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

Smylov : I disagree. I do not claim chess is not a draw. I claim we don't know if chess is a draw and I also claim it is a terribly hard problem. The burden of proof is on the person who says chess is a draw. But I don't really think the discussion is about finding a proof that chess is or is not a draw, it's more about the motivations that one can have to believe chess is drawn and the arguments that can be opposed to these motivations.

non proven math statements are called conjectures. They are never accepted as long as they are not proven. Sometimes one will base his work on admitting a conjecture, but he fully knows that if this conjecture is disproven one day, all his work will become flawed and unusable. Axioms have a special status. They are not proven because they are the basic rules of maths. Asking to prove an axiom would be like asking to prove that bishops move diagonaly. You can't answer anything else than "that is how it works."

I don't know what specialists you are talking about. The game theorists are unanimous that we don't know the answer. Some will say they think chess is likely a draw, but I dare you to find one who will strongly afirm that.

george jetson  I think you are missing that I do not claim to have 100% math proof that chess is a draw. Nobody has such a proof and it is very unlikely that anybody ever will have 100% math proof that chess is a draw.

What we have is indications and evidence.  And these and my own chess experience make me 99% certain that chess is a draw. In fact I would bet my life on it.

It does not matter who has the burden of proof that chess is a draw or is not a draw because we will never have 100% proof so it is a moot question.

Yes, chess playing by computers is getting stronger and yes eventually there will be little to play for at the very top levels.  This is a prediction.

That chess at the top level is becoming more and more drawish is my hypothesis and it is standing up to fact.

There are many who insist that they do not know if chess is a draw or not because it has not been proveen 100%. Fine you do not know. I do not know 100% but I am not afraid to give an opinion based on several different factors.

As for closing this forum--I am all for it as we seem to be talking through each other  One side says chess cannot be proven a draw and nobody disagrees with that. The other side says that while chess cannot be proven to be a draw 100% math wise the evidence is overwhelming that , in fact, chess is a draw.  So you can have people undetermined which means there is no math proof and that includes me. But also people who are willing to say that the evidence points one way.

I would be glad to close this thread as it makes some people upset.

However chess.com will not just let me close the thread. 

You use this same syllogism (bolded above) against everyone who disagrees with you. It's really quite tiresome.

"Logic and Science" constitute only 1/2 of the human conversation.  You can't clap with only one hand.  Sorry.

zborg your quoting is way out of context !

You indeed seem to know a lot about being arrogant and briging nothing new to the discussion. 

i second that!!!

Come back. Open can of worms. Find new circles running, and additionally ad hominem attacks and "party grouping" (ie people will never admit someone on their side has made a lousy argument). Close can. Go away.

Please, if you still want to do that, move to off-topic. This has been in hot topics for a month now.

I do not see anywhere where anyone mentioned "voting" Please someone show me where "voting" was mentioned?  

There is one heck of a difference between individuals expressing an opinion and "voting"

Also regarding the burden of proof is on those who claim chess is a draw.  I do not see anybody on this forum who claims that it is math proven chess is a draw. I, for one, have always indicated you cannot prove 100% that chess is a draw.

I have seen no mention that people must "vote" 

Maybe I misunderstand from the start but I though this thread was about exchanging the our opinions about the nature of chess and the reason behind these opinions.

Ponz's opinion is that chess is a draw, and his reasons to believe it are the evidence he gave.

My opinion is that I have no idea what the outcome is, and I think I gave some good reason that ponz's arguments do not necessarily give good confidence that chess is a draw.

Of course, everyone is free to think what he wants, but I think this exchange of arguments is interesting. Of course nobody is going to change his mind, but that is not the aim of the discussion.

Now ponz, if you want to close the thread, you should say it clearly.

If you want to continue discussing, I still have one question that you haven't answered yet :

You seem to think (correct me if I am wrong) that GMs and engines now play close to perfection. However in ten years we will have stronger engines capable of beating them 100% of the time, which will show they are not the closest to perfection we can get. These very engines that will beat out GMs will themselves be beaten by yet stronger engines, and so on. In 40 year, we will hence have engines capable of consistently beating engines capable of consitenly beating engines capable of consistently beating engines capable of consistently beating today's chess centaurs. Knowing that, how can we be so sure that we play so close to perfection ? There seems to be many steps to go before reaching that state.

Tron   I asked to close the thread but go no reply. 

To answer your question yes, I think that Centaur Chess is getting fairly close to perfection. It is not there yet [maybe in 10-20 years ]but for one thing they have shown if you want to win--no point in opening 1. e4  

I think they will get better in 40 years but not so much better that they could beat today's Centaur players consistently.  

So I do not know in 40 years the chess engines will be good enough to wipe out today's Centaur Players. Nobody knows.  I do not think it will happen but I do not know.

One thing  I am sure of [and so far facts show I am right] is that as chess advances over time--there will be even more draws.

When I say playing close to prefection, I mean playing chess without errors which would change the outcome of the game.

I posted my own thread here in July named "Does winning involve any luck ?" with 5400 views and some interesting comments. My own personal feeling is your state of mind is the most important, you could be tired, preoccupied with some or other problem, without being aware of it or Jack Daniels plays a role. That's when you make a mistake and "luck" kicks in for your opponent. Same goes vice versa.

I have again made a request to close this forum. I like a good debate but do not like when it causes hard feelings. 

george jetson I get your argument that our heuristics might not even converge towards perfect play as they are refined more and more. It looks a bit counter intuitive to me though, since the way they approach chess is by looking foward a certain amount of move. Once their lookahead gets high enough to see a checkmate or a dead draw from the starting position (we're not there yet), I think they'll be playing perfect chess.

I am not sure I understand very well how engines work though. I heard they sometimes decide not to explore a variation, because they estimate the variation is too bad. Maybe these choices of not exploring all positions prevent them from being able to find perfect play. Is that what you mean ?

Those who are upset are free to leave, surely? I have found this thread instructive, entertaining and puzzling (and continue to do so).

george jetson

a I did not say we are at the limits right now--come on-I did not say that at all

b  Also I did not say computing power will not advance.  Why misquote me?

Regarding todays strongest chess engines. I just got the free version of Stockfish and it is comparable to houdini.  I had it analyze one of my best games.  At one point I forced an endgame where I had a good knight vs a bad bishop.  It did not see I had a big advantage.  It took several moves before it saw this.  So, the positional play is not so good.  Also in the endgame I made a sacrifice of my knight for a pawn.  This I had to calculate several moves before it happened.  It did not "see" this sacrifice until the move I made the  sacrifice and then it took a few seconds but then it said I as abolute winning.  So the best chess engines have weaknesses now but if you combine them with a strong player you have something.

Chess engines are usually very good in tactics--they need to become better in such things as a good knight vs a bad bishop.

So chess engines have room for advancement but even so it you team them up with a strong human the humans can compliment the chess engines.

Asymptotic approach to perfection still means the better engine (or Centaur) will beat a worse one, doesn't it? Or will there come a point short of perfection where draws happen 100% of the time? How close are we to that point?

flurton  it depends how much better one engine is than another.

We now have to chess engines which are very close in strength

Stockfish and Houdini.  While neither is perfect -they are strong and if you comvine them with a very good Centaur Player you do not have perfection yet but they are way more than half way there. I would love to see a long match between the two. I would guess 50% draws.

I do  think that short of perfection we will not have draws 100% of the time but would not be surprised if we have draws 90% of the time.

Hard to say a machine or chess engine ever reaches perfection but even a class B player may happen to make the best moves in a game. 

Sorry, that syllogism mentioned above is the method of @George_Jetless.  Yours too, @Tronchen.

You guys are both cut from the same cloth.  You both make the same methodological assertion about logic, science, and uncertainty, ad nauseum.

Any thoughtful reader will pick this up, if they can wade through your diatribes.

George Jetson  you are ducking my assertion that you misquoted me twice.

Pleas address that.