bob wrote:That is a risk yes. But I think it can be greatly reduced by not creating the bitbases as bitbases or by converting them from dtm egtbs, but by converting them from DTZ tablebases.Tony wrote:My concern is that "won" doesn't necessarily translate to 1-0 in the real game. There are lots of positions where capturing or advancing the pawn prematurely, even though the resulting position is won by bitbase info, leads to a draw because of the 50 move rule. KRP vs KR is a good example where you can run the pawn down to the 7th too quickly and see the win evaporate. There are similar cases such as in KNN vs KP where it is easy to let the pawn advance too far which pushes the win beyond 50 moves. This gets magnified with 6 piece endings which is really where the bitbases will become more useful because of the 1tb+ size of th egtbs. Those endings get harder and harder to win with the 50 move rule looming large.Daniel Shawul wrote:I'm not sure. There are variations where you should not take the first chance for transposition. OTOH transposing into a winning position with less pieces should make it simpler.What could you have done in my situation?? Do you expect me to delay
the capture in the hope that it will help in avoiding the 50 move rule later?
You can only find such kind of exact information (where exactly the capture should happen) in EGTB with the mate in N condition. Assuming we have only bitbases, then all I can do at best is to weigh the different factors in to condsderation. If you have other ideas , lets hear it. Other wise stop making such silly claims which are easily refuted.
You should make some tests to settle this. Include positions where the "plain win" is more than 50 moves away without resetting the 50 move counter, so the "searching bitbases" could actually score better.
I haven't found any positions that XiniX messes up, and it didn't even have the full bitbases ( I used a lossy compression). That includes that cm in 107 ( iirc) in a KNNKP ending. But thats doesn't mean there aren't.
Tony
I played around with this a good while back, and decided that a combination of the "swindle-mode" approach, along with an evaluation that had some specific code for particularly problematic endings might work. And for some endings, like KRP vs KR it worked pretty well and was able to win won positions against a version of Crafty with EGTBs which means "optimal" play against the bitbase version. But when I expanded to 6 piece files, the "rules" became way too complex and I stopped working on the problem.
That way you only have to search atmost 99 ply to guarantee the best move. Anything below 99 poses a risk, but I seriously doubt it's a big one.
But then again, I only tried KPPKPP from the 6 pieces, wich has unsurprisingly few problems to make progress.
Tony
New Scorpio bitbase files
Moderator: Ras
Re: New Scorpio bitbase files
-
- Posts: 4186
- Joined: Tue Mar 14, 2006 11:34 am
- Location: Ethiopia
Re: New Scorpio bitbase files
Yes I could have done that but I guess the speed improvement will not be that much. I don't cutoff with a win/loss score inside the 2/3 search tree. So if I do probes there to cull the loosing moves, obviously it will speed up the search a bit. But One should not forget the effect of alpha-beta especially when I am returning scores closer to what is given to a real mate. The tree searched by including all moves will not be that bigger than the one with the culled moves. The speed gain is also diminished by the fact that I have to probe the bitbases inside tree now, which I am not doing right now. Note that culling the moves will not help me in progressing because I probe the bitbases at the boundary line (2/3 depth) and that will tell me if If the move that should have been culled (inside search tree) is actually a non-winning move. The regular eval() ,which could have tampered that, is never called in my case. I think the only advantage is speed which I might test in the future.The pawn term is implicit because of the transpose bonus.
What I didn't see you mention ( and is worthwhile) is the folowing. When in a won bitbase position, prune all the moves that don't lead to a lost position for the opponent.
This also sounds pretty much like the one we have been discussing at the root, except that now it is inside search. For me culling the loosing root moves that does not help in the progression aspect at all.
Daniel
-
- Posts: 4186
- Joined: Tue Mar 14, 2006 11:34 am
- Location: Ethiopia
Re: New Scorpio bitbase files
I randomly setup a KNNKP position to show how scorpio works.
Pleae note how the PV is exactly as long as how I specified it to be. Sometimes it might be shorter when it finds a capture/promotion/pawn push where I cutoff. I will try that in my next post.
[d] 6k1/7p/8/8/8/8/1NN2K2/8 w - - 0 1
With full search depth dedicated to finding winning ways.
With 2/3rd only
Just for fun I tried also 1/3rd
Now if I just set it to a depth inside the queiscence search it starts to use its regular eval and might return a wrong unless I cull the loosing root moves at the root.
Daniel
Pleae note how the PV is exactly as long as how I specified it to be. Sometimes it might be shorter when it finds a capture/promotion/pawn push where I cutoff. I will try that in my next post.
[d] 6k1/7p/8/8/8/8/1NN2K2/8 w - - 0 1
With full search depth dedicated to finding winning ways.
Code: Select all
FEN: 6k1/7p/8/8/8/8/1NN2K2/8 w - - 0 1
Scorpio:
2 00:00 8 200 +50.64 Nb2-d3 Kg8-f7
2 00:00 22 550 +50.80 Kf2-g3 Kg8-f7
2 00:00 22 550 +50.80 Kf2-g3 Kg8-f7
3 00:00 99 707 +50.92 Kf2-g3 Kg8-g7 Kg3-g4
3 00:00 122 813 +50.92 Kf2-g3 Kg8-g7 Kg3-g4
4 00:00 214 1.426 +50.10 Kf2-g3 Kg8-f7 Kg3-g4 Kf7-e6
4 00:00 257 1.713 +50.10 Kf2-g3 Kg8-f7 Kg3-g4 Kf7-e6
5 00:00 800 3.478 +50.09 Kf2-g3 Kg8-g7 Nb2-d3 Kg7-g6 Kg3-f4
5 00:00 1.488 5.314 +50.14 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-g6 Kf3-e4
5 00:00 1.488 5.314 +50.14 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-g6 Kf3-e4
6 00:00 2.457 8.472 +49.40 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-e6 Ke3-e4 h7-h5
6 00:00 2.928 10.096 +49.40 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-e6 Ke3-e4 h7-h5
7 00:00 9.820 20.458 +49.34 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-f6 Kf3-e4 Kf6-g6 Nc2-e3
7 00:00 13.002 24.532 +49.34 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-f6 Kf3-e4 Kf6-g6 Nc2-e3
8 00:00 24.898 44.460 +48.59 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-f6 Ke3-f4 Kf6-e6 Kf4-g5 Ke6-d5
8 00:00 30.635 51.923 +48.60 Kf2-g3 Kg8-g7 Nb2-d1 Kg7-f6 Kg3-f4 Kf6-e6 Kf4-e4 h7-h5
8 00:00 30.635 51.923 +48.60 Kf2-g3 Kg8-g7 Nb2-d1 Kg7-f6 Kg3-f4 Kf6-e6 Kf4-e4 h7-h5
9 00:01 63.912 60.294 +48.54 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-g6 Kg3-f4 Kg6-f6 Kf4-e4 Kf6-g6 Nc2-e3
9 00:01 93.374 47.639 +48.56 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-g6 Ke3-e4 Kg6-g5 Nd3-e5 h7-h5
9 00:01 93.374 47.639 +48.56 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-g6 Ke3-e4 Kg6-g5 Nd3-e5 h7-h5
10 00:02 223.733 104.548 +47.83 Nb2-d3 Kg8-g7 Kf2-e3 Kg7-f6 Ke3-e4 Kf6-g5 Nd3-f4 Kg5-f6 Nc2-e3 h7-h5
10 00:02 238.260 110.818 +47.83 Nb2-d3 Kg8-g7 Kf2-e3 Kg7-f6 Ke3-e4 Kf6-g5 Nd3-f4 Kg5-f6 Nc2-e3 h7-h5
11 00:03 417.525 104.381 +47.81 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-f6 Kf3-g4 Kf6-g6 Nc2-d4 Kg6-f6 Kg4-h5 Kf6-g7 Kh5-g5
11 00:04 507.780 101.963 +47.81 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-f6 Kf3-g4 Kf6-g6 Nc2-d4 Kg6-f6 Kg4-h5 Kf6-g7 Kh5-g5
12 00:05 845.663 154.318 +47.11 Nb2-d3 Kg8-g7 Kf2-f3 Kg7-f7 Kf3-g4 Kf7-g6 Nc2-d4 Kg6-f6 Kg4-h5 Kf6-e7 Kh5-h6 Ke7-d6
12 00:05 939.670 171.472 +47.11 Nb2-d3 Kg8-g7 Kf2-f3 Kg7-f7 Kf3-g4 Kf7-g6 Nc2-d4 Kg6-f6 Kg4-h5 Kf6-e7 Kh5-h6 Ke7-d6
13 00:12 1.654.868 129.997 +47.06 Nb2-d3 Kg8-g7 Kf2-f3 Kg7-f7 Nc2-e3 Kf7-f6 Kf3-f4 Kf6-e6 Kf4-g5 Ke6-f7 Kg5-f5 Kf7-g7 Nd3-f4
13 00:15 1.988.436 130.560 +47.06 Nb2-d3 Kg8-g7 Kf2-f3 Kg7-f7 Nc2-e3 Kf7-f6 Kf3-f4 Kf6-e6 Kf4-g5 Ke6-f7 Kg5-f5 Kf7-g7 Nd3-f4
14 00:17 3.272.777 189.946 +46.36 Nb2-d3 Kg8-g7 Kf2-f3 Kg7-f7 Nc2-e3 Kf7-f6 Kf3-f4 Kf6-e6 Kf4-g5 Ke6-d7 Nd3-f4 Kd7-d6 Kg5-f5 h7-h6
14 00:17 3.669.519 204.088 +46.36 Nb2-d3 Kg8-g7 Kf2-f3 Kg7-f7 Nc2-e3 Kf7-f6 Kf3-f4 Kf6-e6 Kf4-g5 Ke6-d7 Nd3-f4 Kd7-d6 Kg5-f5 h7-h6
15 00:38 6.061.142 157.514 +46.34 Nb2-d3 Kg8-g7 Kf2-f3 Kg7-f7 Nc2-e3 Kf7-f6 Kf3-f4 Kf6-e6 Kf4-g5 Ke6-e7 Nd3-f4 Ke7-d6 Kg5-f5 Kd6-e7 Ne3-f1
15 00:45 6.963.121 153.949 +46.34 Nb2-d3 Kg8-g7 Kf2-f3 Kg7-f7 Nc2-e3 Kf7-f6 Kf3-f4 Kf6-e6 Kf4-g5 Ke6-e7 Nd3-f4 Ke7-d6 Kg5-f5 Kd6-e7 Ne3-f1
16 00:57 15.068.552 259.892 +45.59 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-e6 Kf3-g4 Ke6-f6 Nc2-d4 Kf6-e7 Kg4-f5 Ke7-d7 Nd4-b5 Kd7-c6 Nb5-c3 Kc6-d7 Nc3-e4 Kd7-e7
16 00:59 16.374.545 274.142 +45.59 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-e6 Kf3-g4 Ke6-f6 Nc2-d4 Kf6-e7 Kg4-f5 Ke7-d7 Nd4-b5 Kd7-c6 Nb5-c3 Kc6-d7 Nc3-e4 Kd7-e7
Code: Select all
FEN: 6k1/7p/8/8/8/8/1NN2K2/8 w - - 0 1
Scorpio:
2 00:00 2 2 +51.57 Nb2-d3
2 00:00 4 4 +51.66 Kf2-g3
2 00:00 4 4 +51.66 Kf2-g3
3 00:00 12 12 +50.80 Kf2-g3 Kg8-f7
3 00:00 14 14 +50.80 Kf2-g3 Kg8-f7
4 00:00 22 22 +50.80 Kf2-g3 Kg8-f7
4 00:00 24 24 +50.80 Kf2-g3 Kg8-f7
5 00:00 139 13.900 +50.92 Kf2-g3 Kg8-g7 Kg3-g4
5 00:00 183 18.300 +50.92 Kf2-g3 Kg8-g7 Kg3-g4
6 00:00 323 32.300 +50.10 Kf2-g3 Kg8-f7 Kg3-g4 Kf7-e6
6 00:00 385 38.500 +50.10 Kf2-g3 Kg8-f7 Kg3-g4 Kf7-e6
7 00:00 542 54.200 +50.10 Kf2-g3 Kg8-f7 Kg3-g4 Kf7-e6
7 00:00 584 58.400 +50.10 Kf2-g3 Kg8-f7 Kg3-g4 Kf7-e6
8 00:00 1.496 37.400 +50.09 Kf2-g3 Kg8-f7 Nb2-d3 Kf7-g6 Kg3-f4
8 00:00 2.363 26.255 +50.14 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-g6 Kf3-e4
8 00:00 2.363 26.255 +50.14 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-g6 Kf3-e4
9 00:00 3.618 40.200 +49.40 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-e6 Kf3-e4 h7-h5
9 00:00 4.273 47.477 +49.40 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-e6 Kf3-e4 h7-h5
10 00:00 5.517 61.300 +49.40 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-e6 Kf3-e4 h7-h5
10 00:00 6.003 66.700 +49.40 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-e6 Kf3-e4 h7-h5
11 00:00 18.762 53.605 +49.34 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-f6 Kf3-e4 Kf6-g6 Nc2-e3
11 00:00 25.545 55.532 +49.34 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-f6 Kf3-e4 Kf6-g6 Nc2-e3
12 00:00 44.817 93.368 +48.59 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-f6 Kf3-f4 Kf6-e6 Kf4-g5 Ke6-d5
12 00:00 51.058 102.116 +48.60 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-f6 Kg3-f4 Kf6-e6 Kf4-e4 h7-h6
12 00:00 51.058 102.116 +48.60 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-f6 Kg3-f4 Kf6-e6 Kf4-e4 h7-h6
13 00:00 59.460 112.188 +48.60 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-f6 Kg3-f4 Kf6-e6 Kf4-e4 h7-h6
13 00:00 67.450 124.907 +48.60 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-f6 Kg3-f4 Kf6-e6 Kf4-e4 h7-h6
14 00:00 91.650 117.500 +48.54 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-g6 Kg3-f4 Kg6-f6 Nc2-d4 Kf6-g6 Kf4-e4
14 00:01 141.065 110.207 +48.56 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-g6 Kf3-e4 Kg6-g5 Nd3-f4 h7-h5
14 00:01 141.065 110.207 +48.56 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-g6 Kf3-e4 Kg6-g5 Nd3-f4 h7-h5
15 00:01 258.347 181.934 +47.83 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-g7 Kf3-e4 Kg7-g6 Nd3-f4 Kg6-f6 Nc2-e3 h7-h6
15 00:01 294.182 201.494 +47.83 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-g7 Kf3-e4 Kg7-g6 Nd3-f4 Kg6-f6 Nc2-e3 h7-h6
16 00:01 370.943 237.783 +47.83 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-g7 Kf3-e4 Kg7-g6 Nd3-f4 Kg6-f6 Nc2-e3 h7-h6
16 00:01 386.395 246.111 +47.83 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-g7 Kf3-e4 Kg7-g6 Nd3-f4 Kg6-f6 Nc2-e3 h7-h6
17 00:04 720.906 169.624 +47.81 Nb2-d3 Kg8-f7 Kf2-f3 Kf7-g7 Kf3-g4 Kg7-g6 Nc2-d4 Kg6-f6 Kg4-h5 Kf6-g7 Kh5-g5
Code: Select all
FEN: 6k1/7p/8/8/8/8/1NN2K2/8 w - - 0 1
Scorpio:
2 00:00 5 5 +0.10 Nb2-d3
3 00:00 9 9 +51.57 Nb2-d3
3 00:00 11 11 +51.66 Kf2-g3
3 00:00 11 11 +51.66 Kf2-g3
4 00:00 13 13 +51.66 Kf2-g3
4 00:00 14 14 +51.66 Kf2-g3
5 00:00 16 16 +51.66 Kf2-g3
5 00:00 17 17 +51.66 Kf2-g3
6 00:00 25 25 +50.80 Kf2-g3 Kg8-f7
6 00:00 27 27 +50.80 Kf2-g3 Kg8-f7
7 00:00 35 35 +50.80 Kf2-g3 Kg8-f7
7 00:00 37 37 +50.80 Kf2-g3 Kg8-f7
8 00:00 45 45 +50.80 Kf2-g3 Kg8-f7
8 00:00 47 47 +50.80 Kf2-g3 Kg8-f7
9 00:00 162 4.050 +50.92 Kf2-g3 Kg8-g7 Kg3-g4
9 00:00 206 5.150 +50.92 Kf2-g3 Kg8-g7 Kg3-g4
10 00:00 271 6.775 +50.92 Kf2-g3 Kg8-g7 Kg3-g4
10 00:00 294 7.350 +50.92 Kf2-g3 Kg8-g7 Kg3-g4
11 00:00 323 8.075 +50.92 Kf2-g3 Kg8-g7 Kg3-g4
11 00:00 346 8.650 +50.92 Kf2-g3 Kg8-g7 Kg3-g4
12 00:00 570 14.250 +50.10 Kf2-g3 Kg8-f7 Kg3-g4 Kf7-e6
12 00:00 635 15.875 +50.10 Kf2-g3 Kg8-f7 Kg3-g4 Kf7-e6
13 00:00 775 12.916 +50.10 Kf2-g3 Kg8-f7 Kg3-g4 Kf7-e6
13 00:00 817 13.616 +50.10 Kf2-g3 Kg8-f7 Kg3-g4 Kf7-e6
14 00:00 942 15.700 +50.10 Kf2-g3 Kg8-f7 Kg3-g4 Kf7-e6
14 00:00 984 16.400 +50.10 Kf2-g3 Kg8-f7 Kg3-g4 Kf7-e6
15 00:00 1.986 18.054 +50.09 Kf2-g3 Kg8-f7 Nb2-d3 Kf7-g6 Kg3-f4
15 00:00 3.091 15.455 +50.14 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-g6 Ke3-e4
15 00:00 3.091 15.455 +50.14 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-g6 Ke3-e4
16 00:00 3.611 17.195 +50.14 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-g6 Ke3-e4
16 00:00 3.970 18.904 +50.14 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-g6 Ke3-e4
17 00:00 4.457 13.928 +50.14 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-g6 Ke3-e4
17 00:00 4.698 14.681 +50.14 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-g6 Ke3-e4
18 00:00 6.078 17.876 +49.40 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-e6 Ke3-e4 h7-h5
18 00:00 6.679 18.552 +49.40 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-e6 Ke3-e4 h7-h5
19 00:00 7.663 21.286 +49.40 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-e6 Ke3-e4 h7-h5
19 00:00 8.117 22.547 +49.40 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-e6 Ke3-e4 h7-h5
20 00:00 9.006 24.340 +49.40 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-e6 Ke3-e4 h7-h5
20 00:00 9.413 25.440 +49.40 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-e6 Ke3-e4 h7-h5
21 00:00 23.410 39.677 +49.34 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-f6 Ke3-e4 Kf6-g6 Nc2-e3
21 00:00 28.135 41.992 +49.34 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-f6 Ke3-e4 Kf6-g6 Nc2-e3
22 00:00 34.770 44.576 +49.34 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-f6 Ke3-e4 Kf6-g6 Nc2-e3
22 00:00 38.015 45.255 +49.34 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-f6 Ke3-e4 Kf6-g6 Nc2-e3
23 00:00 44.292 47.625 +49.34 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-f6 Ke3-e4 Kf6-g6 Nc2-e3
23 00:00 46.735 48.682 +49.34 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-f6 Ke3-e4 Kf6-g6 Nc2-e3
24 00:01 64.301 64.301 +48.59 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-f6 Ke3-f4 Kf6-e6 Kf4-g5 Ke6-d5
24 00:01 73.000 72.277 +48.60 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-f6 Kg3-f4 Kf6-e6 Kf4-e4 h7-h6
24 00:01 73.000 72.277 +48.60 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-f6 Kg3-f4 Kf6-e6 Kf4-e4 h7-h6
25 00:01 79.167 76.861 +48.60 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-f6 Kg3-f4 Kf6-e6 Kf4-e4 h7-h6
25 00:01 88.420 85.019 +48.60 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-f6 Kg3-f4 Kf6-e6 Kf4-e4 h7-h6
26 00:01 93.450 88.160 +48.60 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-f6 Kg3-f4 Kf6-e6 Kf4-e4 h7-h6
26 00:01 102.066 95.388 +48.60 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-f6 Kg3-f4 Kf6-e6 Kf4-e4 h7-h6
27 00:01 131.570 96.036 +48.58 Kf2-g3 Kg8-f7 Nb2-d3 Kf7-f6 Kg3-f4 Kf6-g6 Kf4-e5 Kg6-g5 Nc2-e3
27 00:01 179.617 96.568 +48.58 Kf2-g3 Kg8-f7 Nb2-d3 Kf7-f6 Kg3-f4 Kf6-g6 Kf4-e5 Kg6-g5 Nc2-e3
28 00:02 211.580 97.055 +48.54 Kf2-g3 Kg8-f7 Nb2-d1 Kf7-e6 Kg3-f4 Ke6-f6 Kf4-e4 Kf6-g6 Nc2-e3
28 00:02 248.260 98.126 +48.56 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-g6 Ke3-e4 Kg6-g5 Nd3-f4 h7-h5
28 00:02 248.260 98.126 +48.56 Nb2-d3 Kg8-f7 Kf2-e3 Kf7-g6 Ke3-e4 Kg6-g5 Nd3-f4 h7-h5
Code: Select all
FEN: 6k1/7p/8/8/8/8/1NN2K2/8 w - - 0 1
Scorpio:
2 00:00 5 5 +49.95 Nb2-d3
3 00:00 96 96 0.00 Nb2-d3 Kg8-g7 Nc2-e3
3 00:00 116 116 0.00 Nb2-d3 Kg8-g7 Nc2-e3
4 00:00 245 245 +0.10 Nb2-d3 Kg8-f7 Nc2-e3 Kf7-g8
4 00:00 281 281 +0.10 Nb2-d3 Kg8-f7 Nc2-e3 Kf7-g8
5 00:00 884 884 0.00 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-h8 Ne3-g4
5 00:00 1.201 1.201 0.00 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-h8 Ne3-g4
6 00:00 2.464 2.464 +0.10 Nb2-d3 Kg8-f7 Nc2-e3 Kf7-g8 Ne3-f5 Kg8-f7
6 00:00 2.953 2.953 +0.10 Nb2-d3 Kg8-f7 Nc2-e3 Kf7-g8 Ne3-f5 Kg8-f7
7 00:00 10.062 335.400 0.00 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-g8 Ne3-f5 Kg8-h8 Nf5-h6
7 00:00 12.672 422.400 0.00 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-g8 Ne3-f5 Kg8-h8 Nf5-h6
8 00:00 24.254 346.485 +0.10 Nb2-d3 Kg8-f7 Nc2-e3 Kf7-g8 Ne3-d1 Kg8-g7 Nd1-b2 Kg7-g8
8 00:00 27.983 399.757 +0.10 Nb2-d3 Kg8-f7 Nc2-e3 Kf7-g8 Ne3-d1 Kg8-g7 Nd1-b2 Kg7-g8
9 00:00 76.533 637.775 +0.10 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-g8 Ne3-f5 Kg8-h8 Nd3-e5 Kh8-g8 Nf5-h6 Kg8-g7
9 00:00 91.708 611.386 +0.10 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-g8 Ne3-f5 Kg8-h8 Nd3-e5 Kh8-g8 Nf5-h6 Kg8-g7
10 00:00 168.362 801.723 +0.10 Nb2-d3 Kg8-f7 Nc2-e3 Kf7-g8 Ne3-d1 Kg8-g7 Nd1-b2 Kg7-g8 Kf2-f3 Kg8-f7
10 00:00 190.087 826.465 +0.10 Nb2-d3 Kg8-f7 Nc2-e3 Kf7-g8 Ne3-d1 Kg8-g7 Nd1-b2 Kg7-g8 Kf2-f3 Kg8-f7
11 00:00 435.207 946.102 +0.10 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-g8 Ne3-f5 Kg8-h8 Nd3-e5 Kh8-g8 Nf5-h6 Kg8-g7 Ne5-f7 Kg7-f6
11 00:00 493.716 968.070 +0.10 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-g8 Ne3-f5 Kg8-h8 Nd3-e5 Kh8-g8 Nf5-h6 Kg8-g7 Ne5-f7 Kg7-f6
12 00:00 975.161 1.083.512 +0.10 Nb2-d3 Kg8-f7 Nc2-e3 Kf7-g8 Ne3-d1 Kg8-g7 Nd1-b2 Kg7-g8 Kf2-f3 Kg8-f7 Kf3-e4 Kf7-g8
12 00:01 1.078.556 1.078.556 +0.10 Nb2-d3 Kg8-f7 Nc2-e3 Kf7-g8 Ne3-d1 Kg8-g7 Nd1-b2 Kg7-g8 Kf2-f3 Kg8-f7 Kf3-e4 Kf7-g8
13 00:01 1.987.909 1.162.519 +0.10 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-g8 Ne3-f5 Kg8-h8 Nd3-e5 Kh8-g8 Nf5-h6 Kg8-g7 Ne5-f7 Kg7-f6 Kf2-f3 Kf6-g7
13 00:01 2.194.145 1.173.339 +0.10 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-g8 Ne3-f5 Kg8-h8 Nd3-e5 Kh8-g8 Nf5-h6 Kg8-g7 Ne5-f7 Kg7-f6 Kf2-f3 Kf6-g7
14 00:04 4.492.497 998.332 +0.10 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-h8 Kf2-f3 Kh8-g7 Ne3-f5 Kg7-g8 Nf5-h6 Kg8-g7 Nh6-g4 Kg7-g8 Ng4-f6 Kg8-g7 Nf6-h5 Kg7-g6 Nh5-g3 Kg6-f7
14 00:04 4.786.995 1.007.788 +0.10 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-h8 Kf2-f3 Kh8-g7 Ne3-f5 Kg7-g8 Nf5-h6 Kg8-g7 Nh6-g4 Kg7-g8 Ng4-f6 Kg8-g7 Nf6-h5 Kg7-g6 Nh5-g3 Kg6-f7
15 00:08 8.671.260 991.001 +0.10 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-g8 Ne3-f5 Kg8-h8 Nd3-e5 Kh8-g8 Nf5-h6 Kg8-g7 Ne5-f7 Kg7-f6 Kf2-f3 Kf6-g7 Kf3-g4 Kg7-f6
15 00:09 9.320.450 1.007.616 +0.10 Nb2-d3 Kg8-g7 Nc2-e3 Kg7-g8 Ne3-f5 Kg8-h8 Nd3-e5 Kh8-g8 Nf5-h6 Kg8-g7 Ne5-f7 Kg7-f6 Kf2-f3 Kf6-g7 Kf3-g4 Kg7-f6
-
- Posts: 20943
- Joined: Mon Feb 27, 2006 7:30 pm
- Location: Birmingham, AL
Re: New Scorpio bitbase files
Tony wrote:I made mine from direct enumeration for extraction from an egtb pair. I chose KRPKR because that is a somewhat problematic position since it is not only about advancing the pawn, but also about controlling the opponent/your king position. That's where I discovered I needed some eval "help" to win won positions. I finally decided that the best case was bitbases in the search, egtbs at the root and this problem goes away. Then I decided to really make the problem go away and not worry with the bitbases.bob wrote:That is a risk yes. But I think it can be greatly reduced by not creating the bitbases as bitbases or by converting them from dtm egtbs, but by converting them from DTZ tablebases.Tony wrote:My concern is that "won" doesn't necessarily translate to 1-0 in the real game. There are lots of positions where capturing or advancing the pawn prematurely, even though the resulting position is won by bitbase info, leads to a draw because of the 50 move rule. KRP vs KR is a good example where you can run the pawn down to the 7th too quickly and see the win evaporate. There are similar cases such as in KNN vs KP where it is easy to let the pawn advance too far which pushes the win beyond 50 moves. This gets magnified with 6 piece endings which is really where the bitbases will become more useful because of the 1tb+ size of th egtbs. Those endings get harder and harder to win with the 50 move rule looming large.Daniel Shawul wrote:I'm not sure. There are variations where you should not take the first chance for transposition. OTOH transposing into a winning position with less pieces should make it simpler.What could you have done in my situation?? Do you expect me to delay
the capture in the hope that it will help in avoiding the 50 move rule later?
You can only find such kind of exact information (where exactly the capture should happen) in EGTB with the mate in N condition. Assuming we have only bitbases, then all I can do at best is to weigh the different factors in to condsderation. If you have other ideas , lets hear it. Other wise stop making such silly claims which are easily refuted.
You should make some tests to settle this. Include positions where the "plain win" is more than 50 moves away without resetting the 50 move counter, so the "searching bitbases" could actually score better.
I haven't found any positions that XiniX messes up, and it didn't even have the full bitbases ( I used a lossy compression). That includes that cm in 107 ( iirc) in a KNNKP ending. But thats doesn't mean there aren't.
Tony
I played around with this a good while back, and decided that a combination of the "swindle-mode" approach, along with an evaluation that had some specific code for particularly problematic endings might work. And for some endings, like KRP vs KR it worked pretty well and was able to win won positions against a version of Crafty with EGTBs which means "optimal" play against the bitbase version. But when I expanded to 6 piece files, the "rules" became way too complex and I stopped working on the problem.
That way you only have to search atmost 99 ply to guarantee the best move. Anything below 99 poses a risk, but I seriously doubt it's a big one.
But then again, I only tried KPPKPP from the 6 pieces, wich has unsurprisingly few problems to make progress.
TonyAnd more recently I have not even been using egtbs...
-
- Posts: 4186
- Joined: Tue Mar 14, 2006 11:34 am
- Location: Ethiopia
Re: New Scorpio bitbase files
position from MES
[d] k7/5K2/8/8/7p/p6P/P7/8 w - - 0 1
After the Kxa3 capture is found it stops searching because I assume that
fewer number of pieces a zeroed fifty move count is a huge advantage. Not assuming that ofcourse might find you even more winning ways where the capture is delayed. But my search is now faster with not so much risk IMO. You can just invest more search nodes to find those winning ways being slower I guess.
Without the capture/promotion/pawn push cutoff. The pvs are longer now
but still the same pv until the kb4xa3 is maintained even after a very tall search depth. So we can say that you are trying to be more safe than mine, while I try to be fast in this case. With a capture (zeroed fifty move count + reduced number of pieces), the cutting of is pretty safe most of the time. Don't you think? But with a pawn push (zeroing of fifty move count) maybe the safety issue is a bit more important as you suggested.
I will try with the a pawn push example later...
[d] k7/5K2/8/8/7p/p6P/P7/8 w - - 0 1
After the Kxa3 capture is found it stops searching because I assume that
fewer number of pieces a zeroed fifty move count is a huge advantage. Not assuming that ofcourse might find you even more winning ways where the capture is delayed. But my search is now faster with not so much risk IMO. You can just invest more search nodes to find those winning ways being slower I guess.
Code: Select all
FEN: k7/5K2/8/8/7p/p6P/P7/8 w - - 0 1
Scorpio:
2 00:00 13 13 +1.18 Kf7-e6 Ka8-b7
2 00:00 27 27 +1.18 Kf7-e6 Ka8-b7
3 00:00 51 51 +1.28 Kf7-e6 Ka8-b7 Ke6-e5
3 00:00 112 112 +1.28 Kf7-e6 Ka8-b7 Ke6-e5
4 00:00 137 137 +0.78 Kf7-e6
4 00:00 252 252 +1.44 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4
4 00:00 303 303 +1.44 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4
5 00:00 419 419 +1.44 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4
5 00:00 528 528 +1.44 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4
6 00:00 680 680 +0.82 Kf7-f6
6 00:00 960 960 +0.82 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4 Kc6-d5
6 00:00 996 996 +0.82 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4 Kc6-d5
7 00:00 1.103 1.103 +1.39 Kf7-f6
7 00:00 1.358 1.358 +1.41 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4 Kc6-d5 Kh4-g4
7 00:00 1.576 1.576 +1.41 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4 Kc6-d5 Kh4-g4
8 00:00 1.868 1.868 +0.82 Kf7-f6
8 00:00 3.223 3.223 +0.39 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5-h6 Kc6-d5 Kh6-h5 Kd5-e4 Kh5xh4
8 00:00 4.301 4.301 +0.48 Kf7-g7 Ka8-b7 Kg7-f6 Kb7-c6 Kf6-e6 Kc6-c7 Ke6-e7
8 00:00 4.335 4.335 +0.48 Kf7-g7 Ka8-b7 Kg7-f6 Kb7-c6 Kf6-e6 Kc6-c7 Ke6-e7
9 00:00 5.265 5.265 +0.39 Kf7-g7 Ka8-b7 Kg7-f6 Kb7-c7 Kf6-g5 Kc7-c6 Kg5-h6
9 00:00 5.821 5.821 +0.98 Kf7-e6
9 00:00 6.598 6.598 +0.98 Kf7-e6 Ka8-b7 Ke6-f5 Kb7-c6 Kf5-g5 Kc6-d5 Kg5xh4 Kd5-e4 Kh4-g4
9 00:00 6.699 6.699 +0.98 Kf7-e6 Ka8-b7 Ke6-f5 Kb7-c6 Kf5-g5 Kc6-d5 Kg5xh4 Kd5-e4 Kh4-g4
10 00:00 7.704 7.704 +0.98 Kf7-e6 Ka8-b8 Ke6-f5 Kb8-c7 Kf5-g4 Kc7-d6 Kg4xh4 Kd6-e5 Kh4-g4 Ke5-e4
10 00:00 8.522 8.522 +0.98 Kf7-e6 Ka8-b8 Ke6-f5 Kb8-c7 Kf5-g4 Kc7-d6 Kg4xh4 Kd6-e5 Kh4-g4 Ke5-e4
11 00:00 9.907 990.700 +0.94 Kf7-e6 Ka8-b7 Ke6-f5 Kb7-c7 Kf5-g4 Kc7-d6 Kg4-h5 Kd6-e5 Kh5xh4 Ke5-f4 Kh4-h5
11 00:00 10.612 1.061.200 +0.94 Kf7-e6 Ka8-b7 Ke6-f5 Kb7-c7 Kf5-g4 Kc7-d6 Kg4-h5 Kd6-e5 Kh5xh4 Ke5-f4 Kh4-h5
12 00:00 12.882 1.288.200 +0.46 Kf7-e6 Ka8-b8 Ke6-f5 Kb8-c7 Kf5-g4 Kc7-d6 Kg4-h5 Kd6-e5 Kh5xh4 Ke5-f4 Kh4-h5 Kf4-e5
12 00:00 14.607 1.460.700 +0.46 Kf7-e6 Ka8-b8 Ke6-f5 Kb8-c7 Kf5-g4 Kc7-d6 Kg4-h5 Kd6-e5 Kh5xh4 Ke5-f4 Kh4-h5 Kf4-e5
13 00:00 25.710 2.571.000 +0.94 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c4 Kc8-b7 Kc4-b3 Kb7-c6 Kb3xa3 Kc6-d5 Ka3-b4 Kd5-e4 Kb4-c5
13 00:00 26.531 2.653.100 +0.94 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c4 Kc8-b7 Kc4-b3 Kb7-c6 Kb3xa3 Kc6-d5 Ka3-b4 Kd5-e4 Kb4-c5
14 00:00 28.478 949.266 +46.52 Kf7-e6
15 00:00 31.832 1.061.066 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
15 00:00 33.333 1.111.100 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
16 00:00 35.147 1.171.566 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
16 00:00 38.084 634.733 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
17 00:00 41.163 686.050 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
17 00:00 42.860 714.333 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
18 00:00 44.811 746.850 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
18 00:00 48.348 690.685 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
19 00:00 53.559 765.128 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
19 00:00 56.782 811.171 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
20 00:00 58.557 836.528 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
20 00:00 61.831 883.300 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
21 00:00 67.808 398.870 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
21 00:00 71.091 418.182 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
22 00:00 73.762 283.700 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
22 00:00 78.583 302.242 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
23 00:00 85.877 306.703 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
23 00:00 89.526 279.768 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
24 00:00 91.937 287.303 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
24 00:00 97.284 286.129 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
25 00:00 102.689 302.026 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
25 00:00 107.227 306.362 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
26 00:00 109.079 311.654 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
26 00:00 116.942 316.059 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
27 00:00 122.513 314.135 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
27 00:00 125.532 313.830 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
28 00:00 128.349 305.592 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
28 00:00 139.476 324.362 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
29 00:00 145.985 339.500 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
29 00:00 151.131 335.846 +46.52 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c8 Kd5-c5 Kc8-d7 Kc5-b4 Kd7-e6 Kb4xa3
30 00:00 154.303 342.895 +46.52 Kf7-e6 Ka8-b8 Ke6-d5 Kb8-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3
but still the same pv until the kb4xa3 is maintained even after a very tall search depth. So we can say that you are trying to be more safe than mine, while I try to be fast in this case. With a capture (zeroed fifty move count + reduced number of pieces), the cutting of is pretty safe most of the time. Don't you think? But with a pawn push (zeroing of fifty move count) maybe the safety issue is a bit more important as you suggested.
Code: Select all
FEN: k7/5K2/8/8/7p/p6P/P7/8 w - - 0 1
Scorpio:
2 00:00 13 13 +1.18 Kf7-e6 Ka8-b7
2 00:00 27 27 +1.18 Kf7-e6 Ka8-b7
3 00:00 51 51 +1.28 Kf7-e6 Ka8-b7 Ke6-e5
3 00:00 112 112 +1.28 Kf7-e6 Ka8-b7 Ke6-e5
4 00:00 137 137 +0.78 Kf7-e6
4 00:00 252 252 +1.44 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4
4 00:00 303 303 +1.44 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4
5 00:00 419 419 +1.44 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4
5 00:00 528 528 +1.44 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4
6 00:00 680 680 +0.82 Kf7-f6
6 00:00 960 960 +0.82 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4 Kc6-d5
6 00:00 996 996 +0.82 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4 Kc6-d5
7 00:00 1.103 1.103 +1.39 Kf7-f6
7 00:00 1.358 1.358 +1.41 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4 Kc6-d5 Kh4-g4
7 00:00 1.576 1.576 +1.41 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5xh4 Kc6-d5 Kh4-g4
8 00:00 1.868 186.800 +0.82 Kf7-f6
8 00:00 3.223 322.300 +0.39 Kf7-f6 Ka8-b7 Kf6-g5 Kb7-c6 Kg5-h6 Kc6-d5 Kh6-h5 Kd5-e4 Kh5xh4
8 00:00 4.301 430.100 +0.48 Kf7-g7 Ka8-b7 Kg7-f6 Kb7-c6 Kf6-e6 Kc6-c7 Ke6-e7
8 00:00 4.335 433.500 +0.48 Kf7-g7 Ka8-b7 Kg7-f6 Kb7-c6 Kf6-e6 Kc6-c7 Ke6-e7
9 00:00 5.265 526.500 +0.39 Kf7-g7 Ka8-b7 Kg7-f6 Kb7-c7 Kf6-g5 Kc7-c6 Kg5-h6
9 00:00 5.827 582.700 +0.98 Kf7-e6
9 00:00 6.318 631.800 +0.92 Kf7-e6 Ka8-b7 Ke6-f5 Kb7-b6 Kf5-g4 Kb6-c6 Kg4-g5 Kc6-d6 Kg5xh4
9 00:00 6.481 648.100 +0.92 Kf7-e6 Ka8-b7 Ke6-f5 Kb7-b6 Kf5-g4 Kb6-c6 Kg4-g5 Kc6-d6 Kg5xh4
10 00:00 7.467 746.700 +0.82 Kf7-e6 Ka8-b8 Ke6-f5 Kb8-c7 Kf5-g4 Kc7-b6 Kg4-h5
10 00:00 8.512 851.200 +0.82 Kf7-e6 Ka8-b8 Ke6-f5 Kb8-c7 Kf5-g4 Kc7-b6 Kg4-h5
11 00:00 9.682 968.200 +0.82 Kf7-e6 Ka8-b7 Ke6-f5 Kb7-c7 Kf5-g4 Kc7-b6 Kg4-g5 Kb6-c6 Kg5-h5 Kc6-c5 Kh5xh4
11 00:00 10.511 50.052 +0.82 Kf7-e6 Ka8-b7 Ke6-f5 Kb7-c7 Kf5-g4 Kc7-b6 Kg4-g5 Kb6-c6 Kg5-h5 Kc6-c5 Kh5xh4
12 00:00 18.158 78.947 +0.44 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c7 Kd5-c4 Kc7-d8 Kc4-b3 Kd8-e7 Kb3xa3 Ke7-e6 Ka3-b4 Ke6-e5
12 00:00 19.100 83.043 +0.44 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c7 Kd5-c4 Kc7-d8 Kc4-b3 Kd8-e7 Kb3xa3 Ke7-e6 Ka3-b4 Ke6-e5
13 00:00 24.914 99.656 +0.94 Kf7-e6
13 00:00 26.062 104.248 +0.94 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-c6 Kb4xa3 Kc6-d6 Ka3-b4 Kd6-e5 Kb4-c5
13 00:00 26.603 106.412 +0.94 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-c6 Kb4xa3 Kc6-d6 Ka3-b4 Kd6-e5 Kb4-c5
14 00:00 27.443 105.550 +45.72 Kf7-e6
15 00:00 32.218 111.096 +44.27 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-d6 Kb4xa3 Kd6-e5
15 00:00 33.493 108.041 +44.27 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-d6 Kb4xa3 Kd6-e5
16 00:00 36.712 118.425 +44.27 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c7 Kc4-b3 Kc7-d6 Kb3xa3 Kd6-e5
16 00:00 37.317 120.377 +44.27 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c7 Kc4-b3 Kc7-d6 Kb3xa3 Kd6-e5
17 00:00 42.424 101.009 +45.77 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3 Ke6-f5 Ka3-b3
17 00:00 44.133 91.943 +45.77 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-c7 Kd5-c5 Kc7-d7 Kc5-b4 Kd7-e6 Kb4xa3 Ke6-f5 Ka3-b3
18 00:00 49.420 98.840 +43.47 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-d5
18 00:00 50.325 100.650 +43.47 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-d5
19 00:00 58.315 114.343 +43.47 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-d5
19 00:00 60.088 117.819 +43.47 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-d5
20 00:00 70.398 105.071 +45.02 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4
20 00:00 71.912 102.731 +45.02 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4
21 00:00 79.864 106.485 +42.88 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c3 Ka4-b5 Kc3-d4
21 00:00 81.771 107.593 +42.88 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c3 Ka4-b5 Kc3-d4
22 00:00 89.693 114.991 +42.88 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c3 Ka4-b5 Kc3-d4
22 00:00 91.439 117.229 +42.88 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c3 Ka4-b5 Kc3-d4
23 00:00 101.831 117.047 +44.26 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5
23 00:00 106.485 112.089 +44.26 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5
24 00:01 118.946 117.768 +42.27 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c3 Ka4-b5 Kc3-d4 a2-a4 Kd4-e5
24 00:01 120.987 117.463 +42.27 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c3 Ka4-b5 Kc3-d4 a2-a4 Kd4-e5
25 00:01 132.090 124.613 +42.28 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4
25 00:01 136.332 127.413 +42.28 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4
26 00:01 152.677 124.127 +43.56 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4 Kb4-c5
26 00:01 155.852 124.681 +43.56 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4 Kb4-c5
27 00:01 175.429 133.915 +41.68 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4 Kb4-c5 Kf4-e5
27 00:01 184.249 139.582 +41.68 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4 Kb4-c5 Kf4-e5
28 00:01 202.943 148.133 +41.68 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4 Kb4-c5 Kf4-e5
28 00:01 205.139 149.736 +41.68 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4 Kb4-c5 Kf4-e5
29 00:01 225.569 150.379 +42.82 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4 Kb4-c5 Kf4-e5 Kc5-c4
29 00:01 241.910 138.234 +42.82 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4 Kb4-c5 Kf4-e5 Kc5-c4
30 00:01 268.527 147.542 +40.88 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4 Kb4-c5 Kf4-e5 Kc5-c4 Ke5-e4
30 00:01 274.303 149.077 +40.88 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4 Kb4-c5 Kf4-e5 Kc5-c4 Ke5-e4
31 00:01 294.088 155.602 +40.88 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4 Kb4-c5 Kf4-e5 Kc5-c4 Ke5-e4
31 00:01 306.874 159.830 +40.88 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b4 Ke4-f4 a2-a4 Kf4-g3 a4-a5 Kg3-f4 Kb4-c5 Kf4-e5 Kc5-c4 Ke5-e4
32 00:03 388.151 112.182 +42.09 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b2 Ke4-f3 a2-a4 Kf3-e4 Kb2-c2 Ke4-e3 Kc2-d1 Ke3-f3 Kd1-e1 Kf3-f4 Ke1-f2
32 00:03 402.453 114.986 +42.09 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-e4 Ka3-b2 Ke4-f3 a2-a4 Kf3-e4 Kb2-c2 Ke4-e3 Kc2-d1 Ke3-f3 Kd1-e1 Kf3-f4 Ke1-f2
33 00:03 469.574 126.569 +40.10 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c3 Ka4-b5 Kc3-b2 a2-a4 Kb2-c3 a4-a5 Kc3-d4 Kb5-c6 Kd4-c4 a5-a6 Kc4-d4
33 00:03 501.302 131.575 +40.10 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c3 Ka4-b5 Kc3-b2 a2-a4 Kb2-c3 a4-a5 Kc3-d4 Kb5-c6 Kd4-c4 a5-a6 Kc4-d4
34 00:03 552.996 140.711 +40.10 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c3 Ka4-b5 Kc3-b2 a2-a4 Kb2-c3 a4-a5 Kc3-d4 Kb5-c6 Kd4-c4 a5-a6 Kc4-d4
34 00:03 563.519 142.302 +40.10 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c3 Ka4-b5 Kc3-b2 a2-a4 Kb2-c3 a4-a5 Kc3-d4 Kb5-c6 Kd4-c4 a5-a6 Kc4-d4
35 00:05 652.135 124.216 +40.78 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c5 Ka4-b3 Kc5-d5 Kb3-c3 Kd5-e5 a2-a4 Ke5-f5 Kc3-b3 Kf5-g6 Kb3-b2 Kg6-f5 a4-a5
35 00:05 721.274 131.140 +40.78 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c5 Ka4-b3 Kc5-d5 Kb3-c3 Kd5-e5 a2-a4 Ke5-f5 Kc3-b3 Kf5-g6 Kb3-b2 Kg6-f5 a4-a5
36 00:05 843.898 146.764 +39.51 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c3 Ka4-b5 Kc3-d4 Kb5-c6 Kd4-e5 a2-a4 Ke5-d4 Kc6-d6 Kd4-e4 Kd6-e6 Ke4-f4 Ke6-f6 Kf4-e4
36 00:05 890.515 154.872 +39.51 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c3 Ka4-b5 Kc3-d4 Kb5-c6 Kd4-e5 a2-a4 Ke5-d4 Kc6-d6 Kd4-e4 Kd6-e6 Ke4-f4 Ke6-f6 Kf4-e4
37 00:05 960.904 160.150 +39.63 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c5 Ka4-b3 Kc5-d4 Kb3-c2 Kd4-e4 Kc2-d2 Ke4-f3 Kd2-d3 Kf3-f4 Kd3-e2 Kf4-e4 Ke2-f2 Ke4-f4
37 00:06 1.007.333 161.173 +39.63 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c5 Ka4-b3 Kc5-d4 Kb3-c2 Kd4-e4 Kc2-d2 Ke4-f3 Kd2-d3 Kf3-f4 Kd3-e2 Kf4-e4 Ke2-f2 Ke4-f4
38 00:07 1.195.488 154.256 +40.40 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c5 Ka4-b3 Kc5-d4 Kb3-c2 Kd4-e4 a2-a4 Ke4-e3 Kc2-d1 Ke3-f3 Kd1-d2 Kf3-e4 Kd2-e2 Ke4-f4 Ke2-e1
38 00:07 1.370.337 171.292 +40.40 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c5 Ka4-b3 Kc5-d4 Kb3-c2 Kd4-e4 a2-a4 Ke4-e3 Kc2-d1 Ke3-f3 Kd1-d2 Kf3-e4 Kd2-e2 Ke4-f4 Ke2-e1
39 00:08 1.498.931 181.688 +39.32 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c5 Ka4-b3 Kc5-d4 Kb3-c2 Kd4-e4 Kc2-d2 Ke4-f3 a2-a4 Kf3-f2 Kd2-d3 Kf2-e1 Kd3-e4 Ke1-d2 Ke4-f4 Kd2-c2
39 00:08 1.669.499 196.411 +39.32 Kf7-e6 Ka8-b7 Ke6-d5 Kb7-b6 Kd5-c4 Kb6-c6 Kc4-b3 Kc6-d5 Kb3xa3 Kd5-c4 Ka3-a4 Kc4-c5 Ka4-b3 Kc5-d4 Kb3-c2 Kd4-e4 Kc2-d2 Ke4-f3 a2-a4 Kf3-f2 Kd2-d3 Kf2-e1 Kd3-e4 Ke1-d2 Ke4-f4 Kd2-c2
-
- Posts: 4186
- Joined: Tue Mar 14, 2006 11:34 am
- Location: Ethiopia
Re: New Scorpio bitbase files
With pawn pushes also the same story. You try to score them in eval() and not cutoff
inside search which is more on the safe side. I tend to perfer the faster search by taking
more risks. The problem is basically theoretically impossible to solve with bitbases alone.
We don't have perfect information so it just can not be solved with a shallow search (we may have the cm107 kind of
positions). Once in a while these positions will pop in my case and yours also, however impractical they
may seem. It is just a speed/accuracy issue IMO, while none of us being perfect. I think the
issue is clear now to my satisfaction.
Here is a very quick game (which is not helping the progress) between scorpio and ruffian. This is with the current setting that I have. Look at how it progresses and also finally shows a mate score +99.97 once it is in its reach.
[d] 6k1/8/7r/8/8/8/1PR5/1K6 w - - 0 1
With the immediate return after pawn push commented out. Since I don't evaluate pawn ranks it failed to mate within the given time!! (I really didn't expect that) . I think now that may be mixing the pawn advancement term, like you do, may not necessarily be better
. Because the other evaluation terms that you have might swamp the passed pawn scoring sometimes. Even if you give high scores to passed pawns, it will then start preferring advancement of pawns rather than exchanging pieces and staff like that... I think there are many factors coming into play!
inside search which is more on the safe side. I tend to perfer the faster search by taking
more risks. The problem is basically theoretically impossible to solve with bitbases alone.
We don't have perfect information so it just can not be solved with a shallow search (we may have the cm107 kind of
positions). Once in a while these positions will pop in my case and yours also, however impractical they
may seem. It is just a speed/accuracy issue IMO, while none of us being perfect. I think the
issue is clear now to my satisfaction.
Here is a very quick game (which is not helping the progress) between scorpio and ruffian. This is with the current setting that I have. Look at how it progresses and also finally shows a mate score +99.97 once it is in its reach.
[d] 6k1/8/7r/8/8/8/1PR5/1K6 w - - 0 1
Code: Select all
[Event "Computer Chess Game"]
[Site "WINKOOL-E3F9911"]
[Date "2009.02.16"]
[Round "-"]
[White "Scorpio 2.06"]
[Black "Ruffian 1.0.5"]
[Result "1-0"]
[TimeControl "40/60"]
[FEN "6k1/8/7r/8/8/8/1PR5/1K6 w - - 0 1"]
[SetUp "1"]
{--------------
. . . . . . k .
. . . . . . . .
. . . . . . . r
. . . . . . . .
. . . . . . . .
. . . . . . . .
. P R . . . . .
. K . . . . . .
white to play
--------------}
1. b3 {+50.22/11 2.8} Kf7 {-1.72/16 10} 2. Re2 {+45.63/1 0.2} Rh8 3. Kb2
{+46.08/10 1.7} Rb8 4. Kc3 {+46.93/9 1.3} Rc8+ 5. Kd4 {+46.89/9 1.5} Rb8 6.
Re3 {+46.74/8 1.7} Rd8+ {-1.69/14 1.3} 7. Kc5 {+45.94/10 2.1} Rc8+ 8. Kd6
{+45.89/10 1.5} Rb8 {-1.66/15 1.1} 9. Rf3+ {+45.89/10 1.2} Ke8 10. Kc7
{+45.41/11 1.6} Rb4 11. Kc6 {+45.62/11 1.6} Kd8 12. Rh3 {+46.59/8 1.7} Rd4
{-1.58/11 1.5} 13. Rh8+ {+47.51/8 1.9} Ke7 14. Rh7+ {+47.41/9 1.4} Kd8 15.
Rg7 {+46.52/9 2.4} Rf4 16. Rg8+ {+46.57/9 1.6} Ke7 17. Rb8 {+47.37/10 1.7}
Rf6+ {-1.42/12 1.0} 18. Kc7 {+48.17/10 1.5} Rf4 19. Rb6 {+48.97/10 1.5} Rd4
{-1.50/12 1.4} 20. b4 {+49.77/10 1.3} Rc4+ {-1.86/12 1.8} 21. Kb7
{+48.85/13 1.4} Kd7 22. b5 {+49.65/13 1.2} Rb4 {-2.05/14 2.9} 23. Ka7
{+46.28/11 1.0} Kc7 {-2.14/13 3} 24. Rb7+ {+47.08/13 1.4} Kd6
{-2.21/12 1.3} 25. Rb8 {+47.88/13 1.5} Re4 26. b6 {+49.55/9 1.8} Ra4+
{-2.49/10 1.8} 27. Kb7 {+46.78/1 0.1} Kd7 28. Rh8 {+47.10/9 1.5} Rb4
{-3.59/12 2.2} 29. Ka6 {+47.90/11 1.5} Rb2 30. b7 {+49.70/9 1.6} Ra2+
{-4.34/16 1.7} 31. Kb6 {+44.96/12 2.3} Rb2+ 32. Ka7 {+45.69/11 1.3} Ra2+
33. Kb8 {+45.79/10 1.4} Ra1 34. Rc8 {+46.55/9 1.5} Ra4 35. Rc3
{+46.66/9 1.9} Ra1 {-4.75/14 2.3} 36. Rc4 {+45.88/10 1.3} Ra2
{-5.15/13 2.3} 37. Rc7+ {+45.11/12 1.2} Kd8 38. Rh7 {+45.08/12 2.3} Ra1 39.
Rh2 {+45.08/12 1.0} Kd7 40. Rc2 {+44.96/12 0.4} Ra3 {-5.06/14 2.2} 41. Rc7+
{+44.96/13 1.8} Kd8 42. Rc4 {+44.12/14 1.7} Kd7 {-5.30/15 1.8} 43. Rd4+
{+45.23/12 2.1} Ke6 44. Kc7 {+47.17/14 1.6} Rc3+ 45. Kd8 {+47.97/15 1.8}
Rb3 46. Rd7 {+48.77/14 4} Rb5 {-5.34/15 1.6} 47. Rc7 {+49.59/11 1.6} Kd5
48. Kc8 {+50.39/11 2.0} Ke4 49. b8=Q {+53.27/9 1.3} Rxb8+ 50. Kxb8
{+50.77/1 0.7} Kd4 {-5.77/14 1.5} 51. Re7 {+50.19/15 1.4} Kc4 52. Rd7
{+49.89/16 1.4} Kb3 53. Rc7 {+49.69/17 1.2} Kb4 54. Kc8 {+48.89/19 1.4} Kb3
55. Kb7 {+99.82/18 1.4} Kb2 56. Ka6 {+99.84/18 1.3} Ka3 57. Kb5
{+99.86/19 1.0} Kb2 58. Kb4 {+99.92/22 1.1} Ka2 59. Kc3 {+99.94/52 1.0} Kb1
60. Ra7 {+99.96/70 0.2} Kc1 61. Ra1# {+99.98/70}
{Xboard adjudication: Checkmate} 1-0

Code: Select all
[Event "Computer Chess Game"]
[Site "WINKOOL-E3F9911"]
[Date "2009.02.16"]
[Round "-"]
[White "Scorpio 2.06"]
[Black "Ruffian 1.0.5"]
[Result "1/2-1/2"]
[TimeControl "40/60"]
[FEN "6k1/8/7r/8/8/8/1PR5/1K6 w - - 0 1"]
[SetUp "1"]
{--------------
. . . . . . k .
. . . . . . . .
. . . . . . . r
. . . . . . . .
. . . . . . . .
. . . . . . . .
. P R . . . . .
. K . . . . . .
white to play
--------------}
1. Rf2 {+46.98/9 2.9} Kg7 {-2.05/15 9} 2. Kc2 {+46.93/8 1.4} Rc6+ 3. Kd3
{+46.96/8 1.2} Rb6 4. Kc4 {+47.85/7 1.6} Rb8 5. b3 {+47.85/7 1.5} Rc8+ 6.
Kd5 {+47.85/7 1.2} Rb8 {-1.90/14 1.1} 7. Rf3 {+47.03/8 1.7} Kg6 8. Kc4
{+46.93/8 1.5} Rc8+ 9. Kb5 {+46.89/8 1.5} Rb8+ 10. Ka4 {+46.81/9 1.5} Ra8+
11. Kb4 {+46.66/9 1.5} Kg5 12. Rf1 {+46.72/8 1.5} Kg4 13. Kc4
{+46.89/8 1.3} Rc8+ {-1.80/13 1.1} 14. Kb5 {+46.83/9 1.7} Rb8+ 15. Ka4
{+45.87/10 1.6} Ra8+ 16. Kb4 {+46.69/9 1.5} Rb8+ 17. Ka3 {+46.81/9 1.5}
Ra8+ 18. Kb2 {+46.81/9 1.6} Rb8 19. Rf7 {+46.69/8 1.5} Kg3 20. Kc3
{+46.96/8 1.4} Rc8+ 21. Kd4 {+47.03/8 1.6} Rb8 {-1.37/14 1.4} 22. Kc4
{+46.61/9 1.6} Rc8+ 23. Kd5 {+46.52/8 1.5} Rb8 24. Rg7+ {+45.74/10 1.9} Kf4
25. Kc4 {+45.74/1} Rc8+ {-1.38/14 1.6} 26. Kb5 {+0.00/11 11} Rb8+
{-1.28/13 1.1} 27. Ka4 {+0.00/11 0.8} Ra8+ 28. Kb4 {+0.00/11 0.8} Rb8+
{-1.35/15 1.4} 29. Kc4 {+0.00/70} Rc8+ 30. Kd5 {+0.00/9 0.8} Rb8 31. Kc4
{+0.00/1}
{Draw by repetition} 1/2-1/2
-
- Posts: 2851
- Joined: Wed Mar 08, 2006 10:01 pm
- Location: Irvine, CA, USA
Re: New Scorpio bitbase files
I expect that would make the bitbases less compressible, at least in the most difficult cases. Probably not really an important issue, though.Tony wrote:That is a risk yes. But I think it can be greatly reduced by not creating the bitbases as bitbases or by converting them from dtm egtbs, but by converting them from DTZ tablebases.
That way you only have to search at most 99 ply to guarantee the best move. Anything below 99 poses a risk, but I seriously doubt it's a big one.
That's still a lot of plies to search for certainty, even in the end game. You could make things much easier by having both DTZ50 and DTZ25 bitbases. Once you get to a won position for DTZ50 you would start using DTZ25 in your search. If that isn't reliable enough DTZ17 and DTZ34 could be used instead, and the size of all three files would still be much less that the EGTB files, I expect.