Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Hard endgame position

Author: Dave Gomboc

Date: 07:28:56 03/01/00

Go up one level in this thread


On March 01, 2000 at 09:52:45, Bernhard Bauer wrote:

>On March 01, 2000 at 09:15:18, James T. Walker wrote:
>
>>On March 01, 2000 at 08:48:41, Bernhard Bauer wrote:
>>
>>>On March 01, 2000 at 08:13:11, Dave Gomboc wrote:
>>>
>>>>On March 01, 2000 at 03:42:23, Bernhard Bauer wrote:
>>>>
>>>>>On March 01, 2000 at 02:58:49, Dann Corbit wrote:
>>>>>
>>>>>>On March 01, 2000 at 01:37:17, Janos Keinrath wrote:
>>>>>>
>>>>>>>Hi!
>>>>>>>
>>>>>>>Yesterday I saw an intresting endgame position on ICC.
>>>>>>>
>>>>>>>[D]8/8/6k1/3p4/3P4/2P2PKp/pB4b1/8 b - -
>>>>>>>
>>>>>>>This is the final position of a blitz game played by Amateur
>>>>>>>with black. The game ended with rep. draw, but I feel black somehow
>>>>>>>can win. Could you check the position? What is the winning plan?
>>>>>>
>>>>>>Crafty likes Kf5 from the get-go, but the evaluation does not climb over one
>>>>>>pawn until ply 19.  I'd be surprised if programs see a win at normal game-type
>>>>>>time controls (other than 'postal' games).
>>>>>
>>>>>This position is a simple win for black.
>>>>>For example:
>>>>>1...   Kf5
>>>>>2.Ba1  Bf1
>>>>>3.Bb2  Be2
>>>>>4.Ba1  h2
>>>>>5.Kxh2 Kf4
>>>>>5.Kg2  Bxf3
>>>>>6.Kf2  Ke4
>>>>>7.Ke1  Kd3
>>>>>and black wins the white bishop and the game.
>>>>>White may defend in a different way, but without any hope.
>>>>>A human will win this position easyly.
>>>>>
>>>>>Crafty does very poor on this kind of positions. So it's output here
>>>>>is not very usefull.
>>>>>Look at H. Exner's pawn endgame test, pos12
>>>>>8/6Bp/6p1/2k1p3/4PPP1/1pb4P/8/2K5 b - - bm b2;
>>>>>or on the famous position from Topalov-Shirov
>>>>>8/8/4kpp1/3p1b2/p6P/2B5/6P1/6K1 b - - 0 47
>>>>>
>>>>>Kind regards
>>>>>Bernhard
>>>>
>>>>Crafty did just fine on this position, at least on my computer (2xPII-266).
>>>>
>>>>Dave
>>>
>>>Crafty may behave diferent due to hash settings, extension settings and
>>>unknown things especially an multi proc systems.
>>>Without changing the extensions I get
>>>
>>>               17->   5.20   0.61   1. ... Kg5 2. Ba1 Bf1 3. Bb2 Kf5 4.
>>>                                    Ba1 Kf6 5. Bb2 Bg2 6. Ba1 Ke6 7. Bb2
>>>                                    Kf5 8. Kf2 Bh1 9. Ba1 Bg2
>>>               18    10.71   0.60   1. ... Kg5 2. Ba1 Bf1 3. Bb2 Kf5 4.
>>>                                    Ba1 Kf6 5. Bb2 Bg2 6. Ba1 Ke6 7. Bb2
>>>                                    Ke7 8. Ba1 Kd6 9. Bb2 Bf1 10. Kh2
>>>               18    21.61     ++   1. ... Kf5!!
>>>              time=30:00  cpu=200%  mat=0  n=1199751062  fh=1%  nps=666506
>>>              ext-> checks=234624777 recaps=1254995 pawns=1005159 1rep=28133248
>>>thrt:6425
>>>              predicted=0  nodes=1199751062  evals=8539222
>>>              endgame tablebase-> probes done=0  successful=0
>>>              SMP->  split=1340  stop=113  data=9/64  cpu=60:00  elap=30:00
>>>
>>>I wouldn't call this "doing just fine"
>>>
>>>Kind regards
>>>Bernhard
>>
>>Hello Bernhard,
>>I guess I miss your point.  It takes Crafty almost 22 seconds to play Kf5 which
>>is the best first move in your line.  So sure Crafty would not win this at blitz
>>but at tournament time it would probably win this position from the black side.
>>I doubt if most humans would win this position at blitz either.
>>Jim Walker
>
>My point is not that Crafty didn't find the first move, it did.
>And not that Crafty took too long to find it, it didn't.
>My point is that Crafty gives no usefull output, even in a 30 min run.
>This is bad for analysis purposes, not necessarily for playing games.
>IMHO Crafty blows up, you see 234 million check extensions.
>
>Kind regards
>Bernhard

Well, my computer resolved the fail high in twenty seconds or so.  Odd.

Dave



This page took 0.01 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.