Will computers ever solve chess?

Sort:
godsofhell1235

Oh, I guess I need to add that it takes more than a pawn's advantage to win most endgames and that players blunder in proportion to the complexity of a position, and inversely proportional to their skill, and that human players of high human-relative skill blunder rarely even in normal position.

So if skilled players play passively, then it's not unlikely to be perfect.
If there are many thousands of instances where skilled players have played passively, then it's likely a perfect game has been played.

USArmyParatrooper
ponz111 wrote:
USArmyParatrooper wrote:
godsofhell1235 wrote:

It's evidence for a perfect game when I say neither player puts their opponent under pressure.

I totally disagree. Players, engines, whatever can completely miss opportunities without ever knowing it happened. It only means neither side was able to initiate perceived threats.

Of course players can completely miss opportunities without ever knowing it happened. But in the particular game shown--this did not happen.

How do i know? My own analysis of the game. I would bet a strong chess engine would analyze the game the same way...i e there was no point in that game where the theoretical result was anything but a draw...

“Ladies and gentlemen of the jury, the defendant is guilty. I know this from my own analysis of the incident.”

 

Cool story, bro 👍

godsofhell1235

I added some stuff. It's unfortunately on the bottom of the previous page.

USArmyParatrooper
godsofhell1235 wrote:

Oh, I guess I need to add that it takes more than a pawn's advantage to win most endgames and that players blunder in proportion to the complexity of a position, and inversely proportional to their skill, and that human players of high human-relative skill blunder rarely even in normal position.

So if skilled players play passively, then it's not unlikely to be perfect.
If there are many thousands of instances where skilled players have played passively, then it's likely a perfect game has been played.

 

“So if skilled players play passively, then it's not unlikely to be perfect.”

 

Based on what? 

 

Take Stockfish operating full strength on a super computer. Take ANY player and give him no handicap. Just tell him to play passively. How do you think that will work out?

 

That’s JUST Stockfish. Imagine a monster that solved chess. 

godsofhell1235
USArmyParatrooper wrote:
godsofhell1235 wrote:

Oh, I guess I need to add that it takes more than a pawn's advantage to win most endgames and that players blunder in proportion to the complexity of a position, and inversely proportional to their skill, and that human players of high human-relative skill blunder rarely even in normal position.

So if skilled players play passively, then it's not unlikely to be perfect.
If there are many thousands of instances where skilled players have played passively, then it's likely a perfect game has been played.

 

“So if skilled players play passively, then it's not unlikely to be perfect.”

 

Based on what? 

Basically what I'm saying is the starting position is solid and passive i.e. not likely to cause a mistake.

If the players do nothing to change this fact, then a mistake is unlikely to happen... at least one large enough for it to change the 32 man tablebase evaluation.

Take Stockfish operating full strength on a super computer. Take ANY player and give him no handicap. Just tell him to play passively. How do you think that will work out?

Both players have to play passively, as in the game I posted.

I've played blitz with full strength stockfish. Not often, just a few times, just for fun.

Surprisingly sometimes I was reaching equal positions into the late middlegame... like... boarderline draw-able (but I had big practical difficulties, then 1-5 moves later I was 100% lost).

That’s JUST Stockfish. Imagine a monster that solved chess. 

A perfect game between players of mismatched skill is highly unlikely. But that's not what I'm talking about, so I don't know why this is relevant.

 

godsofhell1235

And I only did so well against SF when... you guessed it... symmetrical passive stuff. Like an exchange french, then a lot of pieces come off before move 30.

When it was more complicated I did about as well as you'd expect.

USArmyParatrooper
godsofhell1235 wrote:
USArmyParatrooper wrote:
godsofhell1235 wrote:

Oh, I guess I need to add that it takes more than a pawn's advantage to win most endgames and that players blunder in proportion to the complexity of a position, and inversely proportional to their skill, and that human players of high human-relative skill blunder rarely even in normal position.

So if skilled players play passively, then it's not unlikely to be perfect.
If there are many thousands of instances where skilled players have played passively, then it's likely a perfect game has been played.

 

“So if skilled players play passively, then it's not unlikely to be perfect.”

 

Based on what? 

Basically what I'm saying is the starting position is solid and passive i.e. not likely to cause a mistake.

If the players do nothing to change this fact, then a mistake is unlikely to happen... at least one large enough for it to change the 32 man tablebase evaluation.

Take Stockfish operating full strength on a super computer. Take ANY player and give him no handicap. Just tell him to play passively. How do you think that will work out?

Both players have to play passively, as in the game I posted.

I've played blitz with full strength stockfish. Not often, just a few times, just for fun.

Surprisingly sometimes I was reaching equal positions into the late middlegame... like... boarderline draw-able (but I had big practical difficulties, then 1-5 moves later I was 100% lost).

That’s JUST Stockfish. Imagine a monster that solved chess. 

A perfect game between players of mismatched skill is highly unlikely. But that's not what I'm talking about, so I don't know why this is relevant.

 

“A perfect game between players of mismatched skill is highly unlikely. But that's not what I'm talking about, so I don't know why this is relevant.”

 

Two equal strength players sit at a table. If on move 7 a player makes a move that a stronger player would be able to exploit, the move isn’t perfect - irrespective of whether his opponent was able to exploit it.

 

Now carry that forward to literally the strongest player possible. A computer that solved chess. We can’t know if a perfect game has been played.

godsofhell1235
USArmyParatrooper wrote:

 

 Two equal strength players sit at a table. If on move 7 a player makes a move that a stronger player would be able to exploit, the move isn’t perfect - irrespective of whether his opponent was able to exploit it.

I agree.

And a perfect player may say a move is losing that we were 100% sure was fine. It would really shock us. That's why we can't know for sure.

But because the nature of the starting position is unlikely to cause mistakes, and because it takes a relatively large mistake to lose (we know this because EGTB can show many positions where being a pawn up isn't enough in an endgame) and because skilled players are unlikely to blunder in symmetrical positions with no potential threats, it's likely a perfect game has been played.

USArmyParatrooper
godsofhell1235 wrote:
USArmyParatrooper wrote:

 

 Two equal strength players sit at a table. If on move 7 a player makes a move that a stronger player would be able to exploit, the move isn’t perfect - irrespective of whether his opponent was able to exploit it.

I agree.

And a perfect player may say a move is losing that we were 100% sure was fine. It would really shock us. That's why we can't know for sure.

But because the nature of the starting position is unlikely to cause mistakes, and because it takes a relatively large mistake to lose (we know this because EGTB can show many positions where being a pawn up isn't enough in an endgame) and because skilled players are unlikely to blunder in symmetrical positions with no potential threats, it's likely a perfect game has been played.

That would mean a player chose every single move that neither Stockfish or “God” could exploit to force a win. 

 

It may have happened. I find it unlikely. We have no idea if that ever happened.

godsofhell1235
USArmyParatrooper wrote:
godsofhell1235 wrote:
USArmyParatrooper wrote:

 

 Two equal strength players sit at a table. If on move 7 a player makes a move that a stronger player would be able to exploit, the move isn’t perfect - irrespective of whether his opponent was able to exploit it.

I agree.

And a perfect player may say a move is losing that we were 100% sure was fine. It would really shock us. That's why we can't know for sure.

But because the nature of the starting position is unlikely to cause mistakes, and because it takes a relatively large mistake to lose (we know this because EGTB can show many positions where being a pawn up isn't enough in an endgame) and because skilled players are unlikely to blunder in symmetrical positions with no potential threats, it's likely a perfect game has been played.

That would mean a player chose every single move that neither Stockfish or “God” could exploit to force a win. 

Correct.

But remember at that point God plays himself. He has to exploit the move against God-level defense.

 

It may have happened. I find it unlikely. We have no idea if that ever happened.

I find it likely, but I guess that's not really important.

I agree we can't know for sure.

 

godsofhell1235

An annoying thing for humans is, some positions which are 0.00 are really hard to play.

So we can be making many practical errors over the course of a game, that make it more likely we'll lose, even though we're not technically screwing up yet, at least from the 32 man EGTB perspective.

I only mention this because all this talk of perfect play is making chess sound easy when in reality it just takes 1 move to lose, and it's really easy to make that 1 move tongue.png

USArmyParatrooper
godsofhell1235 wrote:

An annoying thing for humans is, some positions which are 0.00 are really hard to play.

So we can be making many practical errors over the course of a game, that make it more likely we'll lose, even though we're not technically screwing up yet, at least from the 32 man EGTB perspective.

I only mention this because all this talk of perfect play is making chess sound easy when in reality it just takes 1 move to lose, and it's really easy to make that 1 move

Oh I play a lot of gambits and wild openings (fried liver attack, Evans gambit) and also lines of the French that the analyzer doesn’t like.

 

But my opponents are as inept at chess as I am 🤪

godsofhell1235

My primary openings are QGD and the Spanish (Spanish with both colors).

So I can play pretty well when it's classical and solid like that (by which I mean the engine isn't yelling at me every move).

But I'll also mix it up sometimes. I'm trying out the benko in tournaments now for example. Some of those positions I'm really surprised... sometimes the engine is saying every move (by both of us) is a big mistake tongue.png

USArmyParatrooper
s23bog wrote:
USArmyParatrooper escribió:

 

In the context of the thread topic (solving chess) the difference between a mistake and error are semantics. There is only forced mate or draw by force.

Actually, that really isn't a true dichotomy.  It is better to look at the problem like "There is either a forced mate, or there is not a forced mate."

The game can either end in a win or draw. Is there a third option?

chessspy1

Tables-up maybe?

godsofhell1235

In tournament play there's a double forfeit option a tournament director can do...

I forgot when they use it though.

Not that this matter to the discussion.

troy7915
ponz111 wrote:

You do not seem to understand what "perfect" means as regards to a perfect chess game.

 

  It’s rather simple: you can only pronounce a game perfect after you checked all the lines. If you do not check all the lines, then your proclamation of a perfect move or perfect game has no meaning. In your flawed example, 3. Bb5 was deemed ‘perfect’ as part of a perfect game. Do you know that 3. Bb5 does not end in a forced loss? 

 

  You do not, Kasparov doesn’t know, the strongest engine doesn’t know. If you don’t know, then 3. Bb5 can lose by force, in all the lines, meaning with the best defense. On the other hand, 3. d4 may win by force, with the best defense possible.

 

  If that is a possible result, do you understand that 3. Bb5 may not be a perfect move—if it loses with best defense and there is another move that wins with best defensive lines? 

 

  Same goes for the moves preceding 3. Bb5. In order to label them ‘perfect’ you must make sure they don’t end up in a forced loss and other moves don’t result in a forced win at the same time.

 

 Now, apart from having difficulties intellectually grabbing this concept, you also have difficulties understanding what an error or even a mistake is.

 

  Unless it misses a forced mate, like Kramnik’s stupid blunder against Deep Fritz, what you consider ‘error’ or even a ‘mistake’ may turn out to forcefully win. In order to eliminate that possibility, you would have to check all the lines, in order to know that for sure. 

  You can speculate, which is what you do, claim the consensus of others, but that doesn’t mean anything, as I’m sure there are others who speculate and at the same time are unaware they are speculating.

  It doesn’t matter if the game is short or long: your whole lack of logic is based on the fact that actual theory is sound. But actual theory may prove to be a sham, nobody knows the final outcome/picture. We gathered a lot of speculations in this total darkness and then knuckleheads think they’ve got something other than speculations.

 

 Who knows for sure that Ruy Lopez doesn’t lose by force? Does anyone? If nobody knows, why are you making silly statements about the beginning moves of Ruy Lopez being perfect?

 

 Knowledge is not important here: any moron can accumulate decades of experience, analysis, all that. What matters here is the ability to watch your thinking, every step and turn it takes, to see why it thinks the way it does. It takes diligence to do that, as opposed to being lazy and look the other way thinking: I know, you don’t, end of story. Classic indolence. After all, you can’t teach an old dogs new tricks. It lacks alertness.

 

  As for the starting position, we’ve been over this and you haven’t paid attention: it might as well be that it it’s a total zuzgwang: whoever moves first, loses. Who knows?

USArmyParatrooper

troy7915, you just

 

null

 

 

troy7915

You as well, repeatedly, but to no avail it seems.

ponz111
troy7915 wrote:  Ponz in blue
ponz111 wrote:

You do not seem to understand what "perfect" means as regards to a perfect chess game.

 

  It’s rather simple: you can only pronounce a game perfect after you checked all the lines.  This is not true  I can pronouce a game perfect without checking all the lines. To check all the lines is quite impossible as you know.

 

If you do not check all the lines, then your proclamation of a perfect move or perfect game has no meaning. Sure it has meaning--it means that i believe the particular game is a perfect game even if you do not believe it is a perfect game/

 

In your flawed example, 3. Bb5 was deemed ‘perfect’ as part of a perfect game. Do you know that 3. Bb5 does not end in a forced loss?  Yes. i know that 3.Bb5 does not end in a forced loss. I am somewhat surprised that you do not know this?!

 

  You do not, Kasparov doesn’t know, Actually Kasparov DOES KNOW that 3.Bb5 does not lead to a forced loss. Do you really think you can speak for Kasparov? Do you really think you know more than he does?

the strongest engine doesn’t know. Chess engines do not think like humans do so chess engines do not really know anything.

 

If you don’t know, then 3. Bb5 can lose by force, in all the lines, meaning with the best defense. This is just plain silly and very much against logical reasoning. What i know or do not know has no bearing if 3.Bb5 actually loses by force in all lines against the best defense.  Either 3. Bb5 loses by force in all lines against the best defense or it does not lose by for in all lines against the best defense and it does not matter what i think or what you think. Our thinking cannot control the theoretical result of a move or a game.


 

On the other hand, 3. d4 may win by force, with the best defense possible. It does NOT win by force with the best defense possible. However i did win vs a grandmaster with that move.

 

  If that is a possible result, if what is a possible result?  Do you mean winning with 3. d4?? i know it is a possible result as i beat a former United States Champion with that move. But it was only a possible result because he must have made a mistake somewhere in that game.      do you understand that 3. Bb5 may not be a perfect move—if it loses with best defense and there is another move that wins with best defensive lines? Yes, i understand that 3. Bb5 may not be a perfect move IF it loses with best defense and there is another move that wins with best defensive lines. HOWEVER you used an "if" and the "if" is not the case--it is not reality.

 

  Same goes for the moves preceding 3. Bb5. In order to label them ‘perfect’ you must make sure they don’t end up in a forced loss and other moves don’t result in a forced win at the same time. And i have done this. By using much evidence i have done this. 

 

 Now, apart from having difficulties intellectually grabbing this concept, this is insulting, i fully grasp your concept. I do not agree with it for the reasons stated.

you also have difficulties understanding what an error or even a mistake is. Here you are talking about the definition of a couple of words. We have agreed that a "mistake" has nothing to do with centipawns. A  "mistake" in how we have been using that word is a move which would change the theoretical outcome of the game.

 

  Unless it misses a forced mate, like Kramnik’s stupid blunder against Deep Fritz, what you consider ‘error’ or even a ‘mistake’ may turn out to forcefully win. You cannot speak for me as to what i consider and "error" or a "mistake"

 

In order to eliminate that possibility, here you seem to be assuming what i would call a 'mistake per the Kramnik game. Do not do this--i can speak for myself.

 

you would have to check all the lines, in order to know that for sure. If i had to make a decision on what mistake or mistakes were made in that game then for sure I would check the lines. I would check the lines necessary to make the decision. This does not mean all the millions [or more] lines.

  You can speculate, which is what you do, claim the consensus of others, but that doesn’t mean anything, as I’m sure there are others who speculate and at the same time are unaware they are speculating. I do claim that almost all the grandmasters believe the game of chess is a draw when neither side makes an error. All the grandmasters i know believe this. I do not think you could find more than 1 grandmaster [out of more than 1500] who does not believe chess is a draw when neither side makes an error.  

  It doesn’t matter if the game is short or long: your whole lack of logic is based on the fact that actual theory is sound. NO! It is not based on actual theory being sound. Sometimes actual theory is not sound!  I have made new theory myself and make up my own mind if theory is sound or not? Please do not speak for me as to what i think or what i know!

 

But actual theory may prove to be a sham, this is just what i said above. However you are wrong when you state my reasoning comes from "actual theory"!!!

nobody knows the final outcome/picture. This sentence is ambiguous--the final outcome/picture of what??? I know the final outcome of a whole lot of chess games or positions.

We gathered a lot of speculations in this total darkness and then knuckleheads think they’ve got something other than speculations. You gathered speculations of what i think and you were wrong.  But yes speculations are just considerations of a subject and i have many considerations on various subjects about chess. However my speculations are not in "total darkness" as you [wrongly] try to say. I have a ton of evidence for what i believe is true. You really are using Ad hominem when you say my considerations are in "total darkness" You are simply wrong about that.

 

 Who knows for sure that Ruy Lopez doesn’t lose by force? I know, Almost all grandmasters know.

Does anyone? If nobody knows, why are you making silly statements about the beginning moves of Ruy Lopez being perfect?  the problem with what you are saying is that there ARE a whole lot of very strong players who know that the Ruy Lopez does not lose by force. Also there ARE a whole lot of very strong players who know that the Ruy Lopez results in a draw when neither side makes a mistake which would change the theoretical result of that opening.

 

 Knowledge is not important here: any moron can accumulate decades of experience, analysis, all that. What matters here is the ability to watch your thinking, every step and turn it takes, to see why it thinks the way it does. This is exactly what i do as my chess games prove.

It takes diligence to do that, as opposed to being lazy and look the other way thinking: I know, you don’t, end of story. if i thought the way you state i would not have the very good chess record that i have.  Classic indolence. After all, you can’t teach an old dogs new tricks. It lacks alertness. all you are doing is Ad hominem attacks on my chess knowledge and my ability to think and you are simply wrong.

 

  As for the starting position, we’ve been over this and you haven’t paid attention: it might as well be that it it’s a total zuzgwang: whoever moves first, loses. Who knows?

I know and so do the vast majority of very strong players.