Will computers ever solve chess?

1. d4 e6 { A40 Horwitz Defense } 2. Bf4 c5 3. Nf3 cxd4 4. Qxd4 Nc6 5. Qd1 Nf6 6. e3 b6 7. Bg5 Bb7 8. Bxf6 Qxf6 9. Bb5?? { (-0.73 → -2.87) Blunder. c3 was best. } (9. c3) 9... Qxb2 10. Nbd2?? { (-2.82 → -7.38) Blunder. Bxc6 was best. } (10. Bxc6 Bxc6) 10... Qxb5 11. c3 Ba3 12. c4 Qa5 13. O-O Bb4 14. Nb3 Qa6 15. c5 Bxc5 16. Nxc5 bxc5 17. Ng5 h6 18. Qh5 O-O 19. Nf3 Ne7 20. Ne5 Rad8 21. Rfd1 d5 22. a4 Qd6 23. f4 f6 24. Ng6 Nxg6 25. Qxg6 Qc6 26. h4 d4 27. h5 Qe8 28. Qxe8 Rfxe8 29. exd4 Rxd4 30. Rxd4 cxd4 31. Rd1 e5 32. fxe5 fxe5 33. Kf2 Rd8 34. g4 d3 35. Ke3 e4 36. g5 hxg5 37. h6 Bc6 38. h7+ Kxh7 39. Rh1+?! { (-18.28 → Mate in 24) Checkmate is now unavoidable. Kd2 was best. } (39. Kd2 Kg6 40. a5 g4 41. Rf1 e3+ 42. Kxe3 d2 43. Rd1 g3 44. Kf4 g2 45. a6 Ba4) 39... Kg6 40. Rh6+?! { (-65.25 → Mate in 6) Checkmate is now unavoidable. Rb1 was best. } (40. Rb1 d2) 40... Kxh6 41. a5 Ba4 42. Kd2 Rb8 43. a6 Rb5 44. Ke3 Ra5 45. Kxe4 Rxa6 46. Kxd3 Rb6 47. Kd2 Bd7 48. Ke3 Rb5 49. Kf3 a6 50. Kf2 a5 51. Ke2 Rb4 52. Kd2 a4 53. Ke3 a3 54. Kf2 a2 55. Ke2 a1=Q 56. Kd3 Qa3+ 57. Ke2 Rb2+ 58. Kd1 Qa1# { Black wins by checkmate. } 0-1

I was the black pieces. On a different site it gave me a perfect game where I had no in inaccuracies, mistake, or blunders. And it gave me a centipawn loss of 17. I did it a day later again and it did the same thing but I had a loss of 16. On this site they gave me one in I did it a day later again and it did the same thing but I had a loss of 16. On this site it gave me one in inaccuracy and one mistake move 17.

By the way I’ve played one game where I had no inaccuracies, mistakes or blunders and a centipawn loss of zero! From the English opening. If you have an inaccuracy in the opening line that shouldn’t count.

This may appear to be so, but it would be necessary to prove that a problem cannot be recast in a form suited to a quantum computer.

Loosely, consider the mathematical system consisting of the graph of all possible chess positions, related by legal moves. Now consider some subset of this tree. Such a subset has a set of endpoint values. There is a recursive definition of a subset of tree moves which is optimal, in the sense that each move does not make the result worse.

To select a move, we need to consider the positions after every move and form the recursive subset of positions that can be reached with optimal play, and then evaluate the value of any of those positions (some will turn out to be in a tablebase and only one suffices.

It is conceivable that all of this can be expressed in the form of a boolean function such as "can white win?" (you need two of these, because there are 3 possible results, but each can be done separately).

So I am not 100% convinced it is not possible. It would be necessary to have quite a large quantum computer, enough to represent all possible positions, at least. A few hundred qubits should suffice.

Clearly the hard (not necessarily possible) part is implementing the transitions between positions that represent the rules of chess, and the recursive definitions. This would require more knowledge than the little I have. Indeed it is an unsolved research problem.

It might be worse than this: it might require a quantum state to represent an entire legal game. If so a very large number of qubits are needed. But certainly with this idea, we can imagine the superposition of all legal games (or all legal games with optimal play) in the quantum paradigm.

The perfect game pawn loss is at 1

[Event "ECO MegaMatch II! (1|1)"] [Site "https://lichess.org/USQ46aqO"] [Date "2020.05.23"] [Round "1"] [White "Lc0-dev"] [Black "Stockfish"] [Result "1/2-1/2"] [WhiteElo "?"] [BlackElo "?"] [Variant "Standard"] [TimeControl "60+1"] [ECO "D49"] [Opening "Semi-Slav Defense: Meran Variation, Blumenfeld Variation"] [Termination "Normal"] [Annotator "lichess.org"] 1. d4 d5 2. c4 e6 3. Nc3 Nf6 4. Nf3 c6 5. e3 Nbd7 6. Bd3 dxc4 7. Bxc4 b5 8. Bd3 a6 9. e4 c5 10. e5 cxd4 11. Nxb5 { D49 Semi-Slav Defense: Meran Variation, Blumenfeld Variation } axb5 12. exf6 gxf6 13. O-O Qb6 14. Qe2 b4 15. Rd1 Bc5 16. a4 bxa3 17. bxa3 Ba6 18. Bb2 O-O 19. Bxh7+ Kxh7 20. Qe4+ Kh8 21. Qh4+ Kg7 22. Qg4+ Kh8 23. Qh5+ Kg7 24. Qg4+ Kh8 25. Bxd4 Be2 26. Qh4+ Kg7 27. Qg3+ Kh6 28. Qf4+ Kg7 29. Qg4+ Kh8 30. Qh3+ Kg7 31. Qg3+ Kh6 32. Qf4+ Kg7 33. Bxc5 Qxc5 34. Rxd7 Rxa3 35. Rxa3 Qxa3 36. Qg4+ Kh8 37. Qh3+ Kg7 38. Qg4+ Kh8 39. Qh3+ Kg8 40. Kh1 Bxf3 41. gxf3 Qa1+ 42. Kg2 f5 43. Qh4 Qg7+ 44. Kh1 Qa1+ 45. Kg2 Qg7+ 46. Kh1 Qh7 47. Qf4 Qh3 48. Rd1 Kh7 49. Qg3 Qxg3 50. fxg3 Ra8 51. Rd7 Ra1+ 52. Kg2 Ra2+ 53. Kg1 Ra1+ 54. Kg2 Ra2+ 55. Kg1 Kg6 56. f4 f6 57. Rd6 Re2 58. h3 Kh5 59. Kf1 Rh2 60. Rxe6 Rxh3 61. Kg2 Kg4 62. Re3 Rh7 63. Re8 Rh6 64. Rf8 Kh5 65. Rf7 Kg6 66. Rf8 Rh7 67. Rd8 Kh5 68. Rd6 Rh6 69. Rc6 Rg6 70. Rd6 { The game is a draw. } 1/2-1/2

https://lichess.org/USQ46aqO

Your posts have nothing to do with the topic here.

Never

This century.

Surprised they haven't used some quantum computer with our understandings of the quantum world to figure this out. Shouldn't be to hard for a quantum computer just need to set the proper parameters.

#6851
That is easier said than done. Quantum computers exist, e.g. from IBM. Quantum programming languages exist. Open source chess engines exist. A clever programmer needs to translate the chess engine into a quantum programming language in a way that fully utilises the quantum ability of parallel processing. Then it should run on one or more quantum computers. The translation into a quantum programming language is harder than you think.

Moore's law died several years ago when it ran into physics. While the number of transistors continues to increase, and there is scope for parallelisation, this is not enough for adding a factor of 10^20, say.

#6854
The factor 10^20 must come from quantum computing.
60 years from now computers with transistors will look just as obsolete as the 1945 ENIAC computer with vacuum tubes looks now.

I wonder if the best way for a computer to win is to out compute it's opponent or have a winning result for every situation and the correct move. Perhaps there are more ways as well. 

I hope I would have been already dead when this happens.

Or actually like on the verge of dying cuz I wanna try it out 

No they haven't yet

It will not happen. It is already known what happens when both sides play without error, 

Quantum computers will not change that. 

It is not at all sure that quantum computers will ever be useful for chess. 

But even if in the future they are better than today's best chess engines that will not change the fundamental truth about chess.

The truth is the truth regardless of who knows it or who does not know it.

Probably, actually yes, there is just so many people trying

Nothing has really changed for the outcome of this thread.  Computers are still not going to solve chess in our lifetimes, and 10^46.7 is still a much larger number than people can fathom ...