I'v seen a few times someone asking for an engine that searches for draw.
How do you think such engine should work? I understand that the feed position will be bad and the engine should try to save it. But in fact this is what the engines try when they are in a bad position. So what is the difference over a normal engine?
Searching for draw
Moderators: hgm, Rebel, chrisw
-
cdani
- Posts: 2204
- Joined: Sat Jan 18, 2014 10:24 am
- Location: Andorra
Searching for draw
Daniel José - 
http://www.andscacs.com
http://www.andscacs.com-
Uri Blass
- Posts: 10267
- Joined: Thu Mar 09, 2006 12:37 am
- Location: Tel-Aviv Israel
Re: Searching for draw
The difference is that the engine can search for forcing a draw also in a good position.
Here is an example:
[d]4rrk1/pp6/5Q2/8/8/8/3PPP2/4K3 w - - 1 1
white is better but if white search for a forced draw the engine will show the line
1.Qg6+ Kh8 2.Qh6+ Kg8 3.Qg6+ and is going to report draw in 5 plies because it forced repetition inside the search in 5 plies and you can assume repetition inside the search is a draw.
In the following position it is a forced draw in 4 plies
[d]4rrk1/pp6/7Q/8/8/8/3PPP2/4K3 w - - 0 1
Line is 1.Qg6+ Kh8 2.Qh6+ Kg8 and the engine forced repetition of the root position so it is a draw because if you could force repetition of the root position in 4 plies then you can force it again.
Note that repetition of history positions of the game should be relevant only if the position happened twice because the fact that you can repeat some history position does not prove that you can repeat it again.
Here is an example:
[d]4rrk1/pp6/5Q2/8/8/8/3PPP2/4K3 w - - 1 1
white is better but if white search for a forced draw the engine will show the line
1.Qg6+ Kh8 2.Qh6+ Kg8 3.Qg6+ and is going to report draw in 5 plies because it forced repetition inside the search in 5 plies and you can assume repetition inside the search is a draw.
In the following position it is a forced draw in 4 plies
[d]4rrk1/pp6/7Q/8/8/8/3PPP2/4K3 w - - 0 1
Line is 1.Qg6+ Kh8 2.Qh6+ Kg8 and the engine forced repetition of the root position so it is a draw because if you could force repetition of the root position in 4 plies then you can force it again.
Note that repetition of history positions of the game should be relevant only if the position happened twice because the fact that you can repeat some history position does not prove that you can repeat it again.
-
Uri Blass
- Posts: 10267
- Joined: Thu Mar 09, 2006 12:37 am
- Location: Tel-Aviv Israel
Re: Searching for draw
I can add that the target is proving at least a draw and today if an engine say 0.00 I do not know if the engine found a forced draw or it is some equal position.
It is useful for analysis of games because humans sometimes need a draw in a game to win in the tournament so if they lose and analyze they want to know if they missed some way to force a draw that the engine can see and the fact that they had 0.00 score at some time of the game and did a mistake does not tell them in what point of the game they did a practical mistake and got out of a position that they could force a draw to a position that may be equal but the engine does not see a forced draw.
It is useful for analysis of games because humans sometimes need a draw in a game to win in the tournament so if they lose and analyze they want to know if they missed some way to force a draw that the engine can see and the fact that they had 0.00 score at some time of the game and did a mistake does not tell them in what point of the game they did a practical mistake and got out of a position that they could force a draw to a position that may be equal but the engine does not see a forced draw.
-
Uri Blass
- Posts: 10267
- Joined: Thu Mar 09, 2006 12:37 am
- Location: Tel-Aviv Israel
Re: Searching for draw
I can add that I believe that it can be useful also for computer chess to decide which move is better between 0.00 moves
Here is an example that I composed
[d]6k1/1p4p1/6Q1/8/8/1q6/1P4P1/1K6 w - - 0 1
both 1.Qe8+ and 1.g3 give evaluation of 0.00 but I consider 1.Qe8+ as better because white is the side that can force repetition by 1.Qe8+ and I see no proof that black can force a draw when the opposite is for 1.g3.
1.Qe8+ is evaluated as 0+(0.00 but maybe more than it that I do not see)
1.g3 is evaluated as 0-(0.00 but maybe less than it from white point of view that I do not see)
I know of no chess engine that has this calculations and I think that it is better if engines use part of their time to evaluate if one side can force a draw(the idea is to add 0.01 to the evaluation for moves that can force a draw and reduce 0.01 from the evaluation for moves when the opponent can force a draw).
Note that maybe a strong engine can see that black can force a draw also after 1.Qe8+ by searching deep enough but even if black can do it then black cannot do it as fast as white and black will need many plies to force a draw after 1.Qe8+.
Here is an example that I composed
[d]6k1/1p4p1/6Q1/8/8/1q6/1P4P1/1K6 w - - 0 1
both 1.Qe8+ and 1.g3 give evaluation of 0.00 but I consider 1.Qe8+ as better because white is the side that can force repetition by 1.Qe8+ and I see no proof that black can force a draw when the opposite is for 1.g3.
1.Qe8+ is evaluated as 0+(0.00 but maybe more than it that I do not see)
1.g3 is evaluated as 0-(0.00 but maybe less than it from white point of view that I do not see)
I know of no chess engine that has this calculations and I think that it is better if engines use part of their time to evaluate if one side can force a draw(the idea is to add 0.01 to the evaluation for moves that can force a draw and reduce 0.01 from the evaluation for moves when the opponent can force a draw).
Note that maybe a strong engine can see that black can force a draw also after 1.Qe8+ by searching deep enough but even if black can do it then black cannot do it as fast as white and black will need many plies to force a draw after 1.Qe8+.
-
cdani
- Posts: 2204
- Joined: Sat Jan 18, 2014 10:24 am
- Location: Andorra
Re: Searching for draw
I have done a version of Andscacs that evaluates repetition of position as 5000, so it takes it when possible:
https://1drv.ms/u/s!Aj8wlSmr_Hq3ispbyGn ... g?e=Rdzam9
https://1drv.ms/u/s!Aj8wlSmr_Hq3ispbyGn ... g?e=Rdzam9
Daniel José - http://www.andscacs.com
http://www.andscacs.com-
Ferdy
- Posts: 4833
- Joined: Sun Aug 10, 2008 3:15 pm
- Location: Philippines
Re: Searching for draw
Deuterium has options for repetition, insufficient, fiftymove and stalemate scores.
[d]4rrk1/pp6/5Q2/8/8/8/3PPP2/4K3 w - - 1 1
D:\Chess\Engines\Deuterium> .\Deuterium_v2019.2.37.73_64bit_pop.exe
uci
setoption name repetitionscore value 1000
position fen 4rrk1/pp6/5Q2/8/8/8/3PPP2/4K3 w - - 1 1
go movetime 5000
[d]4rrk1/pp6/5Q2/8/8/8/3PPP2/4K3 w - - 1 1
D:\Chess\Engines\Deuterium> .\Deuterium_v2019.2.37.73_64bit_pop.exe
Code: Select all
info string Deuterium v2019.2.37.73 64bit POPCNT
Code: Select all
id name Deuterium v2019.2.37.73
...
option name RepetitionScore type spin default 0 min -29800 max 29800
option name InsufficientMaterialScore type spin default 0 min -29800 max 29800
option name FiftyMoveScore type spin default 0 min -29800 max 29800
option name StaleMateScore type spin default 0 min -29800 max 29800
option name TimeBufferMilliSec type spin default 60 min 0 max 1000
option name MultiPV type spin default 1 min 1 max 300
option name UCI_AnalyseMode type check default false
option name ClearHash type button
uciok
position fen 4rrk1/pp6/5Q2/8/8/8/3PPP2/4K3 w - - 1 1
go movetime 5000
Code: Select all
info depth 1 seldepth 3 multipv 1 score cp 150 time 1 nodes 67 nps 67000 pv f6h6
info depth 2 seldepth 6 multipv 1 score cp 168 time 7 nodes 221 nps 31571 pv f6g6 g8h8 d2d4
info depth 3 seldepth 8 multipv 1 score cp 177 time 14 nodes 457 nps 32642 pv f6g6 g8h8 g6h6 h8g8 d2d4
info depth 4 seldepth 7 multipv 1 score cp 1000 time 24 nodes 475 nps 19791 pv f6g6
info depth 4 seldepth 6 multipv 1 score cp 1000 time 32 nodes 628 nps 19625 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 5 seldepth 6 multipv 1 score cp 1000 time 42 nodes 751 nps 17880 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 6 seldepth 6 multipv 1 score cp 1000 time 49 nodes 895 nps 18265 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 7 seldepth 7 multipv 1 score cp 1000 time 59 nodes 1116 nps 18915 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 8 seldepth 9 multipv 1 score cp 1000 time 69 nodes 1356 nps 19652 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 9 seldepth 9 multipv 1 score cp 1000 time 77 nodes 1784 nps 23168 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 10 seldepth 11 multipv 1 score cp 1000 time 89 nodes 2283 nps 25651 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 11 seldepth 11 multipv 1 score cp 1000 time 101 nodes 2995 nps 29653 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 12 seldepth 12 multipv 1 score cp 1000 time 110 nodes 3839 nps 34900 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 13 seldepth 12 multipv 1 score cp 1000 time 124 nodes 4975 nps 40120 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 14 seldepth 14 multipv 1 score cp 1000 time 134 nodes 6363 nps 47485 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 15 seldepth 14 multipv 1 score cp 1000 time 143 nodes 8109 nps 56706 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 16 seldepth 16 multipv 1 score cp 1000 time 155 nodes 10859 nps 70058 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 17 seldepth 22 multipv 1 score cp 1000 time 169 nodes 14147 nps 83710 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 18 seldepth 20 multipv 1 score cp 1000 time 183 nodes 18595 nps 101612 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 19 seldepth 20 multipv 1 score cp 1000 time 195 nodes 23778 nps 121938 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 20 seldepth 22 multipv 1 score cp 1000 time 211 nodes 29925 nps 141824 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 21 seldepth 23 multipv 1 score cp 1000 time 226 nodes 37362 nps 165318 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 22 seldepth 22 multipv 1 score cp 1000 time 241 nodes 47084 nps 195369 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 23 seldepth 27 multipv 1 score cp 1000 time 260 nodes 59560 nps 229076 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 24 seldepth 25 multipv 1 score cp 1000 time 279 nodes 73386 nps 263032 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 25 seldepth 28 multipv 1 score cp 1000 time 302 nodes 92223 nps 305374 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 26 seldepth 30 multipv 1 score cp 1000 time 358 nodes 140715 nps 393058 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 27 seldepth 32 multipv 1 score cp 1000 time 419 nodes 196836 nps 469775 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 28 seldepth 34 multipv 1 score cp 1000 time 492 nodes 265939 nps 540526 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 29 seldepth 33 multipv 1 score cp 1000 time 531 nodes 306186 nps 576621 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 30 seldepth 36 multipv 1 score cp 1000 time 791 nodes 566180 nps 715777 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 31 seldepth 36 multipv 1 score cp 1000 time 995 nodes 750400 nps 754170 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info nodes 750400 nps 745924
info depth 32 seldepth 38 multipv 1 score cp 1000 time 1177 nodes 930932 nps 790936 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 33 seldepth 38 multipv 1 score cp 1000 time 1321 nodes 1088730 nps 824171 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info depth 34 seldepth 39 multipv 1 score cp 1000 time 1530 nodes 1319678 nps 862534 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info nodes 1759232 nps 876985
info depth 35 seldepth 44 multipv 1 score cp 1000 time 2646 nodes 2436440 nps 920801 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info nodes 2796032 nps 930150
info nodes 3862272 nps 964121
info depth 36 seldepth 44 multipv 1 score cp 1000 time 4519 nodes 4419201 nps 977915 pv f6g6 g8h8 g6h6 h8g8 h6g6 g8h8
info time 5000 nodes 4884480
bestmove f6g6 ponder g8h8
-
Uri Blass
- Posts: 10267
- Joined: Thu Mar 09, 2006 12:37 am
- Location: Tel-Aviv Israel
Re: Searching for draw
It seems that it is not exactly take it when possible and it evaluates draw sometimes as -50.00 and sometimes as +50.00cdani wrote: ↑Thu Mar 11, 2021 10:42 pm I have done a version of Andscacs that evaluates repetition of position as 5000, so it takes it when possible:
https://1drv.ms/u/s!Aj8wlSmr_Hq3ispbyGn ... g?e=Rdzam9
I guess -50 if it is white to move and +50 if it is black to move.
Here it evaluate draw as -50 because the repetition is with white to move.
FEN: 6k1/qpp3p1/1p6/7Q/8/8/8/6K1 w - - 0 1
Andscacs:
1/3 00:00 31 31k -3.72 Qh5-e5
2/5 00:00 247 124k -4.00 Qh5-e5 c7-c6
2/6 00:00 391 196k -3.08 Qh5-d5+ Kg8-f8
2/3 00:00 400 200k 0.00 Qh5-e8+ Kg8-h7
3/7 00:00 502 251k -3.60 Qh5-e8+ Kg8-h7 Qe8-e5
4/6 00:00 2k 820k -3.74 Qh5-e8+ Kg8-h7 Qe8-e5 c7-c6
5/10+ 00:00 1k 396k -3.58 Qh5-e5
5/12+ 00:00 3k 988k -3.31 Kg1-h1
5/13 00:00 6k 1,483k -3.97 Qh5-e8+ Kg8-h7 Qe8-e5 b6-b5+ Kg1-g2 Qa7-b6
6/13+ 00:00 5k 1,253k -3.81 Qh5-d5+
6/12+ 00:00 9k 2,213k -3.65 Qh5-d5+
6/11+ 00:00 9k 2,312k -3.03 Qh5-d5+
6/12 00:00 6k 1,185k -2.88 Qh5-d5+ Kg8-f8 Qd5-d8+ Kf8-f7 Qd8xc7+ Kf7-g6 Qc7-e5 Qa7-a4
7/13 00:00 8k 1,553k -3.36 Qh5-e8+ Kg8-h7 Qe8-e4+ Kh7-h6 Qe4-f4+ Kh6-g6 Qf4xc7 b6-b5+ Kg1-g2 Qa7-e3
8/17 00:00 10k 1,740k -3.39 Qh5-e8+ Kg8-h7 Qe8-e4+ Kh7-h6 Qe4-h4+ Kh6-g6 Qh4-g3+ Kg6-h7 Qg3xc7 b6-b5+ Kg1-g2 Qa7-a6
9/19 00:00 16k 2,285k -3.30 Qh5-e8+ Kg8-h7 Qe8-e4+ g7-g6 Qe4-e7+ Kh7-h6 Qe7xc7 b6-b5+ Kg1-f1 Qa7-a4 Qc7xb7 Qa4-c4+ Kf1-g2
10/25+ 00:00 43k 3,610k -3.30 Qh5-e8+
10/23+ 00:00 71k 3,549k -3.20 Qh5-d5+
10/25 00:00 122k 4,076k -3.35 Qh5-d5+ Kg8-h7 Qd5-h1+ Kh7-g6 Qh1-e4+ Kg6-f7 Qe4-f4+ Kf7-g8 Qf4xc7 b6-b5+ Kg1-f1 Qa7-a1+ Kf1-f2 Qa1-a2+ Kf2-g3 Qa2-d5 Qc7-c8+ Kg8-f7 Qc8-g4
11/23 00:00 125k 4,017k -3.21 Qh5-d5+ Kg8-h7 Qd5-h1+ Kh7-g6 Qh1-e4+ Kg6-f7 Qe4-f4+ Kf7-g8 Qf4xc7 b6-b5+ Kg1-f1 Qa7-a1+ Kf1-e2 Qa1-a2+ Ke2-e3 Qa2-b3+ Ke3-d4 Kg8-h7
11/25+ 00:00 138k 3,738k -3.15 Qh5-d5+
11/22 00:00 141k 3,703k -2.98 Qh5-d5+ Kg8-h7 Qd5-h1+ Kh7-g6 Qh1-e4+ Kg6-f7 Qe4-f4+ Kf7-g8 Qf4xc7 b6-b5+ Kg1-f1 Kg8-h7 Qc7-e5 Qa7-a4 Kf1-e2 Qa4-b3 Qe5-e4+ Kh7-h6
12/23 00:00 145k 3,713k -3.07 Qh5-d5+ Kg8-h7 Qd5-h1+ Kh7-g6 Qh1-e4+ Kg6-f7 Qe4-f4+ Kf7-g8 Qf4xc7 b6-b5+ Kg1-f1 Qa7-a1+ Kf1-e2 Qa1-a2+ Ke2-e3 Qa2-b3+ Ke3-d4 Kg8-h7 Qc7-c5
13/36 00:00 354k 4,783k -3.32 Qh5-d5+ Kg8-h7 Qd5-h1+ Kh7-g6 Qh1-e4+ Kg6-f7 Qe4-f4+ Kf7-g8 Qf4xc7 b6-b5+ Kg1-f1 Qa7-a1+ Kf1-g2 Qa1-a2+ Kg2-g3 Qa2-d5 Qc7-e7 Kg8-h7 Kg3-h4 Kh7-h6 Qe7-e3+ Kh6-h7
14/34+ 00:00 501k 5,165k -3.16 Qh5-d5+
14/37+ 00:00 697k 5,578k -3.12 Qh5-d5+
14/45 00:00 1,109k 6,410k -3.20 Qh5-d5+ Kg8-h7 Qd5-e4+ Kh7-h6 Qe4-e6+ Kh6-g5 Qe6-e5+ Kg5-g6 Qe5-e8+ Kg6-f6 Qe8-d8+ Kf6-e6 Qd8-c8+ Ke6-d6 Qc8-f8+ Kd6-c6 Qf8-e8+ Kc6-c5 Qe8-e5+ Kc5-b4 Qe5-d4+ Kb4-b5 Qd4-d3+ Kb5-c5 Qd3-c3+ Kc5-d6 Qc3-g3+ Kd6-d5 Qg3xc7 b6-b5+ Kg1-g2
15/43 00:00 1,409k 6,743k -3.33 Qh5-d5+ Kg8-h7 Qd5-e4+ Kh7-h6 Qe4-e6+ Kh6-g5 Qe6-e5+ Kg5-g6 Qe5-e8+ Kg6-f6 Qe8-d8+ Kf6-f5 Qd8-d7+ Kf5-e4 Qd7-e7+ Ke4-d5 Qe7-f7+ Kd5-c6 Qf7-e8+ Kc6-c5 Qe8-e5+ Kc5-b4 Qe5-d4+ Kb4-b5 Qd4-d3+ Kb5-c6 Qd3-f3+ Kc6-c5
16/54 00:00 1,751k 7,118k -3.35 Qh5-d5+ Kg8-h7 Qd5-e4+ Kh7-h6 Qe4-e6+ Kh6-g5 Qe6-e5+ Kg5-g6 Qe5-e8+ Kg6-f6 Qe8-d8+ Kf6-e6 Qd8-c8+ Ke6-d6 Qc8-f8+ Kd6-d5 Qf8-f7+ Kd5-c6 Qf7-e8+ Kc6-c5 Qe8-e5+ Kc5-b4 Qe5-d4+ Kb4-b5 Qd4-b2+ Kb5-c4 Qb2-c1+ Kc4-b4 Qc1-f4+ Kb4-b5 Qf4xc7 Qa7-a1+ Kg1-g2 Qa1-d4
17/62+ 00:00 2,256k 7,647k -3.19 Qh5-d5+
17/63+ 00:00 3,042k 7,964k -3.15 Qh5-d5+
17/79+ 00:00 3,246k 8,075k -2.99 Qh5-d5+
17/74 00:00 4,748k 8,665k -3.20 Qh5-d5+ Kg8-h7 Qd5-e4+ Kh7-h6 Qe4-h4+ Kh6-g6 Qh4-g4+ Kg6-f6 Qg4-f4+ Kf6-e6 Qf4-c4+ Ke6-d7 Qc4-f7+ Kd7-d6 Qf7-f8+ Kd6-e6 Qf8-g8+ Ke6-f6 Qg8-d8+ Kf6-g6 Qd8-d3+ Kg6-g5 Qd3-d2+ Kg5-h4 Qd2-f4+ Kh4-h5 Qf4xc7 b6-b5+ Kg1-g2 Qa7-a2+ Kg2-g3 Qa2-d5
18/72+ 00:00 4,971k 8,645k -3.04 Qh5-d5+
18/72+ 00:00 6,045k 8,761k -3.00 Qh5-d5+
18/82 00:00 8,266k 9,174k -3.02 Qh5-d5+ Kg8-h7 Qd5-e4+ Kh7-h6 Qe4-h4+ Kh6-g6 Qh4-g4+ Kg6-f6 Qg4-d4+ Kf6-e6 Qd4-e4+ Ke6-d7 Qe4-g4+ Kd7-c6 Qg4-f3+ Kc6-c5 Qf3-f2+ Kc5-c4 Qf2-f4+ Kc4-b5 Qf4-f1+ Kb5-b4 Qf1-e1+ Kb4-c4 Qe1-c1+ Kc4-d5 Qc1-d1+ Kd5-c5 Qd1-c2+ Kc5-b4 Qc2-e4+ Kb4-c3 Qe4-e5+ Kc3-b3 Qe5-d5+ Kb3-a3
19/103+ 00:01 14,580k 9,280k -3.00 Qh5-d5+
19/90+ 00:01 16,840k 9,413k -2.77 Qh5-d5+
19/67 00:01 17,170k 9,450k -2.78 Qh5-d5+ Kg8-h7 Qd5-e4+ Kh7-h6 Qe4-e3+ g7-g5 Qe3-e6+ Kh6-g7 Qe6-d7+ Kg7-f6 Qd7-d8+ Kf6-e6 Qd8-g8+ Ke6-d7 Qg8-d5+ Kd7-e8 Qd5-e6+ Ke8-d8 Qe6-f6+ Kd8-d7 Qf6-f7+ Kd7-c6 Qf7-e6+ Kc6-c5 Qe6-e7+ Kc5-d5 Qe7xg5+ Kd5-c4 Qg5-f4+ Kc4-d3 Qf4-f5+ Kd3-d2 Qf5-e5 b6-b5+ Kg1-f1 Qa7-a4
20/75+ 00:01 18,617k 9,479k -2.78 Qh5-d5+
20/102 00:02 22,148k 9,766k -2.79 Qh5-d5+ Kg8-h7 Qd5-d3+ g7-g6 Qd3-h3+ Kh7-g7 Qh3-c3+ Kg7-g8 Qc3-c4+ Kg8-h8 Qc4-h4+ Kh8-g7 Qh4-e7+ Kg7-h6 Qe7-e3+ g6-g5 Qe3-h3+ Kh6-g7 Qh3-d7+ Kg7-f6 Qd7-d4+ Kf6-e6 Qd4-c4+ Ke6-d6 Qc4-d3+ Kd6-c5 Qd3-f5+ Kc5-b4 Qf5-e4+ Kb4-c3 Qe4-e5+ Kc3-d3 Qe5-d5+ Kd3-c2 Qd5-f5+ Kc2-c1 Qf5xg5+ Kc1-c2 Qg5-g2+ Kc2-c3 Qg2-g3+ Kc3-d2 Qg3-f4+ Kd2-d3 Qf4-f1+ Kd3-c3 Qf1-e1+ Kc3-d4 Qe1-f2+ Kd4-e4
21/114+ 00:02 24,712k 9,764k -2.63 Qh5-d5+
21/126+ 00:03 35,804k 10,259k -2.77 Qh5-d5+
21/72 00:06 63,145k 10,170k -3.48 Qh5-d5+ Kg8-h7 Qd5-h1+ Kh7-g6 Qh1-g2+ Kg6-f7 Qg2-d5+ Kf7-e7 Qd5-g5+ Ke7-d6 Qg5-g6+ Kd6-d5 Qg6-f7+ Kd5-c6 Qf7-f3+ Kc6-c5 Qf3-c3+ Kc5-d5 Qc3xc7 b6-b5+ Kg1-g2 Qa7-a2+ Kg2-g3 Qa2-a3+ Kg3-g4 Qa3-c5 Qc7xb7+ Kd5-c4 Qb7-f7+ Qc5-d5 Qf7-c7+ Kc4-b3 Qc7-g3+ Kb3-a4 Qg3-e3 b5-b4 Kg4-h4
22/76 00:06 67,707k 10,104k -3.47 Qh5-d5+ Kg8-h7 Qd5-e4+ Kh7-h6 Qe4-e3+ Kh6-h5 Qe3-e5+ Kh5-g6 Qe5-g3+ Kg6-f5 Qg3-f2+ Kf5-e6 Qf2-e2+ Ke6-d6 Qe2-h2+ Kd6-d7 Qh2-h3+ Kd7-e7 Qh3-e3+ Ke7-f7 Qe3-f3+ Kf7-e8 Qf3-h5+ Ke8-f8 Qh5-f5+ Kf8-e7 Qf5-g5+ Ke7-f7 Qg5-d5+ Kf7-f6 Qd5-d8+ Kf6-g6 Qd8-d3+ Kg6-g5 Qd3-d4 c7-c5 Qd4xg7+ Kg5-f5 Qg7-d7+ Kf5-e4
23/92 00:07 74,893k 10,121k -3.41 Qh5-d5+ Kg8-h7 Qd5-e4+ Kh7-h6 Qe4-e3+ Kh6-h5 Qe3-e5+ Kh5-g6 Qe5-e8+ Kg6-f5 Qe8-d7+ Kf5-e4 Qd7-e6+ Ke4-d4 Qe6-g4+ Kd4-d3 Qg4-g3+ Kd3-e2 Qg3-e5+ Ke2-f3 Qe5-f5+ Kf3-e3 Qf5-g5+ Ke3-e4 Qg5-g2+ Ke4-f4 Qg2-d2+ Kf4-e5 Qd2-e2+ Ke5-f6 Qe2-f3+ Kf6-g5 Qf3-d5+ Kg5-h6 Qd5-e6+ g7-g6 Qe6-h3+ Kh6-g7 Qh3-c3+ Kg7-g8 Qc3-c4+ Kg8-f8 Qc4-f4+ Kf8-e8
24/90+ 00:08 89,651k 10,132k -3.41 Qh5-d5+
24/80+ 00:09 92,480k 10,092k -3.16 Qh5-d5+
24/126+ 00:10 101,228k 10,085k -2.98 Qh5-d5+
24/58 00:19 192,958k 10,000k -3.29 Qh5-d5+ Kg8-h7 Qd5-e4+ Kh7-h6 Qe4-e3+ Kh6-g6 Qe3-e6+ Kg6-h5 Qe6-e5+ g7-g5 Qe5-e8+ Kh5-h6 Qe8-f8+ Kh6-g6 Qf8-g8+ Kg6-f6 Qg8-h8+ Kf6-f7 Qh8-h5+ Kf7-e6 Qh5-h3+ g5-g4 Qh3xg4+ Ke6-e5 Qg4-g3+ Ke5-e4 Qg3-g6+ Ke4-d5 Qg6-f5+ Kd5-d4 Qf5-f4+ Kd4-d3 Qf4-f3+ Kd3-c4 Qf3-f7+ Kc4-b4 Qf7-f4+ Kb4-b5 Qf4xc7 Qa7-a1+ Kg1-h2 Qa1-b2+ Kh2-g3 Qb2-b3+ Kg3-f2 Qb3-d5
25/100+ 00:20 205,653k 9,893k -3.45 Qh5-d5+
25/100+ 00:24 235,064k 9,724k -3.29 Qh5-d5+
25/95 00:31 299,443k 9,549k -3.36 Qh5-d5+ Kg8-h7 Qd5-e4+ Kh7-h6 Qe4-e6+ Kh6-g5 Qe6-e5+ Kg5-g6 Qe5-g3+ Kg6-f6 Qg3-c3+ Kf6-f5 Qc3-c2+ Kf5-g4 Qc2-e4+ Kg4-g3 Qe4-e5+ Kg3-f3 Qe5-f5+ Kf3-e3 Qf5-f2+ Ke3-d3 Qf2-f3+ Kd3-c4 Qf3-e4+ Kc4-c3 Qe4-e3+ Kc3-b2 Qe3-e2+ Kb2-a3 Qe2-f3+ Ka3-a2 Qf3-g2+ Ka2-b3 Qg2-g3+ Kb3-a4 Qg3-f4+ Ka4-b5 Qf4-f1+ Kb5-c5 Qf1-f2+ Kc5-c4 Qf2-c2+ Kc4-d4 Qc2-d2+ Kd4-e4 Qd2-e2+ Ke4-f4 Qe2-h2+
26/75+ 00:33 321,311k 9,461k -3.36 Qh5-d5+
26/92+ 00:35 336,303k 9,407k -3.26 Qh5-d5+
26/97+ 01:26 750,987k 8,654k -3.55 Qh5-d5+
26/102 02:00 1,011,638k 8,387k -5.06 Qh5-d5+ Kg8-h7 Kg1-g2 Qa7-a5 Qd5xb7 Qa5-e5 Kg2-h3 Kh7-h6 Kh3-g4 b6-b5 Qb7-h1+ Kh6-g6 Qh1-c6+ Kg6-f7 Qc6-b7 g7-g5 Kg4-f3 Kf7-g6 Qb7-b8 Kg6-h5 Qb8-g8 Qe5-f4+ Kf3-g2 Qf4-e4+ Kg2-g1 c7-c6 Qg8-f7+ Kh5-g4 Qf7-b3 b5-b4 Kg1-h2 c6-c5
27/71+ 02:01 1,020,935k 8,368k -4.90 Qh5-d5+
27/85+ 02:09 1,081,693k 8,342k -4.86 Qh5-d5+
27/101+ 02:32 1,254,842k 8,216k -3.22 Qh5-d5+
27/126 03:05 1,533,518k 8,273k -3.38 Qh5-d5+ Kg8-h7 Qd5-e4+ Kh7-h6 Qe4-e6+ g7-g6 Qe6-h3+ Kh6-g5 Qh3-e3+ Kg5-f6 Qe3-f4+ Kf6-e6 Qf4-e4+ Ke6-f7 Qe4-d5+ Kf7-f6 Qd5-d4+ Kf6-f5 Qd4-d3+ Kf5-e5 Qd3-g3+ Ke5-e4 Qg3-g2+ Ke4-d4 Qg2-f2+ Kd4-d5 Qf2-f3+ Kd5-c5 Qf3-e3+ Kc5-d6 Qe3-d4+ Kd6-e7 Qd4-e5+ Ke7-d7 Qe5-g7+ Kd7-e6 Qg7xg6+ Ke6-e5 Qg6-g3+ Ke5-d5 Qg3-d3+ Kd5-c5 Qd3-f5+ Kc5-c4 Qf5-e4+ Kc4-c3 Qe4-e5+ Kc3-d3 Qe5xc7 b6-b5+ Kg1-f1 Qa7-a1+ Kf1-g2 Qa1-d4 Kg2-g3 Qd4-e4
28/101+ 03:08 1,554,813k 8,266k -3.38 Qh5-d5+
28/92+ 03:17 1,629,638k 8,245k -3.54 Qh5-d5+
28/96+ 04:00 1,968,372k 8,183k -3.95 Qh5-d5+
Here is see +50 because it is black to move in the repetition position
FEN: 6k1/qpp3p1/1p4Q1/8/8/8/8/6K1 w - - 0 1
Andscacs:
1/2 00:00 27 27k -73.82 Qg6xg7+
1/3 00:00 35 35k -4.09 Kg1-g2
1/3 00:00 45 45k -3.86 Qg6-e4
2/4 00:00 210 210k -4.14 Qg6-e4 Qa7-a6
2/8 00:00 512 512k -4.08 Qg6-f5 Qa7-a8
2/6 00:00 453 453k -3.96 Qg6-e6+ Kg8-h7
3/7 00:00 2k 834k -3.60 Qg6-e6+ Kg8-h7 Qe6-e5
4/8 00:00 3k 1,320k -3.74 Qg6-e6+ Kg8-h7 Qe6-e5 c7-c6
5/9 00:00 3k 1,298k -3.66 Qg6-e6+ Kg8-h8 Kg1-g2 Qa7-a8 Qe6-f7
6/12+ 00:00 3k 1,098k -3.50 Qg6-e8+
6/6+ 00:00 6k 1,905k -3.34 Qg6-e8+
6/6+ 00:00 6k 1,181k +39.33 Qg6-e8+
6/6 00:00 6k 1,184k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
7/7 00:00 6k 1,003k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
8/8 00:00 6k 1,026k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
9/9 00:00 6k 1,051k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
10/10 00:00 7k 953k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
11/11 00:00 7k 1,033k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
12/12 00:00 8k 1,099k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
13/13 00:00 8k 1,179k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
14/14 00:00 9k 1,116k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
15/15 00:00 10k 1,233k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
16/16 00:00 11k 1,213k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
17/17 00:00 13k 1,269k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
18/18 00:00 15k 1,470k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
19/19 00:00 16k 1,419k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
20/20 00:00 18k 1,610k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
21/21 00:00 22k 1,829k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
22/22 00:00 25k 2,106k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
23/23 00:00 30k 2,312k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
24/24 00:00 35k 2,709k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
25/25 00:00 44k 3,146k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
26/26 00:00 52k 3,469k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
27/27 00:00 61k 3,793k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
28/28 00:00 78k 4,097k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
29/29 00:00 98k 4,676k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
30/30 00:00 123k 4,543k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
31/31 00:00 143k 4,783k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
32/32 00:00 170k 5,142k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
33/33 00:00 237k 5,150k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
34/34 00:00 283k 5,442k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
35/35 00:00 382k 5,454k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
36/36 00:00 536k 5,468k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
37/37 00:00 550k 5,341k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
38/38 00:00 704k 5,675k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
38/22 00:00 688k 5,290k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
39/39 00:00 735k 5,607k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
40/40 00:00 851k 5,389k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
41/41 00:00 943k 5,581k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
42/42 00:00 1,241k 5,799k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
43/43 00:00 1,422k 5,874k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
44/44 00:00 2,680k 6,106k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
45/45 00:00 3,072k 6,322k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
46/46 00:00 3,667k 6,656k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
47/47 00:00 4,120k 6,890k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
48/48 00:00 5,716k 7,611k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
49/49 00:00 7,859k 7,971k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
50/50 00:01 11,448k 8,449k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
51/51 00:02 18,479k 8,791k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
52/52 00:02 21,065k 8,881k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
52/50 00:03 30,029k 9,217k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
53/53 00:03 30,464k 9,129k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
54/54 00:03 30,720k 9,146k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
55/55 00:03 35,522k 9,246k +50.00 Qg6-e8+ Kg8-h7 Qe8-h5+ Kh7-g8 Qh5-e8+
-
cdani
- Posts: 2204
- Joined: Sat Jan 18, 2014 10:24 am
- Location: Andorra
Re: Searching for draw
Solved!
Just download it again.
If you don't ask me by private message I will not send you a version with nnue enabled much stronger
Daniel José - http://www.andscacs.com
http://www.andscacs.com-
cdani
- Posts: 2204
- Joined: Sat Jan 18, 2014 10:24 am
- Location: Andorra
Re: Searching for draw
Someone has used id? It works as expected? I tried and seems so.
If at least two or three ppl are interested, I can do the same modification to Stockfish.
Daniel José - http://www.andscacs.com
http://www.andscacs.com