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

That's probably true most of the time. But that's not the question. The question isn't is chess a draw between really strong players. Really strong players make huge mistakes all the time. The question is if chess is a draw with best play, not strong play.

Really strong players get further and further away from best play every day. The best computers keep getting better, leaving really strong players in the dust. Really strong players make so many mistakes that when it comes to "best play" their opinion is pretty meaningless. When really strong players draw the worlds best computers every game, or even when the best computers draw every game, then we can start saying chess is a draw with best play.

I would guess that chess computers will still get better. I dont think either stockfish or leela are the very best and we've reached the end of chess ability. If so, then that means every game that was won was probably because the other side made a mistake, which is not best play. I think it will be a long time before best play is discovered. It's likely a hundred years from now a chess computer would win every single game against todays stockfish or leela. 

i used to think that that was a thing. if u put stockfish vs stockfish together, surely they would draw. however, when alpha zero became a thing, i started to rethink my theory. alpha makes moves that neither humans nor stockfish can fully understand, and thus we cannot say whether or not alpha is making "perfect" moves. however, alpha does display more appealing statistics compared to stockfish, so the subject is debateable

It is obvious that Alpha Zero is slightly better than stockfish. While most games between them are drawn--of the games with a result other than a draw--Alpha Zero wins more..

There is a lot more evidence since when I posted a few months ago--it is very obvious for   almost anyone looking at the evidence that chess is a draw. It was already obvious to any strong player that chess is a draw and now it is becoming obvious to anyone who has a utok and takes the time to look at the evidence.

Alpha Zero was just +52 elo stronger than Stockfish 8 based on statistics.

In 3 years later in 2020, both Alpha Zero and Stockfish 8 will horribly lose to current Stockfish 12 dev versions or Leela Zero. Latest SF or Leela in Tcec are well above +100  to Alpha Zero. 

Given enough time, with good opening books, those 3800+ engines will be playing with 95-98% drawish chess , theoretically going to 100% draw with enough thinking time.

I don't understand what is "obvious" about "the evidence."   As you mention, when AlphaZero engaged in "best play" against Stockfish, most games were drawn, but many games were won, and a few were lost.  https://en.wikipedia.org/wiki/AlphaZero#Chess_2

The (existing) evidence shows that best play does not necessitate any guaranteed outcome. 

Obviously.

GWTR   it is rare that we see best play.  But there is a ton of evidence that chess is a draw if neither side makes a mistake. It is also obvious that Alpha Zero is better than stockfish and that sometimes stockfish makes a mistake.

There is evidence that chess is a DRAW THAT HAS LITTLE TO DO WITH STOCKFISH OR ALPHA ZERO

Maybe we all have different opinions of what obviously means. It seems to me, maybe even obvious, that chess computers keep getting better. So nobody knows what "best play" is yet. Best play today is not best play tomorrow, because some computer figured out a better play.

I think it's obvious that we dont know what best play is. Because we can prove it. We can prove that computers keep getting better. So I dont think it's obvious that chess is a draw, because nobody has proven it. I also think it's obvious that white wins more than black, for the same reason. Maybe for some people obvious means it can be proven while for others obvious means just a guess.

what a stupid conversation

No. Just stupid people.

Such as yourself 

Feel free to point it out.

I'm good, but thankyou for the offer

Ya, you haven't read this thread. I'm the voice of reason around here. But if you want other opinions like "chess is a proven draw" or "chess is most likely a win for black" or "chess is neither a win or draw" (somehow) scroll up.

How do you know? The things you use to judge don't make perfect decisions.

We're all here waiting.

JimDiesel it is obvious that the vast majority of games are played by people who make many mistakes. Thus it is quite rare to see a game where neither side makes a mistake.

If you want the latest evidence that chess is a draw--study and look at correspondence chess at the highest levels.

ponz111 To say, "It's rare we see best play" implies statistical certainty where there is none to be had. How rare is rare? Has alpha zero ever played perfectly? If you think alpha zero has played perfectly how do you know? Surely you don't judge with... alpha zero? Why can I find positions alpha zero can't solve without a tablebase?

So the latest evidence is correspondence chess... that's your proof. Not even an explanation. Just that it exists?

Thanks for reminding me how dumb chess players are.

Jim  " It is rare to see perfect play does not imply statistical certainty".   It is just a very obvious statement if more than 99.9% of the chess games have at least one error in them then a game played with no errors is rare.

The latest evidence is given per the highest level of correspondence chess--the evidence is too complex to explain to you here.  That is why I suggest you look it up yourself.  If you are strong enough chess player you might understand the evidence? But I do not know if you are a strong enough of a chess player to understand the evidence?

It's a shame that the obvious evidence is too complex to explain.  That often seems to be the case on these boards.

None the less, are you saying that every game played at the highest level of correspondence chess ends in a draw?

Also, why would games played at the highest level of correspondence chess be more relevant than games played at the highest level of super-computer chess?

I remain convinced that best play from both sides does not always result in a draw nor does it always result in a win.

GWTR   I am claiming the highest level of correspondence chess shows the evidence or a good amount of evidence that chess is a draw. However to understand this evidence you would need to look at the evidence and also you would need to be a strong enough chess player to understand the evidence

THERE IS NOW SOME NEW AND VERY STRONG EVIDENCE THAT CORRESPONDENCE CHESS AT THE HIGHEST LEVEL IS VERY GOOD EVIDENCE THAT CHESS IS A DRAW.

And yes you can research yourself if you think you are a strong enough chess player to understand the evidence?

I certainly cannot make you into a strong enough chess player to understand the evidence.  I don't know    your strength but you may never be strong enough?