Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: endgame from Kasparov-Leko

Author: Vasik Rajlich

Date: 07:21:21 03/06/05

Go up one level in this thread


On March 06, 2005 at 08:00:41, Dieter Buerssner wrote:

>On March 06, 2005 at 06:26:33, Vasik Rajlich wrote:
>
>>ok - I looked again. First of all, it's really easy to beat Shredder 9 in this
>>position. Another example, after 2. .. Kf6: 3. Kh5 Kg7 4. Kg5 Kf7 5. Kf5 Ke7 6.
>>Kg6 Ke6 7. Kg7 Ke7 8. d3 a6 9. b3 Kd7 10. Kf7 and now finally a huge fail low.
>
>Vasik, I might be able to give harder resistance than Shredder :-). After
>3...Kg7 Yace shows increasing score from the start. It suggests 3...a6 instead
>with draw score. After Kg7:

Interesting.

Here is the output from Shredder 9 after 3. .. Kg7 4. Kg5:

 20/25	 0:00 	-0.20 	4...Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a4 b6 8.b3 (441.113) 746
 21/30	 0:02 	-0.21 	4...Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 Ke7 8.a4 Ke6 9.a5
(1.624.341) 783
 22/28	 0:04 	-0.41 	4...Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a4 b6 8.b3 (3.422.641) 798
 23/36	 0:07 	-0.43 	4...Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 a6 8.a4 b6 9.Kg7 Ke7 10.a5
(6.401.181) 802
 24/42	 0:14 	-0.38 	4...Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 Ke7 8.a4 b6 9.Kg7 a6
10.Kg8 Ke8 11.Kh7 Kf7 (11.676.326) 802
 25/42	 0:31 	-0.60 	4...Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 Ke7 8.a4 Ke6 9.a5 Ke7
10.Kg7 Ke8 11.Kf6 (23.476.567) 755
 26/44	 0:52 	-0.60 	4...Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 Ke7 8.a4 Ke6 9.a5 Ke7
10.Kg7 Ke8 11.Kf6 (38.627.411) 741
 27/48	 2:03 	-0.60 	4...Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 Ke7 8.a4 Ke6 9.a5 Ke7
10.Kg7 Ke8 11.Kf6 (89.978.811) 729

Feeding it the next few moves and waiting a while is finally enough, ie. 4. Kg5
Kf7 5. Kf5 Ke7 and now:

 20/20	 0:01 	+0.60 	6.Kg6 Ke6 7.a3 Ke7 8.a4 Ke6 9.a5 Ke7 10.Kg7 Ke8 11.Kf6
(891.750) 816
 21/21	 0:01 	+0.60 	6.Kg6 Ke6 7.a3 Ke7 8.a4 Ke6 9.a5 Ke7 10.Kg7 Ke8 11.Kf6
(1.455.965) 794
 22/22	 0:03 	+0.60 	6.Kg6 Ke6 7.a3 Ke7 8.a4 Ke6 9.a5 Ke7 10.Kg7 Ke8 11.Kf6
(2.799.954) 798
 23/23	 0:04 	+0.60 	6.Kg6 Ke6 7.a3 Ke7 8.a4 Ke6 9.a5 Ke7 10.Kg7 Ke8 11.Kf6
(3.934.970) 793
 24/24	 0:07 	+0.60 	6.Kg6 Ke6 7.a3 Ke7 8.a4 Ke6 9.a5 Ke7 10.Kg7 Ke8 11.Kf6
(6.192.059) 794
 25/44	 0:49 	+0.77 	6.Kg6 Ke6 7.a3 Ke7 8.a4 Ke6 9.a5 Ke7 10.Kg7 Ke8
(38.354.821) 772
 26/45	 1:17 	+1.02++	6.Kg6 Ke6 (59.195.528) 761
 26/45	 1:31 	+1.52++	6.Kg6 Ke6 (69.018.262) 755

>
>  62101595  1:08.8   0.12 12.  4.Kg5 Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 Ke7 8.Kg7 b6
>                               9.d3 a6 10.a4 Kd7 11.Kf6 Kd8 12.Ke6 Kc7 13.Kf7
>                               Kd7 14.Kf8 Kd8 15.Kf7 Kd7 16.b3 a5 17.Kf6 Kc6
>                               18.Ke6 Kc7 19.Ke7 b5 20.cxb5 Kc8 21.b6 Kb7
>                               22.Kxd6 Kc8 23.b7+ Kb8 24.Kd7 c4 25.bxc4 Kxb7
>                               26.d4 Kb8 {HT} {160}
> 108549999  2:00.4   0.13 13t  4.Kg5 Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 Ke7 8.a4 b6
>                               9.Kg7 Ke6 10.Kf8H Kd7H 11.Kf7H Kc6H {HT} {0}
> 142095926  2:37.3   0.13 13.  4.Kg5 Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 Ke7 8.a4 b6
>                               9.Kg7 Ke6 10.Kf8 Kd7 11.Kf7 Kc6 {HT} {0}
> 222576761  4:06.4   0.53 14++ 4.Kg5 Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 Ke7 8.a4 b6
>                               9.Kg7 Ke6 10.Kf8 Kd7 11.Kf7 Kc6 12.Ke7 Kc7
>                               13.d3 Kc8 14.Kxd6 a6 15.b3 a5 16.Kc6 b5H
>                               17.cxb5H {HT} {160}
> 267155553  4:53.3   0.53 14t  4.Kg5 Kf7 5.Kf5 Ke7H 6.Kg6H Ke6H 7.a3H Ke7H
>                               8.Kg7H a6H 9.d3H b5H 10.b4H bxc4H 11.dxc4H d5H
>                               12.exd5H cxb4H 13.axb4H {HT} {80}
> 355448307  6:29.9   0.53 14.  4.Kg5 Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 Ke7 8.Kg7 a6
>                               9.d3 b5 10.b4 bxc4 11.dxc4 d5 12.exd5 cxb4
>                               13.axb4 Kd7H 14.c5H e4H 15.c6+H Kd6H 16.Kf6H
>                               e3H 17.Kf5H Kxd5H {0}
> 583617293 11:17.4   0.93 15++ 4.Kg5 Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 Ke7 8.Kg7 a6
>                               9.a4 b6 10.Kg6 Ke6 11.b3 Kd7 12.Kf6 a5 13.d3
>                               Kc6 14.Ke6 Kb7 15.Kxd6 Kb8 16.Kxe5 Kc8 17.Kd5
>                               Kc7 18.e5 {HT} {160}
> 976119218 18:20.8   1.55 15t  4.Kg5 Kf7 5.Kf5 Ke7 6.Kg6 Ke6 7.a3 Ke7 8.Kg7 a6
>                               9.a4 b6 10.b3 Ke8 11.Kf6 Kd7 12.Kf7 Kc8 13.Ke8
>                               b5H 14.axb5H Kb7H 15.Kd7H d5H 16.exd5H a5H
>                               17.d6H Kb8H 18.b6H Kb7H 19.Kd8H Kc6H 20.Ke8H
>                               Kxd6H 21.b7H Kc7H 22.b8=Q+H Kxb8H {HT} {0}
>
>>So - let me fix this position once again so that Yace won't get it (this time I
>>am 100% sure :)) :
>
>I am not sure, what you exactly mean by "won't get it?". Won't show a sensible
>score or PV? Just would play wrongly? Or do you want convincing score and PV in
>few seconds? THen Yace won't get it.
>
>> 8/1p6/3p1k2/p1p1p3/2P1P2K/1P1P4/P7/8 b - -
>
>
>In the other pos, I had a stable score at 0.0. Here I get slight negative score
>from the first minute, and even the first fail low soon. I'll let it run for a
>while.
>
> 116755354  2:04.0  -0.02 14.  1...Kg6 2.Kg4 b6 3.a3 a4 4.bxa4 Kf6 5.Kh5 Kf7
>                               6.Kh6 Kf6 7.a5 bxa5 8.a4 Kf7 9.Kh7 Kf6 10.Kg8
>                               Kg5 11.Kf8 Kf4 12.Ke7 Ke3 13.Kxd6 Kxd3 14.Kxc5
>                               Kxe4 15.Kd6H Kd4H 16.c5H Kd3H 17.c6H e4H
>                               18.Kd7H e3H 19.c7H e2H 20.c8=QH e1=QH {0}
> 182001917  3:14.4  -0.18 15t  1...Kg6 2.Kg4 b6 3.a3 a4 4.bxa4 Kf6 5.Kh5 Kf7
>                               6.Kh6 Kf6 7.a5 bxa5 8.a4 Kf7 9.Kh7 Kf8 10.Kg6
>                               Ke7 11.Kg7H d5H 12.exd5H e4H 13.dxe4H Kd7H
>                               14.d6H {HT} {-160}
> 254429367  4:34.1  -0.18 15.  1...Kg6 2.Kg4 b6 3.a3 a4 4.bxa4 Kf6 5.Kh5 Kf7
>                               6.Kh6 Kf6 7.a5 bxa5 8.a4 Kf7 9.Kh7 Kf8 10.Kg6
>                               Ke7 11.Kg7 d5 12.exd5 e4 13.dxe4 Kd7 14.d6 {HT}
>                               {-160}
> 293950804  5:17.4  -0.58 16-- 1...Kg6 2.Kg4 b6 3.a3 a4 4.bxa4 Kf6 5.Kh5 Kf7
>                               6.Kh6 Kf6 7.a5 d5 8.exd5 e4 9.dxe4 b5 10.cxb5
>                               Ke5 11.Kg6 c4 12.d6 Kxd6 13.b6 c3 14.e5+ Kd5
>                               15.a4 {HT} {-240}
> 435613647  7:54.2  -0.81 16t  1...Kg6 2.Kg4 Kf6 3.Kh5 Kf7 4.Kh6 Kf6 5.a3 Kf7
>                               6.Kh7 Kf6 7.Kg8 Kg5 8.Kf8 Kf4 9.Ke7 Ke3 10.Kxd6
>                               Kxd3 11.Kxe5 a4 12.bxa4 Kxc4 13.Kd6 Kd4 14.e5
>                               c4 15.e6 c3 16.e7 c2 17.e8=Q {-1001}
> 523569409  9:27.8  -0.81 16.  1...Kg6 2.Kg4 Kf6 3.Kh5 Kf7 4.Kh6 Kf6 5.a3 Kf7
>                               6.Kh7 Kf6 7.Kg8 Kg5 8.Kf8 Kf4 9.Ke7 Ke3 10.Kxd6
>                               Kxd3 11.Kxe5 a4 12.bxa4 Kxc4 13.Kd6 Kd4 14.e5
>                               c4 15.e6 c3 16.e7 c2 17.e8=Q {-1001}

Shredder 9 here:

 20/29	 0:01 	 0.00 	5...Kg6 6.Kg3 Kg5 7.Kf3 Kh4 8.Kf2 Kg4 9.Ke3 Kg5 10.Ke2 Kf4
(1.057.242) 851
 21/27	 0:02 	 0.00 	5...Kg6 6.Kg3 Kg5 7.Kf3 Kh4 8.Kf2 Kg4 9.Ke3 Kg5 10.Ke2 Kf4
(2.101.499) 826
 22/39	 0:05 	 0.00 	5...Kg6 6.Kg3 Kg5 7.Kf3 Kh4 8.Kf2 Kg4 9.Ke3 Kg5 10.Ke2 Kf4
(4.075.698) 807
 23/41	 0:09 	 0.00 	5...Kg6 6.Kg3 Kg5 7.Kf3 Kh4 8.Kf2 Kg4 9.Ke3 Kg5 10.Ke2 Kf4
(7.351.152) 788
 24/43	 0:18 	 0.00 	5...Kg6 6.Kg3 Kg5 7.Kf3 Kh4 8.Kf2 Kg4 9.Ke3 Kg5 10.Ke2 Kf4
(14.161.457) 757
 25/45	 0:33 	 0.00 	5...Kg6 6.Kg3 Kg5 7.Kf3 Kh4 8.Kf2 Kg4 9.Ke3 Kg5 10.Ke2 Kf4
(25.149.794) 747
 26/47	 0:57 	 0.00 	5...Kg6 6.Kg3 Kg5 7.Kf3 Kh4 8.Kf2 Kg4 9.Ke3 Kg5 10.Ke2 Kf4
(42.458.405) 734
 27/50	 1:58 	 0.00 	5...Kg6 6.Kg3 Kg5 7.Kf3 Kh4 8.Kf2 Kg4 9.Ke3 Kg5 10.Ke2 Kf4
(84.745.705) 714
 28/51	 4:16 	 0.00 	5...Kg6 6.Kg3 Kg5 7.Kf3 Kh4 8.Kf2 Kg4 9.Ke3 Kg5 10.Ke2 Kf4
(176.894.963) 690

Shredder has some trouble here.

Anyway, ideally what you would be able to do is evaluate a position like the
following (based on Kasparov-Leko):

[D] 2k2rn1/1p6/3p1p2/p1p1pR2/2P1P2B/1P1P4/P5P1/6K1 w - -

Here, any good human player would be able to calculate 1. g4 Nh6 2. Rxf6 Rxf6 3.
Bxf6 Nxg4 4. Bg5 Kd7 5. Kg2 Ke6 6. Kg3 Nf6 7. Bxf6 as winning. Black's moves are
all forced.

One idea would be to look for some sort of "advanced opposition" cases, where
you notice that one king is getting kicked out of some critical squares. But
there are so many exceptions ...

Vas

>
>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.