True or false? Chess will never be solved! why?


what is "solved"? Chess isn't a crossword... and no I've not read ther thread start either... maybe I really should check it out, but meantime...
Don't bump a year and half old thread (since the last post) if you're not even going to read any of it. Was there a point in it?
Indeed

I'm going to say False. I believe as chess search engines keep getting faster and looking deeper eventually they will mathematically find the "best" reply to every possible move (They will easily cast away nonsensical or "bad" moves) and every game between 2 search engines will follow this same "perfect game" course.
The late GM Sveshnikov wrote:
“Give me five years, good assistants and modern computers, and I will trace all variations from the opening towards tablebases and 'close' chess.”
From John Tromp research:
Number of chess positions possible with 9 chess sets:
N9 = 8726713169886222032347729969256422370854716254
Number of chess positions with 26 men possible with 9 chess sets:
N926 = 162649677964723028114590870024481216875621152
Number of chess positions possible with 1 chess set:
N1 = 19201527561695835455154058755564594798074
Fraction of legal positions: FL = 0.052
Hence: Number of legal chess positions with 26 men possible with 1 chess set
NL126 = FL * N1 * N926 / N9 = 1.9 * 10^37
To proof checkers is a draw required to look only at roughly the square root of the number of possible positions.
Hence: Number of positions needed to solve chess from 26 men starting positions: 4.3 * 10^18.
The Sesse computer evaluates 49582886 nodes per second.
Hence time on Sesse to solve chess from 26 men starting positions: 2757 year
If we split the task in 500 subtasks that each look at only 1 of the 500 ECO codes:
Time to solve 1 ECO code e.g. C67 on Sesse: 5.5 year
So Sveshnikov was right: it takes 5.5 years to solve 1 ECO code of 26 men on Sesse.
It would require 500 Sesse computers and 5.5 years to fully solve chess.
Human assistants should guide the search to start each ECO code from a 26 men position.

The late GM Sveshnikov wrote:
“Give me five years, good assistants and modern computers, and I will trace all variations from the opening towards tablebases and 'close' chess.”
From John Tromp research:
Number of chess positions possible with 9 chess sets:
N9 = 8726713169886222032347729969256422370854716254
Number of chess positions with 26 men possible with 9 chess sets:
N926 = 162649677964723028114590870024481216875621152
Number of chess positions possible with 1 chess set:
N1 = 19201527561695835455154058755564594798074
Fraction of legal positions: FL = 0.052
Hence: Number of legal chess positions with 26 men possible with 1 chess set
NL126 = FL * N1 * N926 / N9 = 1.9 * 10^37
To proof checkers is a draw required to look only at roughly the square root of the number of possible positions.
Hence: Number of positions needed to solve chess from 26 men starting positions: 4.3 * 10^18.
The Sesse computer evaluates 49582886 nodes per second.
Hence time on Sesse to solve chess from 26 men starting positions: 2757 year
If we split the task in 500 subtasks that each look at only 1 of the 500 ECO codes:
Time to solve 1 ECO code e.g. C67 on Sesse: 5.5 year
So Sveshnikov was right: it takes 5.5 years to solve 1 ECO code of 26 men on Sesse.
It would require 500 Sesse computers and 5.5 years to fully solve chess.
Human assistants should guide the search to start each ECO code from a 26 men position.
Your errors in your logical leaps bolded above.
1. You cannot extrapolate chess' exact solve time from checkers....so 10^18 does not follow.
2. You cannot equate Sesse evaluation of nodes/second from engine valuations with solving chess, either. Sesse evaluates positions using standard engine valuations, which is not at all the same as building a tablebase. It's a completely different calculation.
In both of these cases you are trying to put square pegs into round holes.
#1146
1. It is clear that the number of positions necessary to solve chess is far less than the number of possible positions. E.g. in solving C67 we do not have to look at any positions with a white pawn at e2 or d2, a black pawn at e7 or d7, a white light square bishop... To solve checkers Schaeffer only had to look at about the square root of the number of possible positions. If for chess it is also the square root, or more, or less, we do not know until it is done. We will know more once 1 ECO code e.g. C67 is solved.
2. Solving chess is not building a 32 men table base; it is calculating from the opening towards a table base as Sveshnikov said and as Schaeffer did for checkers. Schaeffer solved checkers with a program Chinook, thus chess likewise can be solved with Sesse or equivalent. It is not necessary to look at all possibilities: the top 4 moves for white and the top 1 move for black are enough. Evaluation only guides the calculation: the real evaluation is when the calculation reaches the table base, as Sveshnikov wrote.

The editor removed it because he “didn’t think it was of general interest to the beginning chess player”.

#1146
1. It is clear that the number of positions necessary to solve chess is far less than the number of possible positions. E.g. in solving C67 we do not have to look at any positions with a white pawn at e2 or d2, a black pawn at e7 or d7, a white light square bishop... To solve checkers Schaeffer only had to look at about the square root of the number of possible positions. If for chess it is also the square root, or more, or less, we do not know until it is done. We will know more once 1 ECO code e.g. C67 is solved.
2. Solving chess is not building a 32 men table base; it is calculating from the opening towards a table base as Sveshnikov said and as was done for checkers. Checkers was solved with a program Chinook, thus chess likewise can be solved with Sesse or equivalent. It is not necessary to look at all possibilities: the top 4 moves for white and the top 1 move for black are enough. Evaluation only guides the calculation: the real evaluation is when the calculation reaches the table base, as Sveshnikov wrote.
Same logical error. Chinook != Sesse. You can add all the "thus", "hence", and "ergos" you like, but you are not comparing like to like at several stages, and your premise is flawed.
I guess we can revisit this in 2026-2027, when we can all have a good laugh at your dividing ECOs plan .

The editor removed it because he “didn’t think it was of general interest to the beginning chess player”.
...and here I thought nobody could top the level of crazy. Congrats.
#1149
You can say nay all day and ridicule as you want for lack of real arguments.
Chinook is a program that plays checkers just like Sesse plays chess.
The method is just the same, only chess is more complicated than checkers.

#1149
You can say nay all day and ridicule as you want for lack of real arguments.
Chinook is a program that plays checkers just like Sesse plays chess.
The method is just the same, only chess is more complicated than checkers.
Five years...laughing. You cannot recognize real or fake arguments, as you have demonstrated with your "apples are cherries, hence any pie we make with either will taste the same" type of logic.
For all your posturing on every solving chess thread you can find, you are no closer to anything now than you were when you first mistakenly posited 10^20 would be enough to solve chess. You're still 20 orders of magnitude short.
#1152
All you do is ridicule and make ridiculous comparisons.
10^20 was too much, by starting from a 26 men tabiya 10^18 is enough.
After looking at facts and figures the statement by GM Sveshnikov seems plausible indeed.
The real practical problem is to find funding for a 5 year research program that allocates a supercomputer full time. Maybe the Lomonosov University will do it.

#1152
All you do is ridicule and make ridiculous comparisons.
10^20 was too much, by starting from a 26 men tabiya 10^18 is enough.
After looking at facts and figures the statement by GM Sveshnikov seems plausible indeed.
The real practical problem is to find funding for a 5 year research program that allocates a supercomputer full time. Maybe the Lomonosov University will do it.
There's nothing left but ridicule. The many posters that have tried to show you the errors of your ways (including the Pfrens, the BlueEmus, etc.) lo these many months have not altered your nonsense one iota. Feel free to keep posting it if you like, but get ready to have it debunked over and over (certainly whenever I see it, but also by others when I don't).
You're like that guy who used to post about his ABS System that would defeat any engine, or War-whatever posting his diatribes...just periodic hyperbole and bad logic applied. Enjoy the rest of your weekend.
#1154
I side with GM Sveshnikov. Let the many posters speak for themselves. I did not write about cherries, pies, or ABS or other nonsense like you. You do not even present any argument, only ridicule.

#1154
I side with GM Sveshnikov. Let the many posters speak for themselves. I did not write about cherries, pies, or ABS or other nonsense like you. You do not even present any argument, only ridicule.
No, you posted about Sveshniknov after the fact. But sure, side with the dead GM.
I don't have to post any numbers anymore, you've already been refuted several times by myself and others, and have done absolutely nothing to address the shortfalls in your premise(s).
#1157
Neither you nor any other has refuted anything, you have only ridiculed and made nonsense comparisons to pies and cherries and whatnot. Laughing is no argument, it is lack of argument. You have never posted any credible numbers. Schaeffer has solved checkers. Chess is more complex: more squares, more men, more complex rules, but the same method is applicable.
what is "solved"? Chess isn't a crossword... and no I've not read ther thread start either... maybe I really should check it out, but meantime...
Don't bump a year and half old thread (since the last post) if you're not even going to read any of it. Was there a point in it?