Chess will never be solved, here's why

#2789
Fermat's Last Theorem and the Four Color Theorem were also known to be true before these were proven. 'Provability is a higher degree of truth" that is a phrase I got from an article in Scientific American.
Fischer said "chess is a draw". Kramnik, Adorjan, Spassky, Capablanca, Lasker, Steinitz all said the same thing with more words. Adding a few weasel words 'probably', 'likely' does not make it more true. It is based on centuries of competitive chess and millions of games and many lifetimes of analysis.
People who thought Losing Chess to be a draw did not spent nearly as much time and effort to Losing Chess as has been spent on Chess.

Then post the references you used for both those statements, so we can examine them.

It is called "anecdotal evidence". It may be good for a court of law, not for mathematics.

Indeed, it makes it less true.

Are you using SF8 there? @tygxc apparently plans to use SF14 which can give significantly different numbers.

Be that as it may, what does it give in this similar example?

I use a recent devel version from github, much newer than SF 14.

Crystal is the latest build (from last December) and has nnue enabled.

I do not expect any of these engines to "solve" the above any time soon, as no antifortress code can handle this- white's king has quite a few similarly "bad" choices, while in the first example white's moves are absolutely forced.

I used the chess.com online analysis tool  on maximum depth analysis - try it yourself.  chess.com configured it, not me.  I was using the default configuration of Stockfish which (surprisingly) is not NNUE. I thought that has been the only version in recent years.

It would be interesting to determine the reason for any discrepancy.

Here is a screen grab of an earlier stage of a similar use of the analysis tool. Note that the number -2.71 is a lot less than the -9.xx in the earlier analysis -  after hours of analysis. This number only appeared later in the analysis. Don't ask me why.

Chess.com's Stockfish 14.1 NNUE gives a lower evaluation initially, but again it might change later:

The chess dot com cloud engine isn't suitable for any sort of serious analysis. Even a recent stockfish on your mobile phone would give more reliable output.

@Elroch @pfren

I think it's just the SF version. Here is SF8.  SF14 shows 0.00 when it reaches 37 ply.

I didn't wait for depth 99 because nothing (including the depth) had changed for ages.

But for @Elroch's purposes in his original post the position here would do equally well. No current engine will score that above -9.19.

If you try it, the chess.com evaluation does better with this similar position. (Skip to the final position in analysis to see the evaluation.)

Chess.com's Stockfish 14.1 NNUE - look at the image to see the version -  at 84 ply:  -2.78

I recall in the original 99 depth analysis (which I accidentally left on most of the day) there was a big jump to 9 pawns late on from 6 pawns earlier. As it gets deeper each ply can take quite a while.

Contrary to no current engine doing that, the engine that did so was the default engine used by the chess.com analysis tool - "Stockfish 14.1 (faster)", which I infer is the non-NNUE version, since they also offer "Stockfish 14.1 NNUE (stronger, 45Mb)".

I understand there is a degree of randomness in the Stockfish analysis tree. Here this seems to be having surprisingly large effects.

Well this is what it gave me after about ten minutes for the position I was referring to (-27.6<-9.19).

Yeah, I have used an even simpler 9 bishops of the same colour position to give a similar evaluation. Bottom line: Stockfish sees some dead drawn positions (even ones that are unlosable, like the 9 bishops ones) as massive material advantage.

IMO, chess is an logical art form. However. The objective has always been to win. So if you create a machine that will always win or, if against itself, always draws. Then yes. Chess is solved.

It's called minimax pathology.

That doesn't happen on a cloud cuckoo computer.

Amazingly, with a lot of time the analysis tool can think your position is a forced mate!!

On my computer I got the same result as @Elroch on that draw position, but I used SF12. Crystal is a derivative especially suited for dealing with fortresses, but it seems less selective, so I don't think it's stronger than SF, overall. @MARattigan got 0.00 with SF14 at depth 37, so I cannot understand why on chess.com SF14 at depth 99 with nnue enabled fails in that position.

Here Crystal (which is a custom stockfish build with aditional antifortress code) sticks to 0.00 at depth 21, and a very fresh stockfish-devel at depth 43. 

I see - it is likely the "custom anti-fortress code" which is providing the better evaluation. This obviously matters a lot here.

@haiaku re  this.

Could be I'm using a different NNUE level.

You are assuming an evaluation can't change radically with depth: it can! Also it can vary between analyses due to randomness in branch selection.

I ran SF12 several times at 90+ depth with nnue enabled, but never got more than -8 for White, SF14 is more advanced, though, and I see now that you didn't reach depth 99 with the nnue enabled. The path followed to reach a position can affect the evaluation, if intermediate results are stored in a TT; that should explain the result on chess.com @MARattigan got here.

@pfren I don't think Crystal actually uses an "anti-fortress" code, rather a fortress detection code and a less selective search, but clearly they have an "anti-fortress" effect during play.

@MARattigan that link on your previous post does not work for me.