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

Yes, when you get to elo 4000 [let's say two elo chess machines playing each other] you wil get very close to all draws. 

Right now the best Centaur chess is about 3400 and you see about 90% draws.

In your opinion, of course - we don't know for sure.

Yes, we don't know for sure. I do not know anything for 100% sure. It is possible that I am just an idea in someone's mind and you [fburton] and all the world and stars and planets and people and animals and everything are implanted into whatever I am to make things look real.

So, I am only about 99.9% plus sure that chess is a draw and would be willing to bet my life on it.

This is an interesting argument. However I think there is more than a linear regression to be done here, as you remarked. In particular, the points towards the higher ratings seem not to follow a linear progression. I also think the quick grandmaster draws play a role in the enormous gap between draw rates at the 2200 level and draw rates above that (since all the games used are tournament games). In other words, there is some kind of noise in the signal we are observing.

There must be deeper analyses of the evolution of draw rates available online, maybe we should look for it.

Ponz I'm still waiting for a source that would justify your 90% draw rate between top centaurs.

ponz111, it's just that your opinions tend to come across as statements of fact - the one above about Elo 4000 being a good example.

If I had said that my linear trend fitting experiment had shown that there would be 100% draws at Elo 5000 (or 6000 or 8000), would you have said "Yes, when you get to elo 5000 you will get very close to all draws", or "Actually, you will get close to all draws by 4000"? It is this kind of dogmatism (is that the right word?) that some people find a bit irritating or at least puzzling.

Don't let that put you off though (I'm sure it won't)!

fburton.  Re elo being 4000 and chess engines playing each other I said "you will get very close to all draws." This is not the same as you will get all draws.  So when I am saying "close to" you seem to take that as not "close to" but 100% all draws.  I am careful about my wording and that is why I said "close to"

Dogmatism would be if I said "when you get to 4000 elo you will get all draws"  Or "It is 100% certain chess is a draw"]  Also dogmatism is often sans reasoning and I give a lot of evidence.

What people do not like is that I am so sure chess is a draw that I state I am 99.9% sure and some think that is equivalent to 100% sure and some just do not like me having a very strong opinon on this subject.

Dogmatism is the result of something taught to people over a period of years. Nobody taught me that chess is a draw. I am sans dogmatism as much as anyone you will see.

People who do not like me being so sure [99.9%] chess is a draw do not have whatever is in my mind and whatever chess experience I have had.

Tronchenbiais.  My source for draw rate in Centaur chess for the top players being about 90% is just something I heard to be true and I could be wrong.

I thought I heard this from one of Firebrands assertations but again I could be wrong.   

That's why I hesitated to call it dogmatism. Maybe it isn't the right word. Do you have any evidence for a particular Elo figure at which draws become a 100% certainty, or even 99% likely? (Apart from my rough and highly speculative regression on 7 data points!) Do you think the rise on draw percentage hits the 100% ceiling abruptly, or does it approach it more and more gradually (asymptotically)?

flurton,  To answer your question--I do not know.  I just know it is very reasonable for there to be more and more draws as the ratings go up.

Rating 4000 is  probably close to perfect chess is my guess.  I am guessing there is little or no difference between 4000 elo and 6000 elo.

However this is only a guess. 

I have sometimes things in my mind that are rather unique and I am not afraid to "take the chance of looking very foolish" to put out these ideas to the chess playing public.

Here is an example. It is well known I am a Ponziani fan.  In fact, it is possible I know more about this opening than anyone on this planet.

I have always advocated for this opening and still do  

However I put out a challenge to the best players with the best machines to play the Ponziani AGAINST ME. In other words I would take Black.

Now, [question] do you know anyone who has been the chief supporter of an opening who would turn things around 100% to say he wants a public challenge and he will play against that opening? 

Here is another example. A player who was always "against the Ponziani" and who even downgraded a book review on Play the Ponziani  [which I co authored] was being accused of being an imposter in Ponziani Power Vote Chess [actually one person 3 characters] when all three voted for a move which would lose the queen.

This cause a lot of problems for Ponziani Power.  This accused player, who certainly was no friend of mine, came to me to ask me to help prove his innocence!  [I had thought there was a chance he had done the dirty deed but was not certain at all]

So, I tried to defend this guy and give evidence. At first the Super and others would not even listen to my evidence.  The Super especially was so sure this guy did the bad deed that he was very upset with me for defending the suspect.  

To make a long story short, I was able to prove [with the help of the suspect] that not only did this suspect  not do this imposter thing-- that we were able to show who did this deed.  It was someone who lived near London and who did this just to disrupt Ponziai Power.

Now the super defriended me and cut off communication and guess what?

I am no longer part of Ponziani Power. [for many reasons]

My point is I am willing to take a position if I am certain I am right even though many will think I am being arrogant.

Thanks for your thorough and detailed reply.

To answer your question - I don't know who has done that regarding a favoured opening, but I am surely the wrong person to ask. 

Now, I wonder... has anyone looked at the relationship between Elo and draw percentage for engine matches? It would be a lot of work (for the computers), but at least it should stop results being skewed by 'short draws'.

actually in the link I gave, there is a link to such a study (a link in the link).

http://kirill-kryukov.com/chess/kcec/draw_rate.html

Computers get a lower draw rate. I don't really know what to think of this however. They just seem to win a lot when there are paired with a weaker engine and draw a lot when at the same level. I think the study mainly proves that engines play at a constant level. Maybe there is more to say, I am not sure.

Tranchenbiais

I would expect if chess is a draw that when chess engines get stronger and assuming they are playing against similar strong engines there would be more and more draws. 

Those point clouds sure show a lot of variability!

To you both, as I said I am not sure this study brings much information to our debate. The strength of the engines vary a lot. When you look at the matches between the top engines only (go to the page engines and select the ones you want to look at), we get drawing rates quite similar to those of grandmasters.

I am often surprised at some of the decisive results when I watch engines play each other on my computer. One factor is that they do not become bored. The engine will keep making moves, and if it has an advantage, it will seek to press that advantage.

maybe you have answered this already, but when will we probably get a computer to play 4000 chess ?....using the past timelines to predict it ?....

It is hard to say, but given enough money, I don't thin it would take so long.

From what I read, engines gain about 50 points every year right now. Nothing says this will steadily increase, but if it does, we will have extremely strong engines soon (and a 4000 rating does not look unreachable).

Merci Tronchenbiais.......

 Your question is interesting but might not be quite accurately stated, i dont wish to go into quibbling or mind-reading:

I'd prefer:

True of False: a chess game can show random results with best play by both players (assumed equal), with a slight edge to white.

in other words, if both players are not equal, your question is false! a 2400 guy will always win agains a 200 elo, assuming all else being equal ( perfect play by both means they are both doing their best?)

raptor101010   I was assuming best play rather than best efforts. I agree a 200 rated player will lose to a 2400 rated player [all things being equal]

However don't you think your suggested title to be a little long?

This is a question very hard to pose as there is the term "Best Play" which can be intrepeted several ways.