Perfect chess engine elo ( 32 men TB) can be within 200 of Stocfish in Tcec LTC conditions

Discussion of anything and everything relating to chess playing software and machines.

Moderators: hgm, Rebel, chrisw

mwyoung
Posts: 2727
Joined: Wed May 12, 2010 10:00 pm

Re: Perfect chess engine elo ( 32 men TB) can be within 200 of Stocfish in Tcec LTC conditions

Post by mwyoung »

duncan wrote: Mon Nov 09, 2020 12:33 pm
BrendanJNorman wrote: Mon Nov 09, 2020 12:03 pm

Right. I don't know why others aren't seeing this (obvious) logic.

A real 32 man tablebase already knows the outcome of the game in the beginning, so what the hell is a "complication trick" (which someone mentioned) going to do to save Stockfish?

What will happen, in *EVERY* game, is this:

1. Players shuffle out opening moves.
2. Stockfish gets an edge (if white, maybe).
3. Stockfish makes ONE...MINOR inaccuracy.
4. 32 man TB announces mate in 56.
5. Stockfish is mated in 31 because without tablebases, he chooses an imprecise path to mate (or getting mated in his case).

100-0 match result.

As you said, SF is clueless in 6 man positions, he'd be clueless and defenseless against a 32 man TB engine.
What happens if Stockfish 's evaluation in regular chess is so good, it does not make significant inaccuracies which will lead to a loss against 32 man TB.

In that case, you will have to target's stockfish's weak point that it cannot see beyond 15 moves. The job of 32 man Tb will be to move the game into positions where it is essential to calculate beyond 15 moves.
Why would you think this is possible! And this is regular chess. When stockfish can not show this kind of accuracy in endgames.
"The worst thing that can happen to a forum is a running wild attacking moderator(HGM) who is not corrected by the community." - Ed Schröder
But my words like silent raindrops fell. And echoed in the wells of silence.
mwyoung
Posts: 2727
Joined: Wed May 12, 2010 10:00 pm

Re: Perfect chess engine elo ( 32 men TB) can be within 200 of Stocfish in Tcec LTC conditions

Post by mwyoung »

BrendanJNorman wrote: Mon Nov 09, 2020 12:03 pm
mwyoung wrote: Mon Nov 09, 2020 9:53 am I thought it would be interesting to show just how weak Stockfish would play against a 32 man tablebase. And we can show this by the table bases we do have. This is just one example. But as we know their are countless positions with 6 man positions that Stockfish is clueless. Let alone more complicated positions. And if we could scale this to a 32 man TB. Stockfish would have no chance.

[d]5q1k/6r1/8/8/6R1/2Q5/2K5/8 w - - 0 7

Stockfish says nothing to see here. This positions is a dead draw.

5: 5. Polasek 1-0, Cs. sach 1/2010 1981
5q1k/6r1/8/8/6R1/2Q5/2K5/8 w - - 0 1

Analysis by Stockfish 081120:

7.Rxg7 Qf2+ 8.Qd2 Qf5+ 9.Qd3 Qc5+ 10.Kb3 Kxg7 11.Qc3+ Qxc3+ 12.Kxc3 Kg6 13.Kd3 Kh6 14.Kc2 Kg5 15.Kd2 Kf6 16.Ke2 Kg7 17.Kd1 Kh6 18.Ke1 Kh5 19.Ke2 Kg4 20.Ke3 Kg3 21.Kd3 Kg2 22.Ke4 Kh3 23.Kf4 Kg2
The position is equal: = (0.00) Depth: 81/49 00:19:14 68081MN
(, 09.11.2020)

But the god like power of table bases tells a much different story. And this is just with a 6 man position! :shock: :lol:

[pgn][Event "?"]
[Site "?"]
[Date "????.??.??"]
[Round "?"]
[White "New game"]
[Black "?"]
[Result "1-0"]
[SetUp "1"]
[FEN "5q1k/6r1/8/8/6R1/2Q5/2K5/8 w - - 0 7"]
[PlyCount "115"]

{[#]} 7. Rh4+ {[%eval 27968,1] [%emt 0:00:06]} Kg8 {[%eval 27968,1] [%emt 0:00:
01]} 8. Qc4+ {[%eval 27969,1] [%emt 0:00:01]} Rf7 {[%eval 27969,1] [%emt 0:00:
01]} 9. Rg4+ {[%eval 27970,1] [%emt 0:00:01]} Kh8 {[%eval 27970,1] [%emt 0:00:
01]} 10. Qd4+ {[%eval 27971,1] [%emt 0:00:01]} Rg7 {[%eval 27971,1] [%emt 0:00:
01]} 11. Rh4+ {[%eval 27972,1] [%emt 0:00:01]} Kg8 {[%eval 27972,1] [%emt 0:00:
01]} 12. Qd5+ {[%eval 27973,1] [%emt 0:00:00]} Rf7 {[%eval 27973,1] [%emt 0:00:
01]} 13. Rg4+ {[%eval 27974,1] [%emt 0:00:00]} Kh8 {[%eval 27974,1] [%emt 0:00:
00]} 14. Qe5+ {[%eval 27975,1] [%emt 0:00:00]} Rg7 {[%eval 27975,1] [%emt 0:00:
00]} 15. Rh4+ {[%eval 27976,1] [%emt 0:00:00]} Kg8 {[%eval 27976,1] [%emt 0:00:
00]} 16. Qe6+ {[%eval 27977,1] [%emt 0:00:00]} Rf7 {[%eval 27977,1] [%emt 0:00:
00]} 17. Kb1 {[%eval 27978,1] [%emt 0:00:00]} Qb8+ {[%eval 27978,1] [%emt 0:00:
00]} 18. Kc1 {[%eval 27979,1] [%emt 0:00:00]} Qc7+ {[%eval 27979,1] [%emt 0:00:
00]} 19. Rc4 {[%eval 27980,1] [%emt 0:00:00]} Qe7 {[%eval 27980,1] [%emt 0:00:
00]} 20. Qg4+ {[%eval 27981,1] [%emt 0:00:00]} Kh8 {[%eval 27981,0] [%emt 0:00:
00]} 21. Qh3+ {[%eval 27982,1] [%emt 0:00:00]} Kg8 {[%eval 27982,1] [%emt 0:00:
00]} 22. Qg3+ {[%eval 27983,1] [%emt 0:00:00]} Rg7 {[%eval 27983,1] [%emt 0:00:
00]} 23. Rc8+ {[%eval 27984,1] [%emt 0:00:00]} Kh7 {[%eval 27984,1] [%emt 0:00:
00]} 24. Qh2+ {[%eval 27985,1] [%emt 0:00:00]} Kg6 {[%eval 27985,1] [%emt 0:00:
00]} 25. Qc2+ {[%eval 27986,0] [%emt 0:00:00]} Kf6 {[%eval 27986,1] [%emt 0:00:
00]} 26. Qf2+ {[%eval 27987,1] [%emt 0:00:00]} Kg6 {[%eval 27987,1] [%emt 0:00:
00]} 27. Rc6+ {[%eval 27988,1] [%emt 0:00:00]} Kg5 {[%eval 27988,0] [%emt 0:00:
00]} 28. Rc5+ {[%eval 27989,1] [%emt 0:00:00]} Kg4 {[%eval 27989,1] [%emt 0:00:
00]} 29. Qg2+ {[%eval 27990,0] [%emt 0:00:00]} Kf4 {[%eval 27990,0] [%emt 0:00:
00]} 30. Qh2+ {[%eval 27991,1] [%emt 0:00:00]} Kg4 {[%eval 27991,1] [%emt 0:00:
00]} 31. Qh5+ {[%eval 27992,1] [%emt 0:00:00]} Kg3 {[%eval 27992,1] [%emt 0:00:
00]} 32. Rc3+ {[%eval 27993,1] [%emt 0:00:00]} Kg2 {[%eval 27993,1] [%emt 0:00:
00]} 33. Rc2+ {[%eval 27994,1] [%emt 0:00:00]} Kg3 {[%eval 27994,1] [%emt 0:00:
00]} 34. Qh2+ {[%eval 27995,1] [%emt 0:00:00]} Kg4 {[%eval 27995,1] [%emt 0:00:
00]} 35. Rg2+ {[%eval 27996,0] [%emt 0:00:00]} Kf5 {[%eval 27996,0] [%emt 0:00:
00]} 36. Qh3+ {[%eval 27997,1] [%emt 0:00:00]} Kf4 {[%eval 27997,1] [%emt 0:00:
00]} 37. Rf2+ {[%eval 27998,1] [%emt 0:00:00]} Ke4 {[%eval 27998,0] [%emt 0:00:
00]} 38. Re2+ {[%eval 27999,1] [%emt 0:00:00]} Kd4 {[%eval 27999,0] [%emt 0:00:
00]} 39. Rxe7 {[%eval 27996,1] [%emt 0:00:00]} Rxe7 {[%eval 27976,1] [%emt 0:
00:00]} 40. Kd2 {[%eval 27977,1] [%emt 0:00:01]} Rc7 {[%eval 27977,0] [%emt 0:
00:00]} 41. Qd3+ {[%eval 27978,0] [%emt 0:00:00]} Kc5 {[%eval 27978,1] [%emt 0:
00:00]} 42. Kc3 {[%eval 27979,0] [%emt 0:00:00]} Rc6 {[%eval 27979,1] [%emt 0:
00:00]} 43. Qd4+ {[%eval 27980,0] [%emt 0:00:00]} Kb5+ {[%eval 27980,1] [%emt
0:00:00]} 44. Kb3 {[%eval 27981,1] [%emt 0:00:00]} Rc5 {[%eval 27981,1] [%emt
0:00:00]} 45. Qd6 {[%eval 27982,0] [%emt 0:00:00]} Rc6 {[%eval 27982,1] [%emt
0:00:00]} 46. Qb4+ {[%eval 27983,1] [%emt 0:00:00]} Ka6 {[%eval 27983,1] [%emt
0:00:00]} 47. Ka3 {[%eval 27984,0] [%emt 0:00:00]} Rc1 {[%eval 27984,0] [%emt
0:00:00]} 48. Qe4 {[%eval 27985,1] [%emt 0:00:00]} Rc3+ {[%eval 27985,0] [%emt
0:00:00]} 49. Ka4 {[%eval 27986,0] [%emt 0:00:00]} Rc5 {[%eval 27986,0] [%emt
0:00:00]} 50. Qb4 {[%eval 27987,1] [%emt 0:00:00]} Rc1 {[%eval 27987,0] [%emt
0:00:00]} 51. Qb5+ {[%eval 27988,0] [%emt 0:00:01]} Ka7 {[%eval 27988,1] [%emt
0:00:00]} 52. Qd3 {[%eval 27989,0] [%emt 0:00:00]} Kb8 {[%eval 27989,0] [%emt
0:00:00]} 53. Qe4 {[%eval 27990,0] [%emt 0:00:00]} Rc7 {[%eval 27990,1] [%emt
0:00:00]} 54. Kb5 {[%eval 27991,1] [%emt 0:00:00]} Ka7 {[%eval 27991,1] [%emt
0:00:00]} 55. Qh1 {[%eval 27992,1] [%emt 0:00:00]} Rb7+ {[%eval 27992,0] [%emt
0:00:00]} 56. Kc6 {[%eval 27993,1] [%emt 0:00:00]} Kb8 {[%eval 27993,0] [%emt
0:00:00]} 57. Qh8+ {[%eval 27994,1] [%emt 0:00:00]} Ka7 {[%eval 27994,1] [%emt
0:00:00]} 58. Qd8 {[%eval 27995,1] [%emt 0:00:00]} Rf7 {[%eval 27995,0] [%emt
0:00:00]} 59. Qb6+ {[%eval 27996,0] [%emt 0:00:00]} Ka8 {[%eval 27996,1] [%emt
0:00:00]} 60. Qb3 {[%eval 27997,0] [%emt 0:00:00]} Rg7 {[%eval 27997,1] [%emt
0:00:00]} 61. Qa2+ {[%eval 27998,0] [%emt 0:00:00]} Kb8 {[%eval 27998,1] [%emt
0:00:00]} 62. Qb2+ {[%eval 27999,1] [%emt 0:00:00]} Kc8 {[%eval 27999,0] [%emt
0:00:00]} 63. Qxg7 {[%eval 27999,1] [%emt 0:00:01]} Kb8 {[%eval 27999,0] [%emt
0:00:00]} 64. Qb7# {[%eval 32766,1] [%emt 0:00:00]} 1-0

[/pgn]
Right. I don't know why others aren't seeing this (obvious) logic.

A real 32 man tablebase already knows the outcome of the game in the beginning, so what the hell is a "complication trick" (which someone mentioned) going to do to save Stockfish?

What will happen, in *EVERY* game, is this:

1. Players shuffle out opening moves.
2. Stockfish gets an edge (if white, maybe).
3. Stockfish makes ONE...MINOR inaccuracy.
4. 32 man TB announces mate in 56.
5. Stockfish is mated in 31 because without tablebases, he chooses an imprecise path to mate (or getting mated in his case).

100-0 match result.

As you said, SF is clueless in 6 man positions, he'd be clueless and defenseless against a 32 man TB engine.
That is because many people here reject logic and reason. And are more interested in self delusion. Even when faced with Logic and Reason.

I think they what to believe that today's chess engines are so good, and have a god like power. Because they wanted it to be true. And will defend this idea no matter what!

People have a hard time understanding the true nature of chess. Chess is a 100% tactical game. There is nothing really called positional play. And chess has only 3 true evaluations. White wins in x, Draw, Black wins in x

A 32 man table base is equal to a full width search to the end of the game, meaning this is like searching 1000's or more ply to stockfish's few 10's of ply in a type B search. :lol:
"The worst thing that can happen to a forum is a running wild attacking moderator(HGM) who is not corrected by the community." - Ed Schröder
But my words like silent raindrops fell. And echoed in the wells of silence.
duncan
Posts: 12038
Joined: Mon Jul 07, 2008 10:50 pm

Re: Perfect chess engine elo ( 32 men TB) can be within 200 of Stocfish in Tcec LTC conditions

Post by duncan »

mwyoung wrote: Mon Nov 09, 2020 2:32 pm
duncan wrote: Mon Nov 09, 2020 12:33 pm
BrendanJNorman wrote: Mon Nov 09, 2020 12:03 pm

Right. I don't know why others aren't seeing this (obvious) logic.

A real 32 man tablebase already knows the outcome of the game in the beginning, so what the hell is a "complication trick" (which someone mentioned) going to do to save Stockfish?

What will happen, in *EVERY* game, is this:

1. Players shuffle out opening moves.
2. Stockfish gets an edge (if white, maybe).
3. Stockfish makes ONE...MINOR inaccuracy.
4. 32 man TB announces mate in 56.
5. Stockfish is mated in 31 because without tablebases, he chooses an imprecise path to mate (or getting mated in his case).

100-0 match result.

As you said, SF is clueless in 6 man positions, he'd be clueless and defenseless against a 32 man TB engine.
What happens if Stockfish 's evaluation in regular chess is so good, it does not make significant inaccuracies which will lead to a loss against 32 man TB.

In that case, you will have to target's stockfish's weak point that it cannot see beyond 15 moves. The job of 32 man Tb will be to move the game into positions where it is essential to calculate beyond 15 moves.
Why would you think this is possible! And this is regular chess. When stockfish can not show this kind of accuracy in endgames.
I thought 99% of positions in 7 or less tablebase, stockfish will draw all drawable and win all winnable positions when playing against a tablebase.
mwyoung
Posts: 2727
Joined: Wed May 12, 2010 10:00 pm

Re: Perfect chess engine elo ( 32 men TB) can be within 200 of Stocfish in Tcec LTC conditions

Post by mwyoung »

duncan wrote: Mon Nov 09, 2020 3:56 pm
mwyoung wrote: Mon Nov 09, 2020 2:32 pm
duncan wrote: Mon Nov 09, 2020 12:33 pm
BrendanJNorman wrote: Mon Nov 09, 2020 12:03 pm

Right. I don't know why others aren't seeing this (obvious) logic.

A real 32 man tablebase already knows the outcome of the game in the beginning, so what the hell is a "complication trick" (which someone mentioned) going to do to save Stockfish?

What will happen, in *EVERY* game, is this:

1. Players shuffle out opening moves.
2. Stockfish gets an edge (if white, maybe).
3. Stockfish makes ONE...MINOR inaccuracy.
4. 32 man TB announces mate in 56.
5. Stockfish is mated in 31 because without tablebases, he chooses an imprecise path to mate (or getting mated in his case).

100-0 match result.

As you said, SF is clueless in 6 man positions, he'd be clueless and defenseless against a 32 man TB engine.
What happens if Stockfish 's evaluation in regular chess is so good, it does not make significant inaccuracies which will lead to a loss against 32 man TB.

In that case, you will have to target's stockfish's weak point that it cannot see beyond 15 moves. The job of 32 man Tb will be to move the game into positions where it is essential to calculate beyond 15 moves.
Why would you think this is possible! And this is regular chess. When stockfish can not show this kind of accuracy in endgames.
I thought 99% of positions in 7 or less tablebase, stockfish will draw all drawable and win all winnable positions when playing against a tablebase.
Where did you find this B.S. stat. And how does this help stockfish from losing all games against a 32 man TB. Stockfish would need to show 100% in 7 man to even have a chance.

But by your stat. Stockfish would be total wrong in 4,238,368,356,673 positions of 7 man TB. This is not a small hole, but a rip you could drive a truck through.
"The worst thing that can happen to a forum is a running wild attacking moderator(HGM) who is not corrected by the community." - Ed Schröder
But my words like silent raindrops fell. And echoed in the wells of silence.
duncan
Posts: 12038
Joined: Mon Jul 07, 2008 10:50 pm

Re: Perfect chess engine elo ( 32 men TB) can be within 200 of Stocfish in Tcec LTC conditions

Post by duncan »

mwyoung wrote: Mon Nov 09, 2020 4:12 pm
duncan wrote: Mon Nov 09, 2020 3:56 pm
mwyoung wrote: Mon Nov 09, 2020 2:32 pm
duncan wrote: Mon Nov 09, 2020 12:33 pm
BrendanJNorman wrote: Mon Nov 09, 2020 12:03 pm

Right. I don't know why others aren't seeing this (obvious) logic.

A real 32 man tablebase already knows the outcome of the game in the beginning, so what the hell is a "complication trick" (which someone mentioned) going to do to save Stockfish?

What will happen, in *EVERY* game, is this:

1. Players shuffle out opening moves.
2. Stockfish gets an edge (if white, maybe).
3. Stockfish makes ONE...MINOR inaccuracy.
4. 32 man TB announces mate in 56.
5. Stockfish is mated in 31 because without tablebases, he chooses an imprecise path to mate (or getting mated in his case).

100-0 match result.

As you said, SF is clueless in 6 man positions, he'd be clueless and defenseless against a 32 man TB engine.
What happens if Stockfish 's evaluation in regular chess is so good, it does not make significant inaccuracies which will lead to a loss against 32 man TB.

In that case, you will have to target's stockfish's weak point that it cannot see beyond 15 moves. The job of 32 man Tb will be to move the game into positions where it is essential to calculate beyond 15 moves.
Why would you think this is possible! And this is regular chess. When stockfish can not show this kind of accuracy in endgames.
I thought 99% of positions in 7 or less tablebase, stockfish will draw all drawable and win all winnable positions when playing against a tablebase.
Where did you find this B.S. stat. And how does this help stockfish from losing all games against a 32 man TB. Stockfish would need to show 100% in 7 man to even have a chance.

But by your stat. Stockfish would be total wrong in 4,238,368,356,673 positions of 7 man TB. This is not a small hole, but a rip you could drive a truck through.

I did not find this stat anywhere. Just an impression which could easily be wrong. Anybody can give accurate stats ?


But if stockfish In 99% of positions in 32 or less tablebase, stockfish will draw all drawable and win all winnable positions when playing against a tablebase, would it not mean it would score a lot of draws in matches against a 32 man TB?
mwyoung
Posts: 2727
Joined: Wed May 12, 2010 10:00 pm

Re: Perfect chess engine elo ( 32 men TB) can be within 200 of Stocfish in Tcec LTC conditions

Post by mwyoung »

duncan wrote: Mon Nov 09, 2020 4:26 pm
mwyoung wrote: Mon Nov 09, 2020 4:12 pm
duncan wrote: Mon Nov 09, 2020 3:56 pm
mwyoung wrote: Mon Nov 09, 2020 2:32 pm
duncan wrote: Mon Nov 09, 2020 12:33 pm
BrendanJNorman wrote: Mon Nov 09, 2020 12:03 pm

Right. I don't know why others aren't seeing this (obvious) logic.

A real 32 man tablebase already knows the outcome of the game in the beginning, so what the hell is a "complication trick" (which someone mentioned) going to do to save Stockfish?

What will happen, in *EVERY* game, is this:

1. Players shuffle out opening moves.
2. Stockfish gets an edge (if white, maybe).
3. Stockfish makes ONE...MINOR inaccuracy.
4. 32 man TB announces mate in 56.
5. Stockfish is mated in 31 because without tablebases, he chooses an imprecise path to mate (or getting mated in his case).

100-0 match result.

As you said, SF is clueless in 6 man positions, he'd be clueless and defenseless against a 32 man TB engine.
What happens if Stockfish 's evaluation in regular chess is so good, it does not make significant inaccuracies which will lead to a loss against 32 man TB.

In that case, you will have to target's stockfish's weak point that it cannot see beyond 15 moves. The job of 32 man Tb will be to move the game into positions where it is essential to calculate beyond 15 moves.
Why would you think this is possible! And this is regular chess. When stockfish can not show this kind of accuracy in endgames.
I thought 99% of positions in 7 or less tablebase, stockfish will draw all drawable and win all winnable positions when playing against a tablebase.
Where did you find this B.S. stat. And how does this help stockfish from losing all games against a 32 man TB. Stockfish would need to show 100% in 7 man to even have a chance.

But by your stat. Stockfish would be total wrong in 4,238,368,356,673 positions of 7 man TB. This is not a small hole, but a rip you could drive a truck through.

I did not find this stat anywhere. Just an impression which could easily be wrong. Anybody can give accurate stats ?


But if stockfish In 99% of positions in 32 or less tablebase, stockfish will draw all drawable and win all winnable positions when playing against a tablebase, would it not mean it would score a lot of draws in matches against a 32 man TB?
Why on earth would you think that your fictional 99% stat will hold up as the game gets more complicated. If Stockfish can only be right 99% of the time in simplified endgames. This will clearly not be true at 8 man, 16 man, or 32 man positions.

But do the math. If Stockfish has a 1% chance of making a losing move in every move of a game. This does not help your argument, but mine! :lol:

The math says we are looking at a world wide fish fry.
"The worst thing that can happen to a forum is a running wild attacking moderator(HGM) who is not corrected by the community." - Ed Schröder
But my words like silent raindrops fell. And echoed in the wells of silence.
duncan
Posts: 12038
Joined: Mon Jul 07, 2008 10:50 pm

Re: Perfect chess engine elo ( 32 men TB) can be within 200 of Stocfish in Tcec LTC conditions

Post by duncan »

mwyoung wrote: Mon Nov 09, 2020 4:33 pm
duncan wrote: Mon Nov 09, 2020 4:26 pm
mwyoung wrote: Mon Nov 09, 2020 4:12 pm
duncan wrote: Mon Nov 09, 2020 3:56 pm
mwyoung wrote: Mon Nov 09, 2020 2:32 pm
duncan wrote: Mon Nov 09, 2020 12:33 pm
BrendanJNorman wrote: Mon Nov 09, 2020 12:03 pm

Right. I don't know why others aren't seeing this (obvious) logic.

A real 32 man tablebase already knows the outcome of the game in the beginning, so what the hell is a "complication trick" (which someone mentioned) going to do to save Stockfish?

What will happen, in *EVERY* game, is this:

1. Players shuffle out opening moves.
2. Stockfish gets an edge (if white, maybe).
3. Stockfish makes ONE...MINOR inaccuracy.
4. 32 man TB announces mate in 56.
5. Stockfish is mated in 31 because without tablebases, he chooses an imprecise path to mate (or getting mated in his case).

100-0 match result.

As you said, SF is clueless in 6 man positions, he'd be clueless and defenseless against a 32 man TB engine.
What happens if Stockfish 's evaluation in regular chess is so good, it does not make significant inaccuracies which will lead to a loss against 32 man TB.

In that case, you will have to target's stockfish's weak point that it cannot see beyond 15 moves. The job of 32 man Tb will be to move the game into positions where it is essential to calculate beyond 15 moves.
Why would you think this is possible! And this is regular chess. When stockfish can not show this kind of accuracy in endgames.
I thought 99% of positions in 7 or less tablebase, stockfish will draw all drawable and win all winnable positions when playing against a tablebase.
Where did you find this B.S. stat. And how does this help stockfish from losing all games against a 32 man TB. Stockfish would need to show 100% in 7 man to even have a chance.

But by your stat. Stockfish would be total wrong in 4,238,368,356,673 positions of 7 man TB. This is not a small hole, but a rip you could drive a truck through.

I did not find this stat anywhere. Just an impression which could easily be wrong. Anybody can give accurate stats ?


But if stockfish In 99% of positions in 32 or less tablebase, stockfish will draw all drawable and win all winnable positions when playing against a tablebase, would it not mean it would score a lot of draws in matches against a 32 man TB?
Why on earth would you think that your fictional 99% stat will hold up as the game gets more complicated. If Stockfish can only be right 99% of the time in simplified endgames. This will clearly not be true at 8 man, 16 man, or 32 man positions.

But do the math. If Stockfish has a 1% chance of making a losing move in every move of a game. This does not help your argument, but mine! :lol:
Your main point is right but If Stockfish has a 1% chance of making a losing move in every move of a game and an average game is 100 moves, then would there be no losing moves in every fourth game / fifth game or so ? if correct it will mean 20/25% draws against a 32 man tablebase.
mwyoung
Posts: 2727
Joined: Wed May 12, 2010 10:00 pm

Re: Perfect chess engine elo ( 32 men TB) can be within 200 of Stocfish in Tcec LTC conditions

Post by mwyoung »

duncan wrote: Mon Nov 09, 2020 5:03 pm
mwyoung wrote: Mon Nov 09, 2020 4:33 pm
duncan wrote: Mon Nov 09, 2020 4:26 pm
mwyoung wrote: Mon Nov 09, 2020 4:12 pm
duncan wrote: Mon Nov 09, 2020 3:56 pm
mwyoung wrote: Mon Nov 09, 2020 2:32 pm
duncan wrote: Mon Nov 09, 2020 12:33 pm
BrendanJNorman wrote: Mon Nov 09, 2020 12:03 pm

Right. I don't know why others aren't seeing this (obvious) logic.

A real 32 man tablebase already knows the outcome of the game in the beginning, so what the hell is a "complication trick" (which someone mentioned) going to do to save Stockfish?

What will happen, in *EVERY* game, is this:

1. Players shuffle out opening moves.
2. Stockfish gets an edge (if white, maybe).
3. Stockfish makes ONE...MINOR inaccuracy.
4. 32 man TB announces mate in 56.
5. Stockfish is mated in 31 because without tablebases, he chooses an imprecise path to mate (or getting mated in his case).

100-0 match result.

As you said, SF is clueless in 6 man positions, he'd be clueless and defenseless against a 32 man TB engine.
What happens if Stockfish 's evaluation in regular chess is so good, it does not make significant inaccuracies which will lead to a loss against 32 man TB.

In that case, you will have to target's stockfish's weak point that it cannot see beyond 15 moves. The job of 32 man Tb will be to move the game into positions where it is essential to calculate beyond 15 moves.
Why would you think this is possible! And this is regular chess. When stockfish can not show this kind of accuracy in endgames.
I thought 99% of positions in 7 or less tablebase, stockfish will draw all drawable and win all winnable positions when playing against a tablebase.
Where did you find this B.S. stat. And how does this help stockfish from losing all games against a 32 man TB. Stockfish would need to show 100% in 7 man to even have a chance.

But by your stat. Stockfish would be total wrong in 4,238,368,356,673 positions of 7 man TB. This is not a small hole, but a rip you could drive a truck through.

I did not find this stat anywhere. Just an impression which could easily be wrong. Anybody can give accurate stats ?


But if stockfish In 99% of positions in 32 or less tablebase, stockfish will draw all drawable and win all winnable positions when playing against a tablebase, would it not mean it would score a lot of draws in matches against a 32 man TB?
Why on earth would you think that your fictional 99% stat will hold up as the game gets more complicated. If Stockfish can only be right 99% of the time in simplified endgames. This will clearly not be true at 8 man, 16 man, or 32 man positions.

But do the math. If Stockfish has a 1% chance of making a losing move in every move of a game. This does not help your argument, but mine! :lol:
Your main point is right but If Stockfish has a 1% chance of making a losing move in every move of a game and an average game is 100 moves, then would there be no losing moves in every fourth game / fifth game or so ? if correct it will mean 20/25% draws against a 32 man tablebase.
Not correct, why do you think a perfect engine could only make the game last 100 moves. Again people have no understanding of what perfect play really looks like. :lol:

But lets go with your theory and numbers again. And see if this hold up to "Perfect chess engine elo ( 32 men TB) can be within 200 of Stocfish in Tcec LTC conditions"

Under your stats you clearly agree the above can not be correct. A 25% draw rate and 0% win rate = 338 Elo, or a 20% draw rate and a 0% win rate = 382 Elo.
"The worst thing that can happen to a forum is a running wild attacking moderator(HGM) who is not corrected by the community." - Ed Schröder
But my words like silent raindrops fell. And echoed in the wells of silence.
Alayan
Posts: 550
Joined: Tue Nov 19, 2019 8:48 pm
Full name: Alayan Feh

Re: Perfect chess engine elo ( 32 men TB) can be within 200 of Stocfish in Tcec LTC conditions

Post by Alayan »

The existence of chess positions that are very difficult not to lose for current Stockfish doesn't mean the TB32 player has a way to force those positions to happen.

TB32 using adversarial search, that is, running Stockfish to discover all Stockfish blindspots (lost positions it thinks drawn or good for itself), would be the best attempt. How successful it would be would depend on how much mistakes Stockfish would actually do and on pure randomness as Stockfish wouldn't have deterministic play, meaning that no swindle path is guaranteed to happen and there may or may not be backup swindle attempts.

Current SF is probably still swindlable, but while a few years ago it was quite easy for an ICCF player to demonstrate a win against an unassisted engine, it's becoming much harder. The ceiling for never losing from the start position as white (then as black) is much lower than the ceiling for never losing a drawn position, engines and hardware will reach it in reasonable time controls in the coming years.

Instead of arguing further, I suggest we play a game. I'll run the latest SF as of now (no update during the game), 3 threads, 2GB hash. I'll play the move from the first depth completing after 1B nodes. Let's see how easily you can beat it. Any hardware, book, engine is fair, but if you take up the challenge, try and play new moves within a few days of mines.

1st move : 1. e4 (1.04 B nodes, +0.12)
Last edited by Alayan on Mon Nov 09, 2020 5:43 pm, edited 1 time in total.
BrendanJNorman
Posts: 2582
Joined: Mon Feb 08, 2016 12:43 am
Full name: Brendan J Norman

Re: Perfect chess engine elo ( 32 men TB) can be within 200 of Stocfish in Tcec LTC conditions

Post by BrendanJNorman »

duncan wrote: Mon Nov 09, 2020 5:03 pm
mwyoung wrote: Mon Nov 09, 2020 4:33 pm
duncan wrote: Mon Nov 09, 2020 4:26 pm
mwyoung wrote: Mon Nov 09, 2020 4:12 pm
duncan wrote: Mon Nov 09, 2020 3:56 pm
mwyoung wrote: Mon Nov 09, 2020 2:32 pm
duncan wrote: Mon Nov 09, 2020 12:33 pm
BrendanJNorman wrote: Mon Nov 09, 2020 12:03 pm

Right. I don't know why others aren't seeing this (obvious) logic.

A real 32 man tablebase already knows the outcome of the game in the beginning, so what the hell is a "complication trick" (which someone mentioned) going to do to save Stockfish?

What will happen, in *EVERY* game, is this:

1. Players shuffle out opening moves.
2. Stockfish gets an edge (if white, maybe).
3. Stockfish makes ONE...MINOR inaccuracy.
4. 32 man TB announces mate in 56.
5. Stockfish is mated in 31 because without tablebases, he chooses an imprecise path to mate (or getting mated in his case).

100-0 match result.

As you said, SF is clueless in 6 man positions, he'd be clueless and defenseless against a 32 man TB engine.
What happens if Stockfish 's evaluation in regular chess is so good, it does not make significant inaccuracies which will lead to a loss against 32 man TB.

In that case, you will have to target's stockfish's weak point that it cannot see beyond 15 moves. The job of 32 man Tb will be to move the game into positions where it is essential to calculate beyond 15 moves.
Why would you think this is possible! And this is regular chess. When stockfish can not show this kind of accuracy in endgames.
I thought 99% of positions in 7 or less tablebase, stockfish will draw all drawable and win all winnable positions when playing against a tablebase.
Where did you find this B.S. stat. And how does this help stockfish from losing all games against a 32 man TB. Stockfish would need to show 100% in 7 man to even have a chance.

But by your stat. Stockfish would be total wrong in 4,238,368,356,673 positions of 7 man TB. This is not a small hole, but a rip you could drive a truck through.

I did not find this stat anywhere. Just an impression which could easily be wrong. Anybody can give accurate stats ?


But if stockfish In 99% of positions in 32 or less tablebase, stockfish will draw all drawable and win all winnable positions when playing against a tablebase, would it not mean it would score a lot of draws in matches against a 32 man TB?
Why on earth would you think that your fictional 99% stat will hold up as the game gets more complicated. If Stockfish can only be right 99% of the time in simplified endgames. This will clearly not be true at 8 man, 16 man, or 32 man positions.

But do the math. If Stockfish has a 1% chance of making a losing move in every move of a game. This does not help your argument, but mine! :lol:
Your main point is right but If Stockfish has a 1% chance of making a losing move in every move of a game and an average game is 100 moves, then would there be no losing moves in every fourth game / fifth game or so ? if correct it will mean 20/25% draws against a 32 man tablebase.
Let me make this idea simple for you.

Kai is a statistician, he is very good at compiling data and making predictions about the future.

Probably one of the brightest minds here when it comes to such things.

In his field, and with the right data, it may appear like he can PREDICT THE FUTURE.

So here's my question:

Say we placed Kai head-to-head with God himself in a contest to predict the future...

Who do you think would come out on top when both were given very complicated (chess-level) data to make predictions from?

Here's the answer: God wins EVERY time, because God knows the answers without even thinking. His knowledge is omnipotent.

This is Stockfish vs 32 man tablebase as well. Stockfish is very good, but TBs are perfect and omnipotent.

Who would win a fight? 1985 Mike Tyson or an Abrams tank?

Who swims faster? Mike Phelps or a sailfish?

There is no contest and comparisons seem ridiculous.