Author: Dann Corbit
Date: 10:36:42 02/13/01
Go up one level in this thread
On February 13, 2001 at 10:47:03, Dieter Buerssner wrote:
>On February 13, 2001 at 02:02:37, Dann Corbit wrote:
>
>>title position 100
>
>Yace needs rather long for this. With more time, it will come up with b6+ or
>with Be3, which both are winning. In the log you have shown, there was
>Be3 in between, but for the wrong reason, because there was no b6 in the PV.
>BTW. With this position, I get very bit changes in solution time for very small
>changes for some null move strategies (dynamical null move or not, with little
>material doing still rather aggressive null move. Do it only far away from the
>root or not, etc.) I remember times between 8 seconds and 80 seconds on my
>computer, and finding the solution between depth 15 and 20. Judging from the NPS
>of the log, it looks a bit suspicios, that Yace did not find this within a
>minute on your computer.
>
>>title position 141
>>solution Qxf4
>>This one (WAC 141) was just a little slow. Took 14 seconds to find the right
>>answer and 25 seconds to recognize a checkmate.
>
>You are right. I am also wondering about this. There was allready the
>explaination of null move. I also accidently found, that some small changes, to
>when to do recapture extensions, can make Yace find this solution rather fast.
>
>In recent versions, Yace will often find mates a bit slower than earlier
>versions, because it won't try that hard to extend matethreats anamore, when the
>position looks allread won. So, when an earlier version came up with a mate
>score, now it may come up with a convincingly high score only, and delay the
>mate score fro one or 2 plies.
>
>>title position 163
>>solution Qg2+
>> 5291176 10.634 Mat11 10t 1...Qg2+ 2.Nxg2 hxg2+ 3.Kxg2 Bf3+ 4.Qxf3 exf3+
>> 5.Kg1 Rf5 6.Rae1 Rfh5 7.Re8+ Kf7 8.Re7+ Kxe7
>> 9.Re1+ Kf8 10.Re8+ Kxe8 11.d6 Rh1# {920}
>
>>Just a bit slow on this one again. It took 9.5 seconds to find it and 10.5 to
>>recognize checkmate.
>
>Is 10 seconds bad, for a mate in 11? BTW. Here it is similar than with 141.
>
>>title position 230
>>solution Rb4
>
>This one is unfortunately impossible for Yace :-(
>
>>According to my reconing, Yace only missed one or two (depending upon how we are
>>to interpret WAC 100).
>
>It missed certainly 2. On my computer (AMD K6-2 475 with 20 M hash) I get
>usually 296 with 10 seconds, with the 4 missing ones cited above.
I should mention that Yace's score is the best of any computer program at my
disposal for WAC at 5 seconds per position. If you have improved it, then that
is even more astounding.
I ran two of the most problematic at long time controls on a fast machine with
Crafty. Here is the output:
choose from book moves randomly (using weights.)
choose from 5 best moves.
book learning enabled
result learning enabled
position learning enabled
threshold set to 9 pawns.
4 piece tablebase files found
1302kb of RAM used for TB indices and decompression tables
Crafty v18.3
White(1): st 10000
search time set to 10000.00.
White(1): epdpfga rh.epd rh.out
PFGA: EPD record: 1 ID: WAC 100
end-game phase
clearing hash tables
time surplus 0.00 time limit 166:40 (166:40)
nss depth time score variation (1)
11 0.19 4.10 1. Be3 Bd6 2. Bf2 Kb6 3. Kb3 Bb8 4.
Bg3 Bc7 5. Kc3 Bd6 6. Kd3
11-> 0.25 4.10 1. Be3 Bd6 2. Bf2 Kb6 3. Kb3 Bb8 4.
Bg3 Bc7 5. Kc3 Bd6 6. Kd3
12 0.31 3.91 1. Be3 Bd6 2. Kb3 Kb6 3. Kc3 Bb8 4.
Kd3 Ba7 5. Bd2 Bb8 6. Ke3 Ba7
12 0.49 4.21 1. Kb3 Bb6 2. Kc3 Bc7 3. Kd3 Bd6 4.
Ke3 Bf8 5. Bc3 Be7 6. Ba5 Bd6
12-> 0.57 4.21 1. Kb3 Bb6 2. Kc3 Bc7 3. Kd3 Bd6 4.
Ke3 Bf8 5. Bc3 Be7 6. Ba5 Bd6
13 0.66 4.26 1. Kb3 Bb6 2. Kc3 Bc7 3. Kd3 Bd6 4.
Ke3 Bf8 5. Bc3 Be7 6. Bb2 Kb6 7. Bc3
13-> 0.91 4.26 1. Kb3 Bb6 2. Kc3 Bc7 3. Kd3 Bd6 4.
Ke3 Bf8 5. Bc3 Be7 6. Bb2 Kb6 7. Bc3
14 1.18 4.21 1. Kb3 Bb6 2. Kc3 Bc7 3. Kd3 Bd6 4.
Ke3 Bf8 5. Be1 Bd6 6. Ba5 Bb8 7. Bc3
Bd6
14-> 1.97 4.21 1. Kb3 Bb6 2. Kc3 Bc7 3. Kd3 Bd6 4.
Ke3 Bf8 5. Be1 Bd6 6. Ba5 Bb8 7. Bc3
Bd6
15 2.30 4.24 1. Kb3 Bb6 2. Kc3 Bc7 3. Kd3 Bd6 4.
Ke3 Bc7 5. Bc3 Bd8 6. Kf2 Bc7 7. Kg3
Bd8 8. Kf3
15-> 3.61 4.24 1. Kb3 Bb6 2. Kc3 Bc7 3. Kd3 Bd6 4.
Ke3 Bc7 5. Bc3 Bd8 6. Kf2 Bc7 7. Kg3
Bd8 8. Kf3
16 4.77 4.21 1. Kb3 Bb6 2. Kc3 Bc7 3. Kd3 Bd6 4.
Ke3 Bc7 5. Bc1 Bb6 6. Bb2 Bd8 7. Bc3
Be7 8. Ba5 Bd6
16 7.92 4.30 1. b6+ Bxb6 2. Kb5 Bc7 3. Be3 Bb8 4.
Bxc5+ Ka8 5. Be3 Bd6 6. c5 Bc7 7. c6
Kb8 8. Kc4 Ka8 9. Bd4 Kb8 10. Bxf6
Bxf4
16-> 8.51 4.30 1. b6+ Bxb6 2. Kb5 Bc7 3. Be3 Bb8 4.
Bxc5+ Ka8 5. Be3 Bd6 6. c5 Bc7 7. c6
Kb8 8. Kc4 Ka8 9. Bd4 Kb8 10. Bxf6
Bxf4
17 9.31 4.37 1. b6+ Bxb6 2. Kb5 Bc7 3. Be3 Bb8 4.
Bxc5+ Ka8 5. Be3 Bd6 6. Kc6 Bc7 7.
Kd7 <HT>
17-> 18.96 4.37 1. b6+ Bxb6 2. Kb5 Bc7 3. Be3 Bb8 4.
Bxc5+ Ka8 5. Be3 Bd6 6. Kc6 Bc7 7.
Kd7 <HT>
18 20.30 4.48 1. b6+ Bxb6 2. Kb5 Bc7 3. Be3 Bb8 4.
Bxc5+ Ka8 5. Be3 Bd6 6. Kc6 Bc7 7.
Kd7 Ba5 8. Ke6 Be1 9. Kxf6 Bxh4+ 10.
Kxf5 Bg3 11. Ke6 h4
18-> 38.53 4.48 1. b6+ Bxb6 2. Kb5 Bc7 3. Be3 Bb8 4.
Bxc5+ Ka8 5. Be3 Bd6 6. Kc6 Bc7 7.
Kd7 Ba5 8. Ke6 Be1 9. Kxf6 Bxh4+ 10.
Kxf5 Bg3 11. Ke6 h4
19 42.82 4.53 1. b6+ Bxb6 2. Kb5 Bc7 3. Be3 Bb8 4.
Bxc5+ Ka8 5. Be3 Bc7 6. c5 Kb8 7. c6
Bd6 8. Bb6 Bxf4 9. a7+ Ka8 10. Ka6
Bd6 11. c7 Bxc7 12. Bxc7 f4
19 1:29 4.62 1. Be3 Bd6 2. Ka5 Bc7+ 3. b6+ Bxb6+
4. Kb5 Bc7 5. Bxc5+ Ka8 6. Be3 Bb8
7. Kc6 Bc7 8. Kd7 Ba5 9. Ke6 <HT>
19-> 1:36 4.62 1. Be3 Bd6 2. Ka5 Bc7+ 3. b6+ Bxb6+
4. Kb5 Bc7 5. Bxc5+ Ka8 6. Be3 Bb8
7. Kc6 Bc7 8. Kd7 Ba5 9. Ke6 <HT>
20 1:40 ++ 1. Be3!!
20 1:48 5.18 1. Be3 Bb6 2. Bf2 Kb8 3. Be1 Ka8 4.
Ba5 Ka7 5. Bxb6+ Kxb6 6. a7 Kxa7 7.
Ka5 Kb7 8. b6 Kb8 9. Kb5 Kb7 10. Kxc5
Kc8 11. Kc6 Kd8
20-> 3:36 5.18 1. Be3 Bb6 2. Bf2 Kb8 3. Be1 Ka8 4.
Ba5 Ka7 5. Bxb6+ Kxb6 6. a7 Kxa7 7.
Ka5 Kb7 8. b6 Kb8 9. Kb5 Kb7 10. Kxc5
Kc8 11. Kc6 Kd8
21 3:43 ++ 1. Be3!!
21-> 9:56 5.57 1. Be3 Bb6 2. Bf2 Kb8 3. Be1 Ka8 4.
Ba5 Ka7 5. Bxb6+ Kxb6 6. a7 Kxa7 7.
Ka5 Kb7 8. b6 Kb8 9. Kb5 Kb7 10. Kxc5
Kc8 11. Kc6 Kd8
22 10:39 ++ 1. Be3!!
22 14:07 6.53 1. Be3 Bd6 2. Ka5 Bc7+ 3. b6+ Bxb6+
4. Kb5 Bc7 5. Bxc5+ Ka8 6. Bf8 Bxf4
7. c5 Bc7 8. Bd6 Bd8 9. c6 f4 10. Bxf4
Ka7 11. c7 Bxc7 12. Bxc7 f5 13. Bf4
Ka8 14. Kb6
22-> 27:25 6.53 1. Be3 Bd6 2. Ka5 Bc7+ 3. b6+ Bxb6+
4. Kb5 Bc7 5. Bxc5+ Ka8 6. Bf8 Bxf4
7. c5 Bc7 8. Bd6 Bd8 9. c6 f4 10. Bxf4
Ka7 11. c7 Bxc7 12. Bxc7 f5 13. Bf4
Ka8 14. Kb6
23 36:30 -- 1. Be3
23 48:58 ++ 1. Be3!!
time=166:40 cpu=99% mat=1 n=3109304801 fh=3% nps=310930
ext-> chk=1346898566 cap=0 pp=32846473 1rep=93863291 mate=374202
predicted=0 nodes=3109304801 evals=229435818
endgame tablebase-> probes done=513 successful=513
PFGA: EPD record: 2 ID: WAC 230
clearing hash tables
time surplus 0.00 time limit 166:40 (166:40)
nss depth time score variation (1)
9-> 0.22 -2.73 1. ... a4 2. Kf4 Rh7 3. f3 Rh1 4. Ke3
Bb7 5. Ba3 Kd7
10 0.34 -2.73 1. ... a4 2. Kf4 Rh7 3. f3 Rh1 4. Ke3
Bb7 5. Ba3 Kd7 6. Kf4
10-> 0.38 -2.73 1. ... a4 2. Kf4 Rh7 3. f3 Rh1 4. Ke3
Bb7 5. Ba3 Kd7 6. Kf4
11 1.08 -2.92 1. ... a4 2. f3 Rh7 3. Ba3 Rh5 4. Kg4
Rh1 5. Bf8 Rc1 6. Bb4 Rh1
11 1.67 -2.95 1. ... Rh7 2. Ba3 Rh1 3. Re2 a4 4.
f3 Bb7 5. Rh2 Ra1 6. Bb2 Rf1 <HT>
11-> 1.73 -2.95 1. ... Rh7 2. Ba3 Rh1 3. Re2 a4 4.
f3 Bb7 5. Rh2 Ra1 6. Bb2 Rf1 <HT>
12 2.68 -2.95 1. ... Rh7 2. Ba3 a4 3. Re2 Rh5 4.
Kf4 Rh4+ 5. Kg3 Rh8 6. Rb2 Rh1 7. f4
Bb7
12-> 2.73 -2.95 1. ... Rh7 2. Ba3 a4 3. Re2 Rh5 4.
Kf4 Rh4+ 5. Kg3 Rh8 6. Rb2 Rh1 7. f4
Bb7
13 6.54 -3.00 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
f4 Ba6 5. Kg2 <HT>
13-> 6.66 -3.00 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
f4 Ba6 5. Kg2 <HT>
14 9.19 -2.94 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kf4 Rh2 5. Kg3 Rh8 6. Bb2 Rh5 7. Kf4
Rh4+ 8. Kg3 Rh8
14-> 9.64 -2.94 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kf4 Rh2 5. Kg3 Rh8 6. Bb2 Rh5 7. Kf4
Rh4+ 8. Kg3 Rh8
15 14.02 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Bb7 5. Ra1 Kd7 6. Rb1 Rh2 7. Kg3
Rh5 <HT>
15-> 21.10 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Bb7 5. Ra1 Kd7 6. Rb1 Rh2 7. Kg3
Rh5 <HT>
16 27.57 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Bb7 5. Ra1 Kd7 6. Rb1 Rh2 7. Kg3
Rh5 <HT>
16-> 34.57 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Bb7 5. Ra1 Kd7 6. Rb1 Rh2 7. Kg3
Rh5 <HT>
17 42.07 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Bb7 5. Ra1 Kd7 6. Rb1 Rh2 7. Kg3
Rh5 <HT>
17-> 51.87 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Bb7 5. Ra1 Kd7 6. Rb1 Rh2 7. Kg3
Rh5 <HT>
18 1:03 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Bb7 5. Ra1 Kd7 6. Rb1 Rh2 7. Kg3
Rh5 <HT>
18-> 1:13 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Bb7 5. Ra1 Kd7 6. Rb1 Rh2 7. Kg3
Rh5 <HT>
19 1:49 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Bb7 5. Ra1 Kd7 6. Rb1 Rh2 7. Kg3
Rh7 8. Kg4 Rh8 9. Rb2 Rh1 10. f3 Bc6
19-> 2:06 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Bb7 5. Ra1 Kd7 6. Rb1 Rh2 7. Kg3
Rh7 8. Kg4 Rh8 9. Rb2 Rh1 10. f3 Bc6
20 3:17 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Kb6 5. Ra1 Bd7 6. Bb2 Be8 7. Rb1
Kc6 8. Ra1 Kc7 9. Rb1 Kd7 10. f3 Rh2
11. Kf4
20-> 3:41 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Kb6 5. Ra1 Bd7 6. Bb2 Be8 7. Rb1
Kc6 8. Ra1 Kc7 9. Rb1 Kd7 10. f3 Rh2
11. Kf4
21 5:56 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Kb6 5. Ra1 Bd7 6. Bb2 Be8 7. Ba3
Kb5 8. Rb1 Kc6 9. Bb4 Rh2 10. Kg3 Rh8
11. Rb2 Kd7
21-> 6:50 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Kb6 5. Ra1 Bd7 6. Bb2 Be8 7. Ba3
Kb5 8. Rb1 Kc6 9. Bb4 Rh2 10. Kg3 Rh8
11. Rb2 Kd7
22 11:34 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Kb6 5. Ra1 Bd7 6. Bb2 Be8 7. Ba3
Kb5 8. Rb1 Kc6 9. Bb4 Kd7 10. Ba3 Bf7
11. f3 Rh2 12. Rb2
22-> 13:35 -2.96 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
Kg4 Kb6 5. Ra1 Bd7 6. Bb2 Be8 7. Ba3
Kb5 8. Rb1 Kc6 9. Bb4 Kd7 10. Ba3 Bf7
11. f3 Rh2 12. Rb2
23 72:55 -2.93 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
f4 Bd7 5. Kg2 Rh4 6. Kg3 Rh7 7. Kg2
Rh8 8. Rh1 Rb8 9. Bb2 Ra8 <HT>
23-> 111:18 -2.93 1. ... Rh7 2. Rb1 a4 3. Ba3 Rh5 4.
f4 Bd7 5. Kg2 Rh4 6. Kg3 Rh7 7. Kg2
Rh8 8. Rh1 Rb8 9. Bb2 Ra8 <HT>
time=166:40 cpu=99% mat=1 n=1967244388 fh=3% nps=196724
ext-> chk=222420030 cap=0 pp=58098434 1rep=5994251 mate=110460
predicted=0 nodes=1967244388 evals=1299650885
endgame tablebase-> probes done=0 successful=0
Black(1):
Black(1): quit
execution complete.
If I recall correctly, crafty used to get this one correct after a long think.
Maybe the overall play is better for the more recent versions that don't solve
it. I wonder if Hossa can solve this one, as it has some sort of blockage
recognition (IIRC).
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.