Chess will never be solved, here's why

#1409
No: the problem is rather: white to play, black draws.
It is only necessary to prove black has one response that draws against all reasonable white moves.
It is obvious to try with the most promising responses: either 1 d4 Nf6 or 1 d4 d5.
The good assistants decide which one to try.
If one of these draw, then it no longer matters if 1 d4 a5 draws as well or not. 1 d4 is then solved to a proven table base draw.
If neither 1...Nf6 nor 1...d5 draw, then 1...a5 will not draw either, as 1...a5 does not contribute to the black play as much as 1...d5 or 1...Nf6.

No.

You are guessing that it is impossible for black to have a winning strategy because you can't see how there could be a zugzwang in the opening position.

My superior understanding only permits me to believe this is unlikely (and to know I can't be 100% sure).

 There are only three game theoretic possibilities for 1. d4 a5 - either it is a win, a draw or a loss. Computer analysis indicates that it is most likely to be a draw (but neither other possibility can be excluded for certain). The fact that (imprecise) computer analysis makes the probability lower than that for 1. ... d5 is not significant to its status for the solution of chess. BOTH MOVES (and the other 16) NEED DEALING WITH ON AN EQUAL FOOTING.

I don't know how you get 'infinity' there.
Despite the gigantic size of the task - it is not an infinite task.
Definitely 'finite'.  

@Elroch is technically accurate in most if not all of his positions regarding the task referred to in the forum topic.
Unfortunately though - there's a lot of dancing around red herrings.
The positions being represented or pursued do not amount to progress.
And that could be because nobody here is actually involved in any project to 'solve' chess.  

On occasion - a particular somebody has stated some 'upper bounds' on relevant quantities - but he does so without presenting derivations.
If we're going to discuss the mathematics of the task then suggestions:
Suggestions that could be implemented - as opposed to dominating the discussion.  Nothing need dominate.

Math - like chess - lends itself to proofs. Or even does so directly.
Explicitly !  
In both cases - even built around such ! 
Math is even designed with such in mind !!

Investigate.  Proceed with an accurate premise.  Or premises.
Math well presented - speaks for itself. 
Regardless of how much whoever might want to obsess or ordain about the semantics of 'premise' - with a hand wave of imaginary authority. 

Start with hard and explicit numbers.  Include maths proofs - not links.
We know the number of possible positions is gigantic.
We know that only computers could sort the legal positions from the illegal ones.  
We know that humans and humanity literally and phsyically don't have the time to count up - sort - solve all the positions.
Repeated suggestions that its a matter of money - don't make any progress either.

Regarding the science involved as opposed to the math -
again - math helps tremendously in presenting science.
The power of modern supercomputers is measured in units called 'petaflops'.
Contemporary supercomputers operate at far under 1000 petaflops.
One petaflop is a trillion operations per second.
Apparently that is a straight trillion ...  not a 'byte' trillion as it were.
Am I positive?  Not yet.  It can be checked.
And its apparently an American trillion.  Not a British one.
Yes - they're different.  Unless that's changed.  

Point:  operations per second needs to be linked to rate of progress of solving somehow.
So does needed memory space.
Money ?  No need to even consider that.  If the task can't be physically completed in a practical period of time - no point in throwing money at it !  (but money is thrown at it.  So there's a different motive or motives.)


Are there ways to know that it 'can't be practically done' right now?
Yes - the task is so difficult that even just 7-piece positions couldn't be tablebased without skipping castling and en passant considerations.

Supercomputers have been around for decades -
but that 7-piece 'milestone' represents a very tiny percentage of the 236,196 possible material situations (which generously doesn't include promotion situations) and upon which each term of which in turn represents a varying total of positional arrangements.  For each.
Some of the more difficult of those 236,196 'situations' could each itself take billions of years to 'solve' perhaps.

Regarding the actual project and money put into it - 
maybe the people doing it don't want to publish some of the more gloomy numbers !   (the politics of grants and fundings and budgets).
Positions need to be sorted into three categories.
1) Illegal positions:  (no need to analyze those further - but they still have to be computer-identified or triaged before being dismissed - adding to the task)
2) Checkmate positions.  The next easiest.  Stop! 
No more analysis - its Checkmate folks. 
3)  The others.  Suggestion - put them aside.  Analyze later.
Where 'later' is putting it mildly !  But - count them up.

Projects doing this kind of thing might never publish those numbers ?
Why not ?   Back to financing again.  
Conjecture:  Such projects are perhaps linked to big business - .com interests - universities - or even governments and the military - computer manufacturers - software organizations.
Could it be some billionaire trying to feel more intellectually complete?
Doubtful - people don't become billionaires nor stay that way chasing such things.  

Provocative declarations of the improbable or unestablished -
make good clickbait? 
Say something invalid - so somebody has something to argue with ...

But it works the other way too -
declare something valid - so whoever can argue with it anyway.

     That is the crux of many of the disagreements we argue about here. Whether we can put our opinion of what the ultimate solution might be into the preconditions of the analysis used to "solve" chess. My own opinion is that perfectly-played chess will result in a draw, but I am not so convinced of my infallibility as to assert that it must be true.

well there's the 'I'm just far better' clickbait again.  In post #1416.

White wouldn't have to be 'zugzwanged' to have losing options on his first move.
And 'zugzwang' itself - is an incomplete term.  Often misunderstood.
There are situations where whoever loses - not because its his/her move -
but instead - because of the continuing obligation to move.
A different thing.  Such obligation producing a win for the opponent regardless of whose move it is.
Whether to make one move or a continuing series of moves.

My take on it is that 'no mistakes will produce a draw' - is uncertain.
We simply don't know.
Many 'anecdotes' might so suggest - 
but do the supercomputers 'solve' all positions hypothetically arising from all GM games to so determine?
The short answer is No.
With some proof.   They can't so solve anyway.  Already established.
The fact that they can and do solve some positions does not so prove.
Most players can thoroughly solve K+R versus lone King too.
Fast.  They do.
Without computer assistance.  That doesn't mean they've 'solved' all positions leading to that.  In fact - they haven't.

You might.  Someday.  You might someday follow your own advice.
On something.   Maybe not here though. 

     Whether or not one person is better at guessing than another has no bearing on any particular situation. Even though person A will guess more correctly than person B 80% of the time person B's guess WILL BE more correct once in five tries. Meticulously examined proofs are ALWAYS better than well-informed guesses. 

     Although one's opinions may be sufficient in their own mind, that has no relevance to actual results.

Yes.  Good post.

Today I read a Reddit article (I usually avoid Reddit and 'Quora' and the almost always inferior/useless 'Stack Overflow' websites)  ....
because they don't quickly provide the information sought ....
but the article concerned 'Gestalt' as it pertains to chess. 
So titled players including GM's were talking there about how they look at positions and think about chess ...  and there it was again ...  'pattern recognition' 
Titled players push that all the time ...
so I then posted to that effect - how such 'pattern recognition' nor 'calculations' could Not possibly help anybody with spotting the first solution move of the tactics problem I was posting in.  (71% of the attempts failing - not surprising - since if the option isn't both spotted nor examined then that's it ...  no 'Solved')
'Observation' was key ...  very distinct from 'calculation'
happens constantly in tactics puzzles.  And in games too.
'Doctrine' interfering with effective chess education.
Much of the time GM's give good chess advice - but the 'doctrine' approach is like with science ...
much better are the logic and evidence approaches to the game.
Even a 'history' approach to math and science and chess too - would be better than a 'doctrine' approach.  

Once you reach the SOLVED! stage you will hear a strange electronic piece of music followed by a high pitched japanese geisha voice: "Congratulations! You just completed Chess level 1 and may now proceed to level 2 of 32767 remaining levels. Good luck!"

And:  Introduction to three-dimensional chess.  

Makes me wonder what the weather conditions and incline were like on your way to school as a child. Especially compared to the way home from school. 

It wasn't asked.
It was 'make me wonder' in that post.

Funny?  a gigantic answer was forthcoming - supposedly with a reason ...

We have to establish exactly what "weakly solved" means, as @Elroch pointed out. According to Allis, a game is weakly solved if, for the initial position(s), the game-theoretic value has been determined and a strategy has been determined to obtain at least the game-theoretic value, for both players. "Determined" means not by guessing, using our experience of the game. We don't know which is this game-theoretic value in chess, nor the perfect strategy. The fact that White has an expected score sligthly higher than 0.5 does not prove that Black can only draw, with perfect play. To understand how hazardous this kind of assumption is, think about how many positions we have seen, where one colour is a rook or a queen down: more than 95% of the times, the position is lost for that colour, but we know that there exist positions where a tactical combination, or a particular positional situation, subverts the expected result. Well, one of those very positions could be the initial one. But we would not see that, because of our (and engines) limited calculation ability. So there might exist a line that, unexpectedly, leads to a win for Black and we cannot rule out this possibility, cutting down the exploration of "inferior" openings for Black, otherwise the solution is weak, but in another sense.

The amount of talk on that specific subject is just ridiculous