Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Completely blocked positions, some interesting things

Author: Omid David

Date: 14:45:18 10/04/02

Go up one level in this thread


On October 04, 2002 at 16:35:28, Jeremiah Penery wrote:

>On October 04, 2002 at 08:03:19, Omid David wrote:
>
>>And a third position which is also a draw:
>>
>>[D] 8/8/8/k3p3/1p1pPp2/1PpP1P2/B1P5/1K6 w - - 0 1
>
>No fair, an illegal position! :)

Thank you for posting the results. BTW, the position above isn't legal, it could
arise from the following sequence of moves:

Bf1 -> h3 -> e6 -> a2
And then the b2 pawn moving to b3 and suffocating the bishop on a2 :-)



>
>The algorithm doesn't do very well here.  It sees there is a potential path for
>the king, and doesn't give a drawscore.  I think it's sort of failing at the
>right edge of the board in certain circumstances.
>
>I figured out the problem, and now it does well here:
>
>With:
>
>                1     0.00   0.00   1. Kc1
>                1->   0.00   0.00   1. Kc1
>               ..........................................
>               45     4.33   0.00   1. Kc1 Kb5 2. Bb1 Kc5 3. Kd1 Kd6 4.
>                                    Ke2 Ke6 5. Kf2 Kf6 6. Kg2 Kg5 7. Kh3
>                                    Kh5 8. Ba2 Kg5 9. Bb1
>               45->   4.33   0.00   1. Kc1 Kb5 2. Bb1 Kc5 3. Kd1 Kd6 4.
>                                    Ke2 Ke6 5. Kf2 Kf6 6. Kg2 Kg5 7. Kh3
>                                    Kh5 8. Ba2 Kg5 9. Bb1
>
>  Elapsed Time: 4.42, CPU Time: 4.24, CPU Use: 95%
>  Material Balance: 350, Previous Moves Predicted: 0
>  Nodes: 3952436, Evaluations: 1585606, FH%: 89%, NPS: 894k
>  EXTENSIONS->
>   Threat: 11, Passed Pawn: 10565, Recapture: 6796
>   One-Reply: 3021, Check: 382285
>  PRUNING-> Razors-> White: 2464, Black: 19345
>            Futility-> White: 83766, Black: 247508
>  TABLEBASES-> Probes: 53, Successful: 18
>  HASH TABLES-> Hash Hits: 59% (816165/1365799)
>                Pawn Hash Hits: 99% (2287204/2288627)
>                Hash Table Filled: 24%
>
>
>Without:
>
>                1     0.00   3.40   1. Kc1
>                1->   0.00   3.40   1. Kc1
>               ...........................
>               41     4.43     --   1. Kc1
>               41     6.72   0.00   1. Kc1 Kb6 2. Kd1 Kc6 3. Ke2 Kd6 4.
>                                    Kf2 Ke6 5. Kg2 Kf6 6. Kf1 Kg6 7. Kg2
>                                    Kh6 8. Kh3 Kg5 9. Kh2 Kg6 10. Kh3 Kg5
>               41->   6.72   0.00   1. Kc1 Kb6 2. Kd1 Kc6 3. Ke2 Kd6 4.
>                                    Kf2 Ke6 5. Kg2 Kf6 6. Kf1 Kg6 7. Kg2
>                                    Kh6 8. Kh3 Kg5 9. Kh2 Kg6 10. Kh3 Kg5
>               ...........................
>               45    28.70   0.00   1. Kc1 Kb6 2. Kd1 Kc6 3. Ke2 Kd6 4.
>                                    Kf2 Ke6 5. Kg2 Kf6 6. Kf1 Kg6 7. Kg2
>                                    Kh6 8. Kh3 Kg5 9. Kh2 Kg6 10. Kh3 Kg5
>               45->  28.70   0.00   1. Kc1 Kb6 2. Kd1 Kc6 3. Ke2 Kd6 4.
>                                    Kf2 Ke6 5. Kg2 Kf6 6. Kf1 Kg6 7. Kg2
>                                    Kh6 8. Kh3 Kg5 9. Kh2 Kg6 10. Kh3 Kg5
>
>  Elapsed Time: 28.78, CPU Time: 28.04, CPU Use: 97%
>  Material Balance: 350, Previous Moves Predicted: 0
>  Nodes: 32458587, Evaluations: 5693896, FH%: 97%, NPS: 1127k
>  EXTENSIONS->
>   Threat: 835, Passed Pawn: 500170, Recapture: 8659
>   One-Reply: 70371, Check: 1464672
>  PRUNING-> Razors-> White: 298398, Black: 84185
>            Futility-> White: 4197130, Black: 1149765
>  TABLEBASES-> Probes: 6265, Successful: 1612
>  HASH TABLES-> Hash Hits: 103% (15629675/15116155)
>                Pawn Hash Hits: 99% (13413939/13418667)
>                Hash Table Filled: 86%



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.