Will computers ever solve chess?

Sort:
Avatar of lfPatriotGames
N1llectr wrote:
phillidor5949 wrote:
lfPatriotGames wrote:
pawn8888 wrote:

Someone says the number of moves is equal to the known amount of atoms or some other foolishness. The moves have to be sensible and correspond to the previous move, either defending or attacking. There are probably only 2 or 3 good moves to make in a position. People who say these almost infinite numbers are talking nonsense. 


In this position, could you name the 2 or 3 good moves? In this position there IS a finite number of moves, but it seems like the number grows quickly once the first move is made. No matter what, it would help immensely to know which are the 2 or 3 good moves. Sometimes I feel like I had no chance to win a game from the very beginning because at any given time (including the beginning, middle, and end) of all the possible moves I often have trouble identifying the 2 or 3 good ones.

For what it is worth, here are my results after letting the chess engine evaluate the initial position for a number of hours using Droidfish 1.72 featuring Stockfish Nine (9).

Note: All of White's twenty (20) possible opening moves are evaluated at a uniform search depth of 38 plies.

  1. [38] 0.40: 1.d4 Nf6 2.c4 e6 3.Nf3 b6;

  2. [38] 0.40: 1.c4 e5 2.Nc3 Nf6 3.g3 Bb4;

  3. [38] 0.34: 1.e4 c5 2.Nf3 d6 3.Nc3 e5;

  4. [38] 0.32: 1.Nf3 Nf6 2.d4 e6 3.c4 b6;

  5. [38] 0.30: 1.e3 e6 2.b3 b6 3.Nf3 Bb7;

  6. [38] 0.21: 1.c3 e6 2.d4 Nf6 3.Nf3 Be7;

  7. [38] 0.20: 1.g3 d5 2.Nf3 Nf6 3.d4 c5;

  8. [38] 0.18: 1.a3 e5 2.c4 Nf6 3.Nc3 Nc6;

  9. [38] 0.18: 1.Nc3 d5 2.d4 Nf6 3.Bf4 e6;

  10. [38] 0.17: 1.h3 Nf6 2.d4 e6 3.Nf3 d5;

  11. [38] 0.12: 1.f4 d5 2.Nf3 Bf5 3.b3 e6;

  12. [38] 0.11: 1.b3 e5 2.Bb2 Nc6 3.e3 Nf6;

  13. [38] 0.00: 1.a4 Nf6 2.d4 d5 3.e3 c5;

  14. [38] -0.01: 1.d3 d5 2.Bf4 Nf6 3.Nf3 e6;

  15. [38] -0.02: 1.b4 e5 2.a3 d5 3.e3 Nf6;

  16. [38] -0.15: 1.Nh3 d5 2.d4 Bxh3 3.gxh3 c6;

  17. [38] -0.18: 1.h4 c5 2.c4 e6 3.e3 d5;

  18. [38] -0.24: 1.Na3 e5 2.e3 d5 3.d4 Nd7;

  19. [38] -0.29: 1.f3 e5 2.e3 Nc6 3.c4 Nf6;

  20. [38] -0.57: 1.g4 d5 2.e3 Nc6 3.d4 e5;

there are 24 moves

because the knights

20 moves seems about right. 8 pawn moves one square up, 8 two up, plus 4 for the knights, which are included. I'm just happy to know which are the two good moves from that position. Problem is my games seem to always last more than one move. btickler just summed it up better than me.

Avatar of Yudodattome

combinatorics

Avatar of Elroch
N1llectr wrote:
phillidor5949 wrote:
lfPatriotGames wrote:
pawn8888 wrote:

Someone says the number of moves is equal to the known amount of atoms or some other foolishness. The moves have to be sensible and correspond to the previous move, either defending or attacking. There are probably only 2 or 3 good moves to make in a position. People who say these almost infinite numbers are talking nonsense. 


In this position, could you name the 2 or 3 good moves? In this position there IS a finite number of moves, but it seems like the number grows quickly once the first move is made. No matter what, it would help immensely to know which are the 2 or 3 good moves. Sometimes I feel like I had no chance to win a game from the very beginning because at any given time (including the beginning, middle, and end) of all the possible moves I often have trouble identifying the 2 or 3 good ones.

For what it is worth, here are my results after letting the chess engine evaluate the initial position for a number of hours using Droidfish 1.72 featuring Stockfish Nine (9).

Note: All of White's twenty (20) possible opening moves are evaluated at a uniform search depth of 38 plies.

  1. [38] 0.40: 1.d4 Nf6 2.c4 e6 3.Nf3 b6;

  2. [38] 0.40: 1.c4 e5 2.Nc3 Nf6 3.g3 Bb4;

  3. [38] 0.34: 1.e4 c5 2.Nf3 d6 3.Nc3 e5;

  4. [38] 0.32: 1.Nf3 Nf6 2.d4 e6 3.c4 b6;

  5. [38] 0.30: 1.e3 e6 2.b3 b6 3.Nf3 Bb7;

  6. [38] 0.21: 1.c3 e6 2.d4 Nf6 3.Nf3 Be7;

  7. [38] 0.20: 1.g3 d5 2.Nf3 Nf6 3.d4 c5;

  8. [38] 0.18: 1.a3 e5 2.c4 Nf6 3.Nc3 Nc6;

  9. [38] 0.18: 1.Nc3 d5 2.d4 Nf6 3.Bf4 e6;

  10. [38] 0.17: 1.h3 Nf6 2.d4 e6 3.Nf3 d5;

  11. [38] 0.12: 1.f4 d5 2.Nf3 Bf5 3.b3 e6;

  12. [38] 0.11: 1.b3 e5 2.Bb2 Nc6 3.e3 Nf6;

  13. [38] 0.00: 1.a4 Nf6 2.d4 d5 3.e3 c5;

  14. [38] -0.01: 1.d3 d5 2.Bf4 Nf6 3.Nf3 e6;

  15. [38] -0.02: 1.b4 e5 2.a3 d5 3.e3 Nf6;

  16. [38] -0.15: 1.Nh3 d5 2.d4 Bxh3 3.gxh3 c6;

  17. [38] -0.18: 1.h4 c5 2.c4 e6 3.e3 d5;

  18. [38] -0.24: 1.Na3 e5 2.e3 d5 3.d4 Nd7;

  19. [38] -0.29: 1.f3 e5 2.e3 Nc6 3.c4 Nf6;

  20. [38] -0.57: 1.g4 d5 2.e3 Nc6 3.d4 e5;

there are 24 moves

because the knights

Are you playing on a bigger board?

Avatar of stiggling

24.

A strange number to be off by.

For example I could understand someone guessing 32, or 28, but not 24.

Avatar of troy7915
Elroch wrote:
IMBacon wrote:

Lets assume engines right now solve chess. 

Name me one person that is going to have the brain power to memorize all the necessary moves.

lol

Might be tricky to memorise the values of 10^40 positions in 10^11 neurons.

   The solution involves a much smaller number of moves/position. Proving the solution involves that number, but even there many of them are silly positions, no need to memorize them.

  Secondly, no one knows the connection between the number of neurons and the number of memories they can ‘hold’.

Avatar of Elroch

Actually, I can sort of agree with you. Solving chess has a degree of irreducible complexity that means it is O(10^40) or so at least, which is a computationally impractical. But playing perfect chess could conceivably be achieved with a program of much smaller size that used approximate methods that were just good enough to play absolutely perfectly in every position that could be reached. How small the smallest such program would be is very uncertain. If it happened that it was really tiny compared to the 32 piece tablebase, it could in principle possible for a tiny brain with 10^11 neurons to be able to achieve it.

But, let's be serious, it's not. This is a much harder task than merely playing perfect chess against all of the top chess engines, and humans are never going to be able to do that.

Avatar of DiogenesDue
Elroch wrote:

Actually, I can sort of agree with you. Solving chess has a degree of irreducible complexity that means it is O(10^40) or so at least, which is a computationally impractical. But playing perfect chess could conceivably be achieved with a program of much smaller size that used approximate methods that were just good enough to play absolutely perfectly in every position that could be reached. How small the smallest such program would be is very uncertain. If it happened that it was really tiny compared to the 32 piece tablebase, it could in principle possible for a tiny brain with 10^11 neurons to be able to achieve it.

But, let's be serious, it's not. This is a much harder task than merely playing perfect chess against all of the top chess engines, and humans are never going to be able to do that.

I agree with this, but before anyone else picks this up and falsely claims this as some reason why chess can be "solved" with realistically foreseeable technology or in our lifetimes, I will point out that you would still never know if the sub-32 tablebase playing engine was playing perfect chess until you solve the 10^40 and work your way backwards from there.  So in terms of this thread and the definition of a solved game, I think the answer is still "not in our lifetimes or in any realistically foreseeable future either".

Avatar of stiggling

Yeah, there are a lot of ideal conditions and if statements in there. He's just trying to give the devil his due so to speak.

Prior to any relevant calculations that may show it's impossible, and using a bit of imagination, a human could play chess perfectly... ok. But also it's still (very) likely impossible tongue.png

Avatar of troy7915
Elroch wrote:

Actually, I can sort of agree with you. Solving chess has a degree of irreducible complexity that means it is O(10^40) or so at least, which is a computationally impractical. But playing perfect chess could conceivably be achieved with a program of much smaller size that used approximate methods that were just good enough to play absolutely perfectly in every position that could be reached. How small the smallest such program would be is very uncertain. If it happened that it was really tiny compared to the 32 piece tablebase, it could in principle possible for a tiny brain with 10^11 neurons to be able to achieve it.

But, let's be serious, it's not. This is a much harder task than merely playing perfect chess against all of the top chess engines, and humans are never going to be able to do that.

 

   When the World Champion, with the white pieces, on a very good day, if he’s lucky, may be able to get a draw against the top engines out there, to envision a future where a human can play perfect chess against the future top engines belongs rather to the fantasy land.

 

  As for the capacity to memorize, it is drastically improved when psychological problems like fear, anger, all kinds of frictions between billions of images in the brain, come to an end. New possibilities come into being, but usually the brain doesn’t tap into this potential, regardless of the size of one’s intellect, not to be confused with intelligence.

Avatar of r00k226
I think 💭 it is D
Avatar of Titled_Patzer

It is like asking will computers ever solve tick-tack-toe to be a win.

An irrelevant question. 

 

Avatar of Titled_Patzer

The very Nature of the game prohibits a forced win. Observe the latest top level tournaments, where Black is winning the majority of games, a result of White over pressing.

Avatar of Titled_Patzer

Checkers 'solved' after years of number crunching. The ancient game of checkers(or draughts) has been pronounced dead. The game was killed by the publication of a mathematical proof showing that draughts always results in a draw when neither player makes a mistake.

Avatar of Titled_Patzer

  'As for the capacity to memorize, it is drastically improved when psychological problems like fear, anger, all kinds of frictions between billions of images in the brain, come to an end. New possibilities come into being, but usually the brain doesn’t tap into this potential, regardless of the size of one’s intellect, not to be confused with intelligence."- Troy

That's the biggest load of mularky I've seen in a long time. 

"Capacity to memorize is dependent on psychological problems" ??? Nonsense. The statement is entirely "made-up" without a shred of any verification.

Avatar of troy7915
  •   You can see the logical connection: the brain remembers many things in connection with the self-image. Insults, childhood hurts, fears of losing various objects of attachments, like images of people (not actual people, but images of people), various ideas and ideologies one is attached to, millions of beliefs in connection with a perceived center from which one thinks they act, desires, frustrations, various objects of hate (images), and many more.
  •  All that is one continuous movement, sustained by this perceived center.
  • An insight makes the momentum of this movement come to an end.
  •  Just like a computer, this frees up a whole lot of space. 99%.
  •  Two things then happen:

            a) technical memory improves dramatically

            b) the brain is filled with energy.

 

      The brain has the potential to tap into an energy much more powerful than knowledge—before it becomes knowledge. As it functions now, with so many frictions between useless images, the brain can only work with low energy patterns—any pattern is already low energy, so the point is to discover an energy which is not already lowered into a pattern.

 

Avatar of troy7915
Titled_Patzer wrote:

The very Nature of the game prohibits a forced win. Observe the latest top level tournaments, where Black is winning the majority of games, a result of White over pressing.

  That’s the point, if the game would be solved, White wouldn’t over press. That only happens because chess hasn’t been solved. Therefore, for the moment we don’t really know if any side is over pressing or not, with certainty. We are only speculating, which is the only thing we can do, as long as we understand that we’re not doing something else.

Avatar of Titled_Patzer

You are making unlimited assumptions about the "Brain." The brain is the least understood phenomenon of existence, yet you pretend to be an expert on it's workings. Anyone who portrays themselves to understand as such ,"the brain" is no different than the preacher standing on his egg carton.   

Avatar of Titled_Patzer

Smacks of Scientology

You are a Scientologist Troy ? 

Avatar of troy7915

 The brain is not understood because man is trying to understand it ‘from the outside’ as it were, measuring various elements and functions. Such ‘observation’ cannot detect the illusion contained in the pattern itself. 

I’m saying what I’m saying because the brain has actually been through that: patterns coming to an end, new energy tapped into. I’m not making logical assumptions. I only exposed logically an actual experience. You or anybody else can verify it too by undergoing this fundamental change. The potential is the same for most brains. Not trying to convince you of anything. After all, there is no practice to force this change into being. Which is why it’s not happening in the first place.

  I’m just exposing a fact: from an experience that goes beyond the personal realm, the brain has great potential, not only in terms of capacity to accumulate and store knowledge,  but in terms of something much bigger than knowledge.

Avatar of troy7915
Titled_Patzer wrote:

Smacks of Scientology

You are a Scientologist Troy ? 

 I’m nobody, dear Patzer.

Avatar of Guest2208748186
Please Sign Up to comment.

If you need help, please contact our Help and Support team.