True or False Chess is a Draw with Best Play from Both Sides

Sort:
ponz111

JimDiesel of course I knew the ratings were not related per your chart. But very probably many people looking at your chart did not know ratings were not related and you should have explained this to the people who did not know.

Also. I know that you did not believe stockfish vs alpha zero drew 15% of their games. You simply made a one word mistake and then changed that one word. It would have been simpler and more ethical just to mention you made that one word mistake.  This would have also been better if you had responded to my question about this situation.  Dodging a simple question is not really good practice.

 

Thee_Ghostess_Lola

but he can do that cuz he's free. he's free ponzie. say it with me. FREE. isnt that beautiful ?

and as much as i like u ?...this is amercia. and its way-WAY more important to be free than right.

FilipinaAdventures

True

Ziryab
JimDiesel22 wrote:

btw @IfPatriotGames

This is white's win rate plotted against rating according to various databases.

 

Who compiled this data. I'd like to read the footnotes.

ArthurEZiegler

ponz111 now has changed my original thought that due to it's complexity you cannot give odds if chess is a draw on not. I have not read the 3000 plus posts looking for evidence, but I accept that skilled correspondence players can believe it is highly likely to be a forced draw based on their experience of the overwhelming preponderance of draws in high level games. However, my certainty of chess being a draw is more now like 90%, not the 99.999% touted by ponz111. It is a fact that all the experience of all the games ever played by both man and machine is insignificant compared to the beyond astronomical number of possible variations. That leaves a lot of space for winning lines of play not envisioned by either man nor machine.

I also want to say to those who think the small opening move advantage of white could be that which if chess is solved produces a win for white. Maybe so, but black seems to have opportunities to neutralize that advantage during the course of a game. It is also  possible chess is a forced win for black. Unlike the Chinese game of GO in chess you must make a move when it is your turn and perhaps white would be forced to make one that produces an inferior position leading to a loss. I don't think you can estimate the odds if this is so or not. Any mathematical experts in game theory want to chime in on this?

ponz111

Arthur  I am 99.9999% sure chess is a draw based on the evidence I have seen.  You are about 90% sure based on the evidence you have seen.  I have seen a lot more evidence than you have seen.

I don't know if you would increase your % or not if you had seen all the evidence and if you had played [and beat] all the masters and grand masters I have played/beat?  

Probably if you were a stronger player you would consider it as no chance that Black can ever force a win with best play on both sides.

 

pfren
MeleiroZ έγραψε:

Checkers is solved, Chess, more complex, is a matter of time.

 

Official checkers championships are played on 10*10 board, and to my knowledge so far only the simpler 8*8 variant has been solved.

Chess will never be mathematically solved, as the universe does not have enough atoms to store the data of the solution.

ponz111

Agree chess will never be solved per math.

 

 

ArthurEZiegler

You should be glad to have modified my position on this, but it is not so much the evidence but some thinking about the game itself. I try to think like a mathematician. Perhaps with a board of just 64 squares and the set moves of the pieces it may be that no combination of moves will produce a forced win. It may be that a slight advantage may be gained but the game does not go on long enough for that advantage to produce a win. Maybe these are the insights that lead correspondence players like yourself to consider chess to ultimately be a drawn game.

ArthurEZiegler

pfren - If for example a game comes down to two kings and a rook it is not needed to exam every possible move variation to win, just an understanding of how to use the combination of the king and rook to force your opponent to the edge of the board for the mate. Could not a computer eventually understand how to produce a forced win without looking at every variation?

ponz111

Arthur  Computers do that now [I think] The computers know the result of all chess positions of 6 pieces [or is it 7?]  So king  vs king and rook is known for all positions. 

The trick is to get the 6 or 7 piece endgame up to 8 and then 9 but as more pieces are added--it is quite certain computers will never get to 32 piece starting position.

ponz111

Arthur  you mentioned a couple of possible insights correspondence players might have re chess but there are a lot more than the 2 you mentioned.   

lfPatriotGames
MeleiroZ wrote:

Checkers is solved, Chess, more complex, is a matter of time.

Anyway, we don't need to be scientifically sure of many things in life, otherwise, for example, no one would be convicted in court.
There is no way, in any opening, for White to obtain a decisive advantage. The increasing number of high-level correspondence chess draws indicates that. There is an ICCF tournament that has one single win in 196 games.
I don't understand this stubbornness when everything indicates that chess, perfectly played by both parties is a draw. I would like to know which opening gives White a decisive advantage. That does not exist.
This, however, does not matter in low-medium level chess. Correspondence chess is different: I doubt that this variant has a short-medium term future.

I dont think comparing solving chess to convicting someone in court is a good example. Chess is a game. More like a puzzle or math problem that can be solved using numbers. Court convictions dont require that. They just need enough evidence.

Using the court example, there have probably been times someone was convicted, so it was "true" that they committed a crime. It was considered fact. But then later new evidence emerged that changed everything. So what was "true" is now false. Right now we could say there is a preponderance of evidence that chess is a draw. But that leaves a LOT of room for it to be a forced win. I agree we'll probably never know for sure because computers will never be able to calculate that much. But what could happen is 100 years from now or so some advanced intelligence could stumble upon a forced win without even coming close to exploring all the possibilities. This still wouldn't be proof because even if every single game that intelligence played was won, there would still be the possibility the forced draw was not discovered yet. 

I imagine what would happen is the exact same discussion. There would be those who say they are 99.99999% sure chess is a forced win, based on the winning streak of that intelligence. And then there would be others that would say they just dont know because the intelligence hadn't explored every possibility where a forced draw might be possible. 

Prometheus_Fuschs
pfren escribió:
MeleiroZ έγραψε:

Checkers is solved, Chess, more complex, is a matter of time.

 

Official checkers championships are played on 10*10 board, and to my knowledge so far only the simpler 8*8 variant has been solved.

Chess will never be mathematically solved, as the universe does not have enough atoms to store the data of the solution.

Then you don't know what "solve" means in this context.

ponz111

Prometheus   Am curious?  Do you think chess will be math solved in the next million years?

and what is your definition of math solved?

ponz111

PATRIOT  I do NOT assume that because someone has been convicted in court that it means that person did the crime.  I certainly would need a lot more evidence than the mere fact that the person has been convicted in court. I certainly would NOT consider it a "fact" that the person did the crime. 

The analogy I gave several times provided a whole lot of evidence that [in the analogy] the person [brother] actually did the crime.

I am from the USA and know our court systems are greatly flawed and this means I would Never believe that someone did a crime with just the evidence that he had been convicted. 

Also, I would NOT say that there is preponderance of evidence that chess is a draw. Preponderance of evidence is not enough.  

Also what is true cannot just become false per opinion. "True: or "false" requires [for me] a lot more than just someones opinion.  In fact if 8 billion people think something is true--that does not make it true.  If 8 billion people think something is false--that does not make it false.

It does not matter to me if someone is 99.9999% sure that chess is a win.  Opinion only makes a difference depending on who holds the opinion.  For example you and I probably trust our doctor to give good medical opinions  We may get additional opinions from other doctors But we in general trust the opinions of doctors as they have studied in their profession for years AND Also doctors tend to agree with each other.

If some intelligent person says chess is a win--we would demand evidence of his claim.

 

 

 

Prometheus_Fuschs
ponz111 escribió:

Prometheus   Am curious?  Do you think chess will be math solved in the next million years?

and what is your definition of math solved?

When it is said checkers is solved, it means the game state of the initial position of the game is known. You do not need to store all of the other game states nor the game tree, let alone an algorithm that can play perfectly (we already discussed what it means to play perfectly).

Prometheus_Fuschs

For more information read the wiki article, it even explicitly says a non-constructive proof is valid.

https://en.wikipedia.org/wiki/Solved_game

Eden013

I have never seen such a useless forum...you have one guy that repeats the same damn thing 3000 times and others who are pointing flaws in his arguments (if he even has any) in which he ignores and proceeds to repeat the same generic response. I already know exactly what this dude's gonna say.

Thee_Ghostess_Lola
GK feels this may be his best game ever. probably cuz AK never made a defined inaccuracy.