Author: Ferdinand S. Mosca
Date: 22:04:32 03/18/01
Go up one level in this thread
On March 18, 2001 at 15:53:35, Dieter Buerssner wrote:
>On March 18, 2001 at 04:12:41, Uri Blass wrote:
>
>>>>>> 8/8/pq1PQ1kp/6p1/8/7P/5PK1/8 b - -
>
>>Crafty18.03 under chessbase(64 mbytes hash) cannot reproduce it.
>
>I tried yet again, and I can reproduce the lines I had posted.
>
>>The score of it is 7.47 at depth 12 but 7.79 at depth 13 when the main line is
>>1...Kg7 2.Qd7+ at depth>=13
>>The difference between depthes 11 and 12 is 0.4 pawns and not 0.39 pawns.
>>
>>At depth 14 the score is +8.02 and it finished depth 14 in a few minutes on
>>pIII800 that is clearly less than your 20 minutes.
>
>I tested with 12M hash and 5M hashp. Perhaps because of the bigger transposition
>tables you used, you got even higer speedup, than the hardware difference would
>suggest. When I retried with slightly bigger hash, I have seen the next output
>after 13:46.
>
>Ok, once again, this time with 24M hash, 5M hashp, TBs disabled (to make it
>easier reproducable):
>
>Crafty v18.6
>[...]
>Black(1): st 100000
>search time set to 100000.00.
>Black(1): go
>end-game phase
> time surplus 0.00 time limit 1666:40 (1666:40)
> nss depth time score variation (1)
>[...]
> 7-> 0.95 8.30 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb3 4.
> d7 g4 5. Qd6+ Kg7 6. d8=Q Qxh3+
> 8 1.45 8.05 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
> d7 Qe4+ 5. Kh2 Qd4 6. d8=Q Qxf2+ 7.
> Kh1 Kf5 <HT>
> 8-> 1.48 8.05 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
> d7 Qe4+ 5. Kh2 Qd4 6. d8=Q Qxf2+ 7.
> Kh1 Kf5 <HT>
> 9 2.23 ++ 1. ... Kg7!!
> 9-> 3.90 7.66 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
> d7 Qe4+ 5. Kh2 Qd4 6. d8=Q Qxf2+ 7.
> Kh1 Kf5 8. Qf8+ <HT>
> 10 5.78 ++ 1. ... Kg7!!
> 10-> 10.99 7.27 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
> d7 Qe4+ 5. Kh2 Qd4 6. d8=Q Qxf2+ 7.
> Kh1 Kf5 8. Qf8+ <HT>
> 11 14.45 7.61 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
> d7 Qe4+ 5. Kh2 Qd4 6. Kg1 g4 7. d8=Q
> Qxd8 8. Qxd8 gxh3
> 11-> 14.68 7.61 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
> d7 Qe4+ 5. Kh2 Qd4 6. Kg1 g4 7. d8=Q
> Qxd8 8. Qxd8 gxh3
> 12 45.33 ++ 1. ... Kg7!!
> 12-> 1:15 7.22 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
> d7 Qe4+ 5. Kh2 Qd4 6. Kg1 g4 7. d8=Q
> Qxd8 8. Qxd8 gxh3
> 13 2:35 ++ 1. ... Kg7!!
> 13-> 3:56 6.83 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
> d7 Qe4+ 5. Kh2 Qd4 6. Kg1 g4 7. d8=Q
> Qxd8 8. Qxd8 gxh3
> 14 13:46 -- 1. ... Kg7
> 14 19:18 7.90 1. ... Kg7 2. Qd7+ Kg6 3. h4 g4 4.
> Qxg4+ Kf6 5. Qf4+ Ke6 6. Qxh6+ Kd5
> 7. Qh5+ Kxd6 8. Qh6+ Kc5 9. Qxb6+ Kxb6
> 10. h5 a5 11. h6 a4
> 14-> 19:20 7.90 1. ... Kg7 2. Qd7+ Kg6 3. h4 g4 4.
> Qxg4+ Kf6 5. Qf4+ Ke6 6. Qxh6+ Kd5
> 7. Qh5+ Kxd6 8. Qh6+ Kc5 9. Qxb6+ Kxb6
> 10. h5 a5 11. h6 a4
>Black(0): ?
> time=22:39 cpu=99% mat=0 n=241885205 fh=88% nps=177905
> ext-> chk=29059370 pp=1117756 1rep=1452948 mate=44333
> predicted=0 nodes=241885205 evals=16868856
> endgame tablebase-> probes done=0 successful=0
> hashing-> trans/ref=42% pawn=8% used=99%
>
>
>Can anybody can reproduce this? Otherwise, it would mean, that I have set up
>something totally wrong, which certainly would interest me.
Hello Dieter,
Here is the log on P3 667
using hash=24, hashp=8(as specified on crafty.rc)
Crafty v18.6
White(1): st 10000
search time set to 10000.00.
White(1): setb 8/8/pq1PQ1kp/6p1/8/7P/5PK1/8 b - -
Black(1): go
end-game phase
clearing hash tables
time surplus 0.00 time limit 166:40 (166:40)
nss depth time score variation (1)
8 0.41 8.05 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
d7 Qe4+ 5. Kh2 Qd4 6. d8=Q Qxf2+ 7.
Kh1 Kf5 <HT>
8-> 0.47 8.05 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
d7 Qe4+ 5. Kh2 Qd4 6. d8=Q Qxf2+ 7.
Kh1 Kf5 <HT>
9 0.83 ++ 1. ... Kg7!!
9-> 1.53 7.66 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
d7 Qe4+ 5. Kh2 Qd4 6. d8=Q Qxf2+ 7.
Kh1 Kf5 8. Qf8+ <HT>
10 2.41 ++ 1. ... Kg7!!
10-> 4.93 7.27 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
d7 Qe4+ 5. Kh2 Qd4 6. d8=Q Qxf2+ 7.
Kh1 Kf5 8. Qf8+ <HT>
11 6.84 7.61 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
d7 Qe4+ 5. Kh2 Qd4 6. Kg1 g4 7. d8=Q
Qxd8 8. Qxd8 gxh3
11-> 6.93 7.61 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
d7 Qe4+ 5. Kh2 Qd4 6. Kg1 g4 7. d8=Q
Qxd8 8. Qxd8 gxh3
12 21.45 ++ 1. ... Kg7!!
12-> 35.77 7.22 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
d7 Qe4+ 5. Kh2 Qd4 6. Kg1 g4 7. d8=Q
Qxd8 8. Qxd8 gxh3
13 1:13 ++ 1. ... Kg7!!
13-> 1:58 6.83 1. ... Kg7 2. Qe7+ Kg6 3. Qc7 Qb1 4.
d7 Qe4+ 5. Kh2 Qd4 6. Kg1 g4 7. d8=Q
Qxd8 8. Qxd8 gxh3
14 7:50 -- 1. ... Kg7
14 10:38 7.90 1. ... Kg7 2. Qd7+ Kg6 3. h4 g4 4.
Qxg4+ Kf6 5. Qf4+ Ke6 6. Qxh6+ Kd5
7. Qh5+ Kxd6 8. Qh6+ Kc5 9. Qxb6+ Kxb6
10. h5 a5 11. h6 a4
14-> 10:54 7.90 1. ... Kg7 2. Qd7+ Kg6 3. h4 g4 4.
Qxg4+ Kf6 5. Qf4+ Ke6 6. Qxh6+ Kd5
7. Qh5+ Kxd6 8. Qh6+ Kc5 9. Qxb6+ Kxb6
10. h5 a5 11. h6 a4
15 15:43 -- 1. ... Kg7
15 23:00 1/3* 1. ... Kg7
I think it is similar.
There might be some differences under chessbase?
On my new setting, Yace does not even consider Qe7, after Kg7, here it is:
white ( 1): input: st 10000
white ( 1): input: setb 8/8/pq1PQ1kp/6p1/8/7P/5PK1/8 b - -
Stored 0 learned positions into hash table
black ( 1): input: go
usetime = 9999.00, mintime = 9999.00 maxtime = 9999.00 tl 9999.00 ml 0
4 0.051 -0.20 1t 1...Kh5 {0}
6 0.055 -0.20 1. 1...Kh5 {0}
65 0.085 -0.26 2t 1...Kh5 2. Qg4+ Kg6 {0}
104 0.122 -0.26 2. 1...Kh5 2. Qg4+ Kg6 {0}
1181 0.132 -0.28 3t 1...Kh5 2. Qe2+ Kg6 3. Qe8+ Kf5 4. d7 Qc6+ {0}
1473 0.140 -0.28 3. 1...Kh5 2. Qe2+ Kg6 3. Qe8+ Kf5 4. d7 Qc6+ {0}
4082 0.177 -0.31 4t 1...Kh5 2. Qe2+ Kg6 3. Qe8+ Kf5 4. d7 Qc6+ 5.
f3 {0}
4619 0.186 -0.31 4. 1...Kh5 2. Qe2+ Kg6 3. Qe8+ Kf5 4. d7 Qc6+ 5.
f3 {0}
16496 0.393 -0.41 5t 1...Kh5 2. Qe2+ Kg6 3. Qe8+ Kf6 4. d7 Qc6+ 5.
f3 Qc2+ 6. Kg3 Qc7+ {0}
18417 0.407 -0.41 5. 1...Kh5 2. Qe2+ Kg6 3. Qe8+ Kf6 4. d7 Qc6+ 5.
f3 Qc2+ 6. Kg3 Qc7+ {0}
34489 0.554 -1.26 6-- 1...Kh5 2. Qf7+ Kh4 3. Qc7 Qd4 4. d7 Qe4+H 5.
Kh2H Qf4+H 6. Qxf4+H gxf4H {0}
45139 0.646 -18.83 6t 1...Kh5 2. Kg3 Qb3+ 3. Qxb3 Kg6 4. d7 Kf6 5.
Qb6+ Kf5 6. d8=Q {-1922}
51177 0.663 -14.03 6t+ 1...Kh7 2. Qf7+ Kh8 3. d7 Qd4 4. Qf8+ Kh7 5.
d8=R
70023 0.728 -8.40 6t 1...Kh7 2. Qf7+ Kh8 3. Qc7 Qxc7 4. dxc7 a5 5.
c8=Q+ Kg7 {-921}
74383 0.744 -8.39 6t+ 1...Kg7 2. Qe5+ Kg8 3. Qe8+ Kg7 4. Qc8 Kg6 5.
f3 {0}
95706 0.810 -7.67 6t 1...Kg7 2. Qd7+ Kg6 3. Qc7 Qb3 4. d7 g4 5. d8=Q
Qxh3+ 6. Kg1 {-841}
95706 0.817 -7.67 6. 1...Kg7 2. Qd7+ Kg6 3. Qc7 Qb3 4. d7 g4 5. d8=Q
Qxh3+ 6. Kg1 {-841}
129282 0.888 -7.75 7t 1...Kg7 2. Qd7+ Kg6 3. Qc7 Qb3 4. d7 g4 5. Qd6+
Kg7 6. d8=Q Qxh3+ 7. Kg1 {-841}
132416 0.903 -7.75 7. 1...Kg7 2. Qd7+ Kg6 3. Qc7 Qb3 4. d7 g4 5. Qd6+
Kg7 6. d8=Q Qxh3+ 7. Kg1 {-841}
200422 1.082 -7.80 8t 1...Kg7 2. Qd7+ Kg6 3. Qc7 Qb3 4. d7 g4 5. Qd6+
Kg7 6. Qe7+ Kg6 7. Qe4+ Kf7 8. d8=Q Qxh3+ 9.
Kg1 {-841}
204735 1.104 -7.80 8. 1...Kg7 2. Qd7+ Kg6 3. Qc7 Qb3 4. d7 g4 5. Qd6+
Kg7 6. Qe7+ Kg6 7. Qe4+ Kf7 8. d8=Q Qxh3+ 9.
Kg1 {-841}
449437 1.772 -7.89 9t 1...Kg7 2. Qd7+ Kg6 3. Qc7 Qb3 4. d7 Qd5+ 5.
Kh2 Qd4 6. d8=Q Qxf2+ 7. Kh1 {-841}
460849 1.822 -7.89 9. 1...Kg7 2. Qd7+ Kg6 3. Qc7 Qb3 4. d7 Qd5+ 5.
Kh2 Qd4 6. d8=Q Qxf2+ 7. Kh1 {-841}
1447447 4.673 -7.49 10++ 1...Kg7 2. Qd7+ Kg6 3. Qc7 Qb3 4. d7 Qd5+ 5.
Kh2 Qd4 6. d8=R Qxf2+ 7. Kh1 {-340}
5245364 17.283 -0.53 10t 1...Kg7 2. Qd7+ Kg6 3. h4 a5 4. hxg5 Kxg5 5.
Qg7+ Kh5 6. f3 Qb4 7. d7 Qh4 {0}
5323482 17.582 -0.53 10. 1...Kg7 2. Qd7+ Kg6 3. h4 a5 4. hxg5 Kxg5 5.
Qg7+ Kh5 6. f3 Qb4 7. d7 Qh4 {0}
8828718 28.760 -0.55 11t 1...Kg7 2. Qd7+ Kg6 3. h4 Qd4 4. Qe6+ Qf6 5.
Qe8+ Kg7 6. d7 Qc6+ 7. f3 Qc2+ 8. Kg3 gxh4+ 9.
Kxh4 Qf2+ 10. Kg4 Qg2+ 11. Kf4 {0}
8927532 29.111 -0.55 11. 1...Kg7 2. Qd7+ Kg6 3. h4 Qd4 4. Qe6+ Qf6 5.
Qe8+ Kg7 6. d7 Qc6+ 7. f3 Qc2+ 8. Kg3 gxh4+ 9.
Kxh4 Qf2+ 10. Kg4 Qg2+ 11. Kf4 {0}
11733688 38.184 -0.95 12-- 1...Kg7 2. Qd7+ Kg6 3. h4 Qd4 4. h5+ Kxh5 5.
Qe8+ Kh4 6. Qe6 Kh5 7. d7 a5 8. Qe8+ Kh4 9.
d8=Q Qd7 {-841}
17490192 55.875 -1.73 12t 1...Kg7 2. Qd7+ Kg6 3. h4 gxh4 4. Qe8+ Kf6 5.
d7 Qc6+ 6. Kh2 Qc7+ 7. Kh3 Qc3+ 8. Kxh4 Qc4+ 9.
Kg3 Qc7+ 10. f4 Qc3+ 11. Kg4 h5+ 12. Kxh5 {-80}
17623480 56.330 -1.73 12. 1...Kg7 2. Qd7+ Kg6 3. h4 gxh4 4. Qe8+ Kf6 5.
d7 Qc6+ 6. Kh2 Qc7+ 7. Kh3 Qc3+ 8. Kxh4 Qc4+ 9.
Kg3 Qc7+ 10. f4 Qc3+ 11. Kg4 h5+ 12. Kxh5 {-80}
20261962 1:04.4 -2.13 13-- 1...Kg7 2. Qd7+ Kg6 3. h4 gxh4 4. Qe8+ Kf6 5.
d7 Qc6+ 6. Kh2 Qc7+ 7. Kh3 Qc3+ 8. Kxh4 Qc4+ 9.
Kg3 Kf5 10. d8=Q a5 {-921}
32293344 2:04.8 -2.76 13t 1...Kg7 2. Qd7+ Kg6 3. h4 gxh4 4. Qe8+ Kg7 5.
Qe7+ Kh8 6. d7 Qc6+ 7. Kh2 Qc7+ 8. Kh3 Qc3+ 9.
Qe3 Qf6 10. Qe8+ Kh7 11. d8=R Qxf2 {-260}
32755120 2:09.4 -2.76 13. 1...Kg7 2. Qd7+ Kg6 3. h4 gxh4 4. Qe8+ Kg7 5.
Qe7+ Kh8 6. d7 Qc6+ 7. Kh2 Qc7+ 8. Kh3 Qc3+ 9.
Qe3 Qf6 10. Qe8+ Kh7 11. d8=R Qxf2 {-260}
39530451 2:34.3 -3.16 14-- 1...Kg7 2. Qd7+ Kg6 3. h4 gxh4 4. Qe8+ Kg7 5.
Qe7+ Kh8 6. d7 Qc6+ 7. Kh2 Qc7+ 8. Kh3 Qc3+ 9.
Qe3 Qf6 10. Qe8+ Kg7 11. d8=Q a5 {-841}
66395034 3:50.9 -2.86 14t 1...Kg7 2. Qd7+ Kg6 3. h4 gxh4 4. Qe8+ Kg7 5.
Qe7+ Kh8 6. d7 Qc6+ 7. Kh2 Qc7+ 8. Kh3 Qc3+ 9.
Qe3 Qf6 10. Qe8+ Kh7 11. d8=R Qxf2 12. Qg8#
{-260}
67824525 3:55.3 -2.86 14. 1...Kg7 2. Qd7+ Kg6 3. h4 gxh4 4. Qe8+ Kg7 5.
Qe7+ Kh8 6. d7 Qc6+ 7. Kh2 Qc7+ 8. Kh3 Qc3+ 9.
Qe3 Qf6 10. Qe8+ Kh7 11. d8=R Qxf2 12. Qg8#
{-260}
82149397 4:38.3 -3.26 15-- 1...Kg7 2. Qd7+ Kg6 3. h4 gxh4 4. Qe8+ Kg7 5.
Qe7+ Kh8 6. d7 Qc6+ 7. Kh2 Qc7+ 8. Kh3 a5
174048747 9:09.6 -7.84 15t 1...Kg7 2. Qd7+ Kg6 3. h4 gxh4 4. Qe8+ Kf6 5.
Qe7+ Kf5 6. Qf7+ Ke5 7. d7 Kd4 8. Qe7 Qc6+ 9.
Kh2 Qc7+ 10. Kh3 Kc3 11. Qf6+ Kb3 12. d8=Q Qxd8
13. Qxd8 {HT} {-841}
177484978 9:20.1 -7.84 15. 1...Kg7 2. Qd7+ Kg6 3. h4 gxh4 4. Qe8+ Kf6 5.
Qe7+ Kf5 6. Qf7+ Ke5 7. d7 Kd4 8. Qe7 Qc6+ 9.
Kh2 Qc7+ 10. Kh3 Kc3 11. Qf6+ Kb3 12. d8=Q Qxd8
13. Qxd8 {HT} {-841}
209641268 10:54.6 -7.84 15f. 1...Kg7 2. Qd7+ Kg6 3. h4 gxh4 4. Qe8+ Kf6 5.
Qe7+ Kf5 6. Qf7+ Ke5 7. d7 Kd4 8. Qe7 Qc6+ 9.
Kh2 Qc7+ 10. Kh3 Kc3 11. Qf6+ Kb3 12. d8=Q Qxd8
13. Qxd8 {-841}
209641268 Nodes, 15.72% Leavenodes, 320223 Nodes/sec
106309315 eval, 48.50% score, 1140706406 genmoves, 3.69% captures le 270/333
ext: pawn 1148992, rcp 161238, chk 23467042, repchk 5350954, null 87788, prune
7194541
htable: 142913525 store, 0 rejected, 207666159 probe, 42.9% f/p, 62.3% f/s
entries 2666664 age 1 renew 0
egtb probes 2409, found 753
The changes are:
icheck 0.6
irepcheck 1.0
My idea at first is to discourage Yace from making some checks, Yace is very
sensitive to checks, especially at endings, and I raise the value of irepcheck
on the idea that there is a serious bussiness when there is repetition of check.
If you can I need some explanation on extension parameters.
BTW a position from a game played by Yace:
Yace 0.99.09 (Paderborn) - LG2000v3(1-0), G10
[D]8/5k2/2p5/2Pq4/5PpP/4Q1K1/8/8 w - - 0 1
solution fastest to see f5
On a new setting Yace sees f5 quickly than the default.
12.860 1.44 12t+ (new setting)
2:35.8 1.87 13++ (default)
This is also the setting where Yace tried not to like Rc3 from my other post.
Currently I have run some test games around this setting.
It won against LG2000v3 at G15 in a 10 game match, I'll have to examine yet
those games.
Regards,
Dinan
>
>>I can add that Qe7+ is also an easy win for white when the best line includes a
>>repetition
>>1.Qe7+ Kg6 2.Qe6+ Kg7 3.Qd7+
>
>Sure. But I think, it is quite unlikely, that a chess engine will show a main
>line starting like this.
>
>Regards,
>Dieter
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.