Will computers ever solve chess?

I said it is a general trend over the years. If one chess engine is much stronger than another chess engine--then the stronger chess engine/machine/whatever you  want to call it--will sometimes win.

However when chess enginess are closely matched there are more and more draws.

Yes, those two things you mentioned may run counter to my argument--that is why i said "over the years".

Well, it's a 2-3 year trend that there are less and less draws among the most evenly matched engines, Stockfish, Komodo, and Houdini...so that is actually the current trend.  Your data is obsolete at this point, much like saying "over the decades, computers vs. humans, humans have come out on top more years then they have lost".  Irrelevant.

And in any case, as they say in the stock market:  "past performance is no guarantee of future returns"

Here is another piece of evidence/also a trend.

It it well known that the strongest chess engines [other than Alpha Zero which appears to be strongest of them all by a good margin] by themselves are not playing the strongest chess. The strongest chess is played by these chess engines with a strong human player guiding.

Let me explain as this is in regard to correspondence chess which used to be my speciality... When I last played correspondence chess, chess computers were not used. This was because chess computers were not as strong as the players in the Final Round of the USA Correspondence Chess Championship. For example in the 14 games i played there was only 1 draw!

After that [except for some exhibition matches] i gave up playing correspondence chess.  Why? Because i did not wish to compete in the new situation. I was not good with computers. I did not even know how to use a data base! I did not know how to research games of my opponents etc etc.

After the 7th US Correspondence Championship there were more and more games ending in a draw. Also for the winners, there were more and  more draws.

This continued. This trend continued.  The other day a grandmaster friend of mine showed me a tournament he won. This was a tournament with chess engines helped by a human. This was a correspondence chess tournament.

He won the tournament [really a tie for first] with a score of 12 draws and 2 wins and no losses!! There were a bunch of players who scored 14 draws with no losses or wins! And the worse result [from looking at the crosstable] was 12 draws and 2 losses!

So this trend is a very good indication that when the strength of the chess players combined with a chess engine becomes stronger--there are more and more draws.

[by the way i am not disparaging the new breed of correspondence players. Most of the very top players were already near the top when i was playing. They have skills i did not have and do not have.

and [this is important] their games are very high quality--even higher than the games of the best super GMs today]

 by the way the recent game where Black played Rg8!? was not really something new--it was new to the over-the-board players--but had been seen more than once in correspondence play!

For those who play correspondence chess through chess.com of course chess engines are not allowed... 

What matters is what he says now and the context of his previous statement [and what exactly he said]

But i think that was several years ago and he was mainly talking about his matches against chess engines and the state of chess engines at the time.

I would guess if today you asked Kasparov  "Is chess a draw when neither side makes an error?" I am guessing he  would say "yes"

If you asked all the grandmasters in the world the same question--you would get a whole lot of "yes"

Chess is extremely drawish game with big margin or error. White half move adv is nothing/useless to force win. Even one pawn adv ( potential +9 or one queen ) adv is draw in many situations. To lose a game a player must do significent big error. ( Something like massive deviation/out of path from 6 lanes highway)

Here is a post from Eric Fleet  post #206 in the Forum True or False, Chess is a Draw with Best Play from Both Sides.

There are 3 types of positions White will  win with best play for both sides, Black will win with best play for both sides, and the position will be a draw with best play for both sides.

It is pretty evident that there is no forced win for Black from the starting position. So the question comes down to this: In the initial position is there a forced mate in xx moves for White?

While it is true that chess has not been solved, humans have been at it for centuries and looked at hundreds and thousands of lines without finding any opening that provides a forcing win for White. Computers which can crunch millions of moves per second also do not find such a winning opening.

Advantages tend to peter out with best play by Black and it is doubtful that such lines exist.

So, while is  not proven, I can say wih 99.999% certainty the opening position is not a forced win for either side.

By the way what drmrboss posted is very strong to me [in my own opinion] and that posting alone is enough to satisfy to me that the chance that chess is a draw when both sides make no mistakes is over 99%  

Here is something else to consider...Suppose a new game was invented just like our chess except Black would start with no queen.

Then suppose i said that this new game is a win for White if neither side makes an error...

 Some would make the argument that i could not know the game is a win for White because the game has not been solved with a 31 piece table base.

Now suppose one million games would be played and the result that more than 95% of the games were won by White...now we would have some posters say that one million games is "nothing" compared to the super number of possible games played and so those one million games mean nothing

Here is another piece of evidence. It is something NM Frederick Rhine wrote which turned into a well followed Wikipedia articlee:

First Move Advantage in Chess.

Here is another thing to consider [this is from a friend and is probably not an exact quote]  Most words have more than 1 meaning and someone cannot undersand a word completely without understanding all the possible meanings.

Chess is always a draw if you use felt tips to move the pieces

There are dangers associated with playing without a queen.

I was working as a brick mason on a building site near Stains (just outside London) The foreman, a guy called Spider (because of the web tattooed between his ear and neck, courtesy of his time in the forces) Wanted to play chess at break times but he wasn't very good, so I gave him queen odds in every game.

Eventually the inevitable happened and he beat me. He went around the site feeling very pleased with himself for beating me. 

 In my opinion it has NOT met the beyond a reasonable doubt threshold. But that’s the point.   We can all state reasons for why we think chess will prove to be a draw when solved. In my opinion most of ponz111 evidence isn’t very convincing, but I do have my own reasons why I also think it’s a draw. 

 It’s ludicrous to suggest anyone “knows“ it’s a draw.

 Folks, things like this is what ponz111 considers “evidence“ that chess is likely a draw. 

And once again, this is NOT evidence that chess is likely a draw. This is your personal appraisal of people who disagree with you. 

 You’ve pretty much been throwing out every random thought in your head and labeling it “evidence.“

Also, believing it is a draw is vastly different than claiming it has been proven. Gary Kasparov stated it has NOT been proven. 

I also *believe* it is a draw.

But yes, even IF that were true it is an argumentum ad populum fallacy.

Perhaps a good way to see if chess is a draw would be to play with 6 squares instead of 8. Take away a bishop and knight from both sides and it would still be even. Since there are only 36 squares instead of 64, it would be easier to solve. Whatever the outcome it would apply to the 64 square game as well - maybe it would be fun to play and easier.