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

i TOTALLY LUV pink octopusses...ty ty ty !

https://www.today.com/parents/nicu-babies-hospital-gives-stuffed-octopus-toys-preemies-t107823

It could be. In some cases, both play great, but at the end, one player gives the endgame away.

keep calm...im a physician who has removed a funny bone AND a wrenched ankle.

That's right. There was a time where there was no sign certain positions were forced wins. They were draws. But as computers got better, that forced win was found, even if it was many dozens of moves. 

So there isn't going to be any sign, at all, right now that there is a forced win. Which is why we see so many draws. The stronger the computer, and the longer the time control, the more draws. Because that is THE very best they can do right now. 

what about engines

#9653
"There was a time where there was no sign certain positions were forced wins. They were draws. But as computers got better, that forced win was found, even if it was many dozens of moves."
++ Yes, that is right, here is an example of a win in 400 moves:

However, this position cannot arise from the initial position by best play from both sides.
There are two black dark square bishops, so black must have promoted a pawn to a black square bishop. That is not best play for black: if black had promoted that pawn to a queen, then this position would not have arisen.

It's probably been posted before, but this is a 549 forced checkmate from a table base.  This ought to be enough to convince anyone that our intuition cannot be trusted and that heuristics from regular play are irrelevant.  The original question is unknown and remains unknown regardless of what we think the probable correct answer is. 

One might say that in every lost game of chess we've been able to find a mistake, therefore perfect play seems like it should be a draw.  But that misses an important observation.  In real games we are able to analyze less than 1 billionth of all future paths.  So in a lost game, we are able to find another variation that is not yet lost according to what we see.  But how would we know for sure if that path is not also lost when we see such a tiny percentage of the full analysis "tree"?

...and let us not forget the fortress. in this classic white is way up yet cant convert.

hmmm this is controversial and i did see this weeks ago but yeah i think engines will solve the game with white always having a slight advantage doing to opening but black may equalize

I'm not sure what the draw rate would be for a Grandmaster vs a computer, but looking online I get the impression the computer would most likely get a win over a draw in a very large percentage of the games. When computers match up at the highest level of their play, I read the draw rate is like 98%. My question is if a engine comes along with a calculating ability several orders of magnitude higher than any current machine, then would it be able to crush current computers as easy as nowadays AlphaZero could a Grand Master? Would this hypothetical situation then refute the argument that Chess at the highest level must end as a draw? Would that be an indication that Chess might have a forced win after all since experience would then have shown that a sufficient advancement of computing ability seems to trend towards a greater chance of a won game?

That's probably true. This particular long forced mate probably cannot be forced from the beginning position. So this is not the solution to chess. But it gives us an idea just how long the forced winning combination probably is.  A winning position for white can be forced from the beginning. It might take 500 moves. It might take 5,000 moves. Computers have not found it yet. It's possible the forced winning position from the beginning would take an additional 10,000 moves to force the mate. 

My guess is that as computers get better, future performance will dominate the best offered today. And as that happens, I would expect the draw rate to keep climbing. From 98%, to 99% to probably 99.999% or more. 

But the draw rate will never reach 100%. And because of that, "best play" will never be a draw. 

Game theory optimal (i.e. "best") play is either a draw or a win for one side or the other. No shades of gray.

so iows asymptotics control stuff, right ?

...okay.

Its a draw with optimal play while we have not brute forced out that answer think about super computers of the future that could brute force every possibility now imagine finding a forcing line hundreds of moves deep that starts from the opening where every single variation of its trillions of lines leads to checkmate because it only takes one that is a draw to make every single one of those other lines worthless.  TLDR Its a draw theoretically speaking using abstract logic in my opinion 

That post is not too long.

Like my neighbor always says, you be you. 

You don't need every variation to lead to a checkmate.  The winning side just needs one variation at each point that maintains the win.  After making that one move, my opponent may have 10 moves to choose from.  For each of those 10 moves, I might have 10 choices where 9 out of those 10 lead to losses or draw, but I only need 1 out of my 10 choices to maintain the win.  No matter how many losing choices I have at each possible choice, I only need 1 winning move for each of my choices.  Any of my choices that fails to keep the forced win can be pruned immediately from the tree no matter how many wins versus losses lie down that path.

The search method would employ a "minimax" algorithm and could use "alpha-beta pruning" (explained here: https://en.wikipedia.org/wiki/Alpha%E2%80%93beta_pruning)

You've not used abstract logic, and the question remains unknowable any any foreseeable future.

For example, in the above position after 2. g4 , black has around 30 moves that may lead to trillions of interesting games.  But none of them need be evaluated because 2... Qh4# ends the game with a white loss.  So we no for sure that 2. g4 is not optimal play for white no matter how many possible wins could result from that.  Perfect play analysis, after finding that Qh4# can stop evaluating 2. g4 even if there are still 25 moves for black that have never been considered.  When evaluating perfect play for white (and assuming perfect play for black), trillions and trillions of possible lines of play are trimmed this way before exploring all alternatives within the subtree.  

#9657
"This ought to be enough to convince anyone that our intuition cannot be trusted and that heuristics from regular play are irrelevant."
++ Yes, that is right, there exist 5-men, 6-men, 7-men, and 8-men positions won in > 50 moves.
However, there is no proof that any of those positions can be reached from the initial position with best play from both sides.
There is evidence of the contrary. In ICCF correspondence players may claim a win (even if it exceeds 50 or 75 moves without capture or pawn move) or a draw  based on the 7-men endgame table base. Such 7-men endgame table base win claims do not happen, 7-men endgame table base draw claims happen in 10% of drawn games. ICCF games end in draws in 39 moves average, 74% by agreement, 16% by 3-fold repetition, and 10% by 7-men endgame table base draw claim. Over the last 10 years the draw rate has gone up from 63% to 93%.