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

Sort:
ponz111

PATRIOT You are using"STRAWMAN" again.  I never said or implied . that now [2020] is the time when chess computers stop improving. 

And you simply do not have the facts or chess knowledge to understand why chess engines of the future will not be able to always beat the best chess engines of today. If you are really interested in this subject I suggest you makes a very extensive review as to what is recently happening in ICCF correspondence chess at the highest levels  and some new happenings associated with that subject?

lfPatriotGames
ponz111 wrote:

PATRIOT You are using"STRAWMAN" again.  I never said or implied . that now [2020] is the time when chess computers stop improving. 

And you simply do not have the facts or chess knowledge to understand why chess engines of the future will not be able to always beat the best chess engines of today. If you are really interested in this subject I suggest you makes a very extensive review as to what is recently happening in ICCF correspondence chess at the highest levels  and some new happenings associated with that subject?

Yes. You did. I quoted you. You very much implied that now is the time computers stop improving. You said, and I quote,

I very much disagree with your idea that super advanced chess engines of the future will be able to win most games from our now super chess engines.

 

lfPatriotGames

Chess computers in 1988 (at the time) were the most advanced chess computers in the world. But who would say that 30 years later those future chess computers would not win most games against a 1988 super chess computer? That would be pretty naive. 

ponz111

Arthur  Sure I can claim that chess is a draw?  Why? because of the ton of evidence that chess is a draw. Also there is more than one version of stockfish and I think both Lela and stockfish are progressing. So the study you mentioned means little except one chess engine was a little stronger than the other. 

I certainly do not have to prove chess is a draw in each and every variation to prove chess is a draw.  That would be so many variations that it would take a zillion life times to prove much of anything. The question of this forum does not ask if there is a 32 piece data base to prove chess is a draw!!! There are other ways to prove something and to me the ton of evidence proves chess is a draw.

Sasha4325

Привет!

ponz111

PATRIOT it is you who are naive if you thought that in 1988 people did not think that  future chess computers [of say 2020] would win  most games against a 1988 super chess computer. 

Many knowledgeable people knew this was possible. 

I knew it was possible and that is one reason I gave up ICCF chess at a very high level about that time!! 

PATRIOT  You quoted me as saying "I very much disagree with your idea of super advanced chess engines of ,the future will be able to win most games from our now super chess engines."

This DOES NOT IMPLY THAT NOW IS THE TIME THAT THIS IS THE TIME COMPUTERS STOP IMPROVING!!! You are showing a lack of basic logic here. .

lfPatriotGames
ponz111 wrote:

PATRIOT it is you who are naive if you thought that in 1988 people did not think that  future chess computers [of say 2020] would win  most games against a 1988 super chess computer. 

Many knowledgeable people knew this was possible. 

I knew it was possible and that is one reason I gave up ICCF chess at a very high level about that time!! 

PATRIOT  You quoted me as saying "I very much disagree with your idea of super advanced chess engines of ,the future will be able to win most games from our now super chess engines."

This DOES NOT IMPLY THAT NOW IS THE TIME THAT THIS IS THE TIME COMPUTERS STOP IMPROVING!!! You are showing a lack of basic logic here. .

I think you are having a hard time here. If you didn't mean now when you used the word "now" what did you mean? Those are YOUR words. If you didn't mean it, then dont say it. 

lfPatriotGames
Optimissed wrote:
lfPatriotGames wrote:

Chess computers in 1988 (at the time) were the most advanced chess computers in the world. But who would say that 30 years later those future chess computers would not win most games against a 1988 super chess computer? That would be pretty naive. 

Hi, in the 1970s it was known that semiconductor advancement would accelerate and continue to progress, so that a basic rule of thumb that processor speed could double every two years was floated and it seemed to be true. It was based on empirical data. Over the last few years there are signs that processor speed advancement is slowing and it probably depends on another breakthrough. Peripheral circuits, especially memory, can be developed though, to make the processing environment faster, but there seems to be some evidence that computers may not keep on getting faster and faster. So this seems to me to suggest a more algorithmical approach to things like chess software .... an increasing departure from the brute force ideas that served us well up until ten years ago but which produced software that produced extremely boring games.

I would suggest that the next phase is a more automated software design strategy for complex calculations like chess.

I'll take your word for it on how chess computers advance. But it seem to me there is always a breakthrough of some kind. I'll bet there will be many breakthroughs in the next 200 years. (assuming people and computers are even still here 200 years from now). I was just having a hard time following Ponz's statement that he thought future chess computers wouldn't win most games against todays chess computers. Barring some kind of unforeseen catastrophe like WW3 or something, I think advancement is inevitable. I dont see any reason why chess computers wont keep improving, even to the point where a future computer will beat todays computers easily. 

pfren
MeleiroZ έγραψε:
ponz111 wrote:

Of course in Correspondence play at the highest levels if there are any wins--White should do far better than Black. 

I won the 7th United States ICCF Correspondence Championship. This was the last Championship before computers.  I set a USA record which will NEVER be tied or broken. I won All my 7 games with Black!  With White I only scored 6 1/2 out of 7.

Nowadays black wins in high level Correspondence Chess are very rare. I won 2 recently, but one of them was a simple "mouse slip" (if you can use this expression in CC...) and the other win was the only one in 5 or 6 years. You are right, nobody will win 7 games playing black, like you did.

 

I have won several games in CC with Black- it is all about lines where computers fail to grasp the essence of a position. Here is a very characteristic line, where I deviated from previous practice at move 29, and won shortly afterwards:

 

 

It just needed to resist the temptation of winning back the exchange (which actually is what the computer is advocating), and build up an attack based to the power of the bishop pair. Actually I think that white should not allow the exchange sac at d5, as Black has terrific positional compensation, and of course realizing that the monster at e5 is much more useful than white's rook. White's position disntegrated very fast, and in the final position the knight is a spectator, and cannot come back to life without heavy material losses.

Opponent was a strong player (German NM) but with limited CC experience. I do not think an experienced CC player would agree entering this line.

ArthurEZiegler

I can't resist adding yet another comment. I think that ponz11 point is that as the playing advances there are fewer wins and those are results of mistakes, therefore eventually all games with out mistakes will eventually be ties, even those played by super computers from the future.  Still, he should admit that this is not a proof in a mathematical sense. As for computer advancements still in early stages of development are quantum computers.  don't claim to understand how they work but know that instead of thinking in binary (zeroes and ones) the data is able to be in an undetermined state, like the famous Schrodinger's cat, and somehow this enables it to process many lines simultaneously. Would a super quantum computer of the far future be able to determine hard to find winning combinations? It appears that Optimissed thinks this not the case with optimum play for both sides, I am not so sure.

ponz111

I Never said computers would not continue to advance!!! In fact I expect them to continue to advance!!  But since it is very obvious chess is a draw with best play--No computer in 200 years or 500 years is going to turn  the game of chess from a draw to a win for either White or Black.  What will happen is computers will give us even more confirmation that chess is a draw.

ponz111

PATRIOT When I used the word "now" I meant the present time [2020]

JimDiesel22

Does anyone know of correspondence chess databases with and without computer assistance? I've seen ICCF and Lechenicher, but ICCF has a very limited amount and skill range and Lechenicher supposedly has both computer assisted and unassisted games, but I can't find the link.

Direct links only, thanks

JimDiesel22

Also, long OTB game/unassisted game databases and engine vs engine databases.

ponz111

PATRIOT  If I understand you correctly you seem to think/guess as computers get stronger over time that this means the best chess engines of 2020 will usually lose to the best chess engines of the future [say 2220]

I wish you would stop saying that I do not believe chess engines will get better over time. Nothing could be farther from the truth as to what I believe!

What I believe is of course computers will get stronger over time. But the same thing [or similar thing] will happen in chess as it did in checkers.

In checkers there were many top players who were absolutely convinced that with best play--checkers was a draw. Of course these top players did not have 100% math proof--but they had much evidence. 100% math proof was not available at the time.

What happened is that computers got stronger and stronger over time. But this did not mean there was a forced win from the opening position.  Finally checkers was 100% math proved and as the top players already knew--checkers is a draw with best play.

ponz111

Jim  after I retired from correspondence chess--more and more correspondence chess at the highest levels was human with the help of computers.

Back in the day more than 30 years ago--there were data bases of top level correspondence games which were played without computer help. But now such data bases are not so relevant--who would buy them except old guys like me?  People often buy data bases to find play at the highest levels--and this means computer assisted correspondence games from the best players or games of super chess engines such as Alpha Zero.

JimDiesel22
ponz111 wrote:

Jim  after I retired from correspondence chess--more and more correspondence chess at the highest levels was human with the help of computers.

Just shut up if you aren't of help.

ponz111

JimDiesel,  ha ha you sound like a certain president of a country who wants to shut people up!

tongue.pngtongue.pngtongue.pngtongue.pngtongue.png

ponz111

Pawned  It is true that after White opens 1. e4  [or 1. d4 or 1. Nf3 or 1 c4] he will enjoy a minute practical advantage. But this does not equate to a forced win.

For example White might continue 1. e4  e5  2. Nf3  Nc6  3. Bb5 but this has been analyzed to a draw. 

DrFrank124c

Maybe!