's based on SF15.
IF using SF15 gives perfectly accurate play (as in Black Rybka/Nalimov in the example I posted) a CAPS2 score of 66.9 that might suggest the best play we have today from the best chess engine we have today is actually not very good (while recognising that anything else that plays without tablebases is considerably worse).
And to compare Carlsen to SF, you should ask both Carlsen and SF if you want an unbiased result. If you just ask SF, it's likely to agree with itself whether it's right or not.
Again, that's more appropriate as a remark about yourself. Completely amazing how everything supports your artificial rationale. You really are completely bonkers.
See? Much better. You should follow my advice more often. If people didn't know it was you, this would sound relatively normal.
Chess engines don't uniformly improve necessarily.
I practice KNN v KP quite a lot against engines.
I had a version of Rybka specially strengthened for endgame play about 20 years ago that could reliably do White mates in 50 (though not usually in 50 moves). No version of SF can; the limit is about 36. Also SF12 is considerably weaker at that endgame than SF11 and weaker than SF8.
ICCF players are like lab assistants. They run around between engines gathering results and pick the best one. Occasionally they remind the engines to check a line the engines had been ignoring..."you missed this test tube over here".
And it is unbias. As Stockfish games were not played in retrograde. Or analyzed at 1 second per game in retrograde.
I don't understand that conclusion.
Can you give some more detail. What, exactly, do you mean by, "played/analyzed in retrograde"?
3.56 grams. I suppose it's because they're bronze. That's about half a kilo.
In the interest of actually helping you, I will point out that light hearted withdrawal does work, but if you post 3 times in succession, it belies the nonchalance you are attempting to project. I have some UK coinage left over, I'll swap with you the next time I am going from London to Fort William.
When computers beat consistently the best human players, chess is solved, more difficult for the GO game though...
And it is unbias. As Stockfish games were not played in retrograde. Or analyzed at 1 second per game in retrograde.
I don't understand that conclusion.
Can you give some more detail. What, exactly, do you mean by, "played/analyzed in retrograde"?
Retrograde analysis is a method used to solve game positions for optimal play. By backwards analysis of KNOWN outcomes.
So is constructing tablebases, but so far they don't allow us to analyse complete games and determine where the blunders are or what the blunder rates are, nor even the part of the game with less than eight men if the game repeats any basic rules positions with less than eight men.
What flavour of retrograde analysis did you actually use? What were the KNOWN outcomes?
So you know a mistake has happened in the game. And you work backwards to find the root cause of that mistake.
How do you know in advance that a mistake has happened if you don't know of one already? By, "the root cause of that mistake", do you mean the first blunder? That still wouldn't allow you to fix a blunder rate.
But mainly, what exactly is the procedure you used for finding the root cause?
It seems SF was used in the procedure? Would the same results have been obtained if you had instead used Magnus Carlsen? (Not as cheaply, obviously.)
If not, how can you say that the comparison is unbiased?
No problem. Just describe in detail the steps that were performed to produce your figures for SF and MC, so we can understand what they represent.
Just saying, "retrograde analysis", isn't enough for that. It could lead to a Nalimov tablebase, a Syzygy tablebase or one of Marc B.'s DTC tablebases or the figures you included in your post. I'm interested in what the last are.
When computers beat consistently the best human players, chess is solved, more difficult for the GO game though...
Just as when cars manage to go faster than the best human runners we can say we've cracked warp drive.
When computers beat consistently the best human players, chess is solved, more difficult for the GO game though...
These are the definitions of "solved" being discussed:
https://en.wikipedia.org/wiki/Solved_game
When computers beat consistently the best human players, chess is solved, more difficult for the GO game though...
These are the definitions of "solved" being discussed:
No they're not really.
None of them mention any time period.
A strong solution of a soluble version of chess can be produced in less time than it takes to produce a set of rules for the soluble version. Such solutions have been known since chess was invented.
The Wikipaedia definitions call only for an effective procedure and say nothing about practicality or tractability.
What is being discussed, irrespective of the definitions being quoted by people discussing it, are the definitions I gave here and here. (Excepting @tygxc who is discussing big red telephones and @Optimissed who is discussing his mother in law, the meaning of life and his unfeasibly large brain.)
I'll repeat them.
A (timely) weak solution means that for the initial position either a proven (timely) strategy has been determined for one player that achieves a win for that player against any opposition, or a proven (timely) strategy has been determined for each player that avoids a loss for that player against any opposition.
A (timely) strong solution means that for all legal positions either a proven (timely) strategy has been determined for one player that achieves a win for that player against any opposition, or a proven (timely) strategy has been determined for each player that avoids a loss for that player against any opposition.
In this discussion time is of the essence, so, for example, the Checkers solution which reportedly needs about two minutes to produce a move, is not a solution of 1 minute bullet Checkers and the strong solution of chess I said eatlier could be simply written out would not be a solution of any form of chess that could actually be played, because in all probability nothing that was capable of playing chess could persist long enough or find enough working space to carry out the strategy.
No problem. Just describe in detail the steps that were performed to produce your figures for SF and MC, so we can understand what they represent.
Just saying, "retrograde analysis", isn't enough for that. It could lead to a Nalimov tablebase, a Syzygy tablebase or one of Marc B.'s DTC tablebases or the figures you included in your post. I'm interested in what the last are.
Lets just demonstrate the power of Retrograde analysis.
Use Stockfish and analyze the starting position.
Stockfish is clueless on how to play this positions. And insists on playing Bxh5??
Now plug the end result of the position as it should have been played. And with Stockfish seeing the outcome first. Move backwards slowly allowing Stockfish to cache the position into the hash tables with Stockfish running back to the starting position.
Now Stockfish understands its mistake, and will now play the correct Bc6 instantly.
And now you have a scoring difference of what should have been played vs What was played!
I just have a program that just automates this process for the games, and can now score the games errors.
Sorry. Clear as mud to me.
You apparently decide first of all where MC or SF went wrong in a series of games. How?
You then decide on an optimal path that should have been played. How?
You then run a series of games in SF starting at the end of your new path and working your way back one move at a time through the path for the next game in the series, after which you say SF will correct SF's or MC's mistake.
Does that always work and why?
Does SF then follow your amended path at all points and why?
You say you then have a scoring difference.
What is it?
You show:
Stockfish dev-20221221: 7/4/10/14/1/4/2/2/41/6/3/12/1/17/2/4/3/4/17/5/0/11/1/1/3/26/6/2/2/10/14/2/4/28/9/3/2/3/5/8/2/4/6/5/9/8/4/7/2/0 => Average=0.07
Carlsen Magnus: 2/10/10/21/14/23/23/4/30/22/49/8/4/15/18/35/6/8/16/5/11/4/9/13/14/7/5/12/7/20/12/27/7/14/1/13/8/2/18/7/11/10/10/6/4/10/2/4/13/7 => Average=0.12
but the averages after the double arrow are obviously not averages of the preceding numbers between the slashes. What are the numbers, how do they relate to the average and what is the average an average of?
Some people try to use forced twins in 549 moves from such positions as an indication that chess isn't drawn with best play; and of course they don't take into account that the position shown is completely unbalanced even though roughly equal in force.
Do they?
I just use it to comment on @tygxc's "proof" that no chess game can last longer than 100 moves.
Doesn't advance a solution of chess much, I admit.
Hoping to cut down on @tygxc's manure rate. Don't know what we can do about yours.