Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Revised ECM Test Suite (some additional analysis needed)

Author: Uri Blass

Date: 17:38:21 07/19/04

Go up one level in this thread


On July 19, 2004 at 20:34:16, Uri Blass wrote:

>On July 19, 2004 at 20:30:15, Uri Blass wrote:
>
>>On July 19, 2004 at 20:15:19, Uri Blass wrote:
>>
>>>On July 19, 2004 at 09:47:48, Dan Andersson wrote:
>>>
>>>>[D]2bq1b2/1p3k2/1r1p1p1p/pNpP1PpP/P1P3P1/6B1/1P5Q/2K1R3 w - -
>>>> This isn't a complete solution tree. But it contains almost all white manouvers
>>>>that make Re6 a kinch move.
>>>>1.Re6 Bxe6 2.fxe6+ Kg8 3.Qc2 Qe8 4.Be1 Qe7 5.Bc3 Bg7 6.Qg6 Rxb5 7.cxb5 b6 8.
>>>>Bxa5
>>>>
>>>>MvH Dan Andersson
>>>
>>>The problem is that I suspect that there are many alternative winning moves.
>>>White can play for example 1.Re3 and later Re6
>>>
>>>Here is my effort with yace to give winning tree that was not succesful.
>>>
>>>I could prove more than +2 after Kg8
>>>After Ke8 Qc2 the score was only +0.97(maybe I did not give yace to search deep
>>>enough) and when I got back to the root position it cannot find 1.Re6 and
>>>prefers 1.Re3 or 1.b3 with evaluation of +1.0x for white.
>>>
>>>There are lines when white gets a won endgame when programs need a deep search
>>>to find the win.
>>>
>>>
>>>
>>>
>>>[Event "?"]
>>>[Site "?"]
>>>[Date "????.??.??"]
>>>[Round "?"]
>>>[White "New game"]
>>>[Black "?"]
>>>[Result "*"]
>>>[SetUp "1"]
>>>[FEN "2bq1b2/1p3k2/1r1p1p1p/pNpP1PpP/P1P3P1/6B1/1P5Q/2K1R3 w - - 0 1"]
>>>[PlyCount "23"]
>>>
>>>1. Re6 Bxe6 2. fxe6+ Kg8 (2... Ke8 3. Qc2) 3. Qc2 Qe7 4. Be1 Qg7 (4... Ra6 5.
>>>Qg6+ Qg7 (5... Bg7 6. Nc3 Ra8 (6... f5) 7. Ne4 b6 8. Bc3) 6. Qe8 Qe7 7. Qb8) (
>>>4... Qh7 5. Bxa5) 5. Nc7 Rxb2 (5... Qxc7 6. Bxa5 Rxb2 (6... Qg7 7. Bxb6) 7.
>>>Bxc7 Rxc2+ 8. Kxc2 Be7 9. Kd3 Kf8 10. Ke4 Kg8 11. Kf5 Kg7 12. Bb6) (5... f5 6.
>>>gxf5) 6. Qxb2 Qxc7 7. Qxf6 b6 (7... Qe7 8. Qxe7 Bxe7 9. Bc3) (7... Qg7 8. Qxg7+
>>>Bxg7 9. Bxa5) (7... Qh7 8. Bc3 Be7 9. Qg6+ Qxg6 10. hxg6) 8. Bc3 Qh7 9. Qg6+
>>>Qxg6 10. hxg6 Be7 11. Kc2 Bd8 12. Kd3 *
>>>
>>>
>>>
>>>Uri
>>
>>Maybe Qc2 is not the right move after Ke8 and Be1 is better and yace can see
>>+1.1 for white some minutes after Fritz's suggestion (Be1).
>>
>>I also plan to analyze all the position of the original suite with movei at 40
>>seconds per move and report results tommorow.
>>
>>I will try to compare movei at 40 seconds per move(Athlon64-3000) with the
>>commercial at 10 seconds per move(A3200)
>>After all movei is still significantly weaker than them so I am not sure if
>>movei can perform better.
>>
>>
>>New game
>>[D]3qkb2/1p6/1r1pPp1p/pNpP2pP/P1P3P1/8/1P5Q/2K1B3 b - - 0 1
>>
>>Analysis by Yace 0.99.87:
>>
>>3...Rxb5 4.axb5
>>  ±  (1.23)   Depth: 1   00:00:00
>>3...Qb8 4.Bxa5
>>  ²  (0.52)   Depth: 1   00:00:00
>>3...Qa8
>>  ³  (-0.28)   Depth: 1   00:00:00
>>3...Qa8
>>  ³  (-0.28)   Depth: 1   00:00:00
>>3...Qa8 4.Nc7+ Kd8 5.Nxa8
>>  +-  (7.85)   Depth: 2   00:00:00
>>3...Qa8 4.Nc7+ Ke7 5.Nxa8
>>  +-  (9.16)   Depth: 2   00:00:00
>>3...Qa8 4.Nc7+ Ke7
>>  +-  (9.16)   Depth: 2   00:00:00
>>3...Qb8 4.e7
>>  +-  (5.41)   Depth: 2   00:00:00
>>3...Qb8 4.Bxa5
>>  ²  (0.52)   Depth: 2   00:00:00
>>3...Qe7 4.Qe2
>>  ²  (0.31)   Depth: 2   00:00:00
>>3...Qe7 4.Bxa5
>>  =  (0.14)   Depth: 2   00:00:00
>>3...Ra6 4.Qe5
>>  =  (-0.01)   Depth: 2   00:00:00
>>3...Ra6 4.Kd1
>>  =  (-0.05)   Depth: 2   00:00:00
>>3...Ra6 4.Kd1
>>  =  (-0.05)   Depth: 2   00:00:00
>>3...Ra6 4.Kd1 Qe7
>>  ³  (-0.43)   Depth: 3   00:00:00
>>3...Ra6 4.Kd1 Qe7
>>  ³  (-0.43)   Depth: 3   00:00:00
>>3...Ra6 4.Qf2 Qe7 5.Qf5
>>  ³  (-0.30)   Depth: 4   00:00:00
>>3...Ra6 4.Qf2 Qe7 5.Qf5
>>  ³  (-0.30)   Depth: 4   00:00:00
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Kd1
>>  ³  (-0.32)   Depth: 5   00:00:00
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Kd1
>>  ³  (-0.32)   Depth: 5   00:00:00
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Rb6
>>  =  (-0.25)   Depth: 6/16   00:00:00  32kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Rb6
>>  =  (-0.25)   Depth: 6/20   00:00:01  35kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Rb6 7.Nxd6 Rxd6 8.Bxd6 Qxd6 9.Qxf6+ Qe7
>>  ³  (-0.30)   Depth: 7/20   00:00:01  72kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Rb6 7.Nxd6 Rxd6 8.Bxd6 Qxd6 9.Qxf6+ Qe7
>>  ³  (-0.30)   Depth: 7/23   00:00:01  78kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Kc8 7.Kd1 Kd8 8.Nxd6 Rxd6
>>  =  (-0.17)   Depth: 8/23   00:00:01  185kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Kc8 7.Kd1 Kd8 8.Nxd6 Rxd6
>>  =  (-0.17)   Depth: 8/23   00:00:01  202kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Kc8 7.Kd1 Rb6 8.Qf7
>>  =  (-0.05)   Depth: 9/23   00:00:01  426kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Kc8 7.Kd1 Rb6 8.Qf7
>>  =  (-0.05)   Depth: 9/26   00:00:01  455kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Kc8 7.Kd1 Rb6 8.Nxd6+ Rxd6 9.Bxd6 Qxd6
>>10.Qxf6
>>  =  (-0.11)   Depth: 10/26   00:00:01  1075kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Kc8 7.Kd1 Rb6 8.Nxd6+ Rxd6 9.Bxd6 Qxd6
>>10.Qxf6
>>  =  (-0.11)   Depth: 10/29   00:00:02  1142kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Kc8 7.Kd1 Rb6 8.Be1 Ra6 9.Bg3 Rb6 10.Be1 Ra6
>>11.Bg3
>>  =  (0.00)   Depth: 11/32   00:00:02  2619kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Kc8 7.Kd1 Rb6 8.Be1 Ra6 9.Bg3 Rb6 10.Be1 Ra6
>>11.Bg3
>>  =  (0.00)   Depth: 11/32   00:00:02  2830kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Rb6 7.Kd1 Kc8 8.Be1 Ra6 9.Bg3 Rb6 10.Be1 Ra6
>>11.Bg3 Rb6
>>  =  (0.00)   Depth: 12/36   00:00:04  6484kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bg3 Rb6 7.Kd1 Kc8 8.Be1 Ra6 9.Bg3 Rb6 10.Be1 Ra6
>>11.Bg3 Rb6
>>  =  (0.00)   Depth: 12/36   00:00:05  6940kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bc3 Bg7 7.Qh7 Qf8 8.Kd1 Ke7 9.Qf5 Qe8 10.b3
>>  =  (0.07)   Depth: 13/39   00:00:15  20698kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Bc3 Bg7 7.Qh7 Qf8 8.Kd1 Ke7 9.Qf5 Qe8 10.b3
>>  =  (0.07)   Depth: 13/39   00:00:16  21915kN
>>3...Ra6 4.Nc3 Qe7 5.Ne4 Kd8 6.Qf2 f5 7.gxf5 Bg7 8.f6 Bxf6 9.Nxf6 Qxf6 10.Qxf6+
>>Kc8
>>  ²  (0.47)   Depth: 14/45   00:00:44  54827kN
>>3...Ra6 4.Nc3
>>  ²  (0.47)   Depth: 14/45   00:00:48  60752kN
>>3...Ra6 4.Nc3
>>  ²  (0.47)   Depth: 14/45   00:00:53  65577kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Nc3 Qg7 7.Ne4 f5 8.Qxf5 Qe7 9.Qf6 Qxf6 10.Nxf6
>>Bg7 11.Ng8 Rb6 12.Bxa5
>>  ±  (0.76)   Depth: 15/51   00:02:44  184620kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Nc3 Qg7 7.Ne4 f5 8.Qxf5 Qe7 9.Qf6 Qxf6 10.Nxf6
>>Bg7 11.Ng8 Rb6 12.Bxa5
>>  ±  (0.76)   Depth: 15/51   00:02:52  193207kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Nc3 Qe8 7.Qxe8+ Kxe8 8.Ne4 Be7 9.Ng3 Bf8 10.Nf5
>>Rb6 11.Kc2 Ra6 12.Bc3 Be7 13.Nxh6
>>  ±  (1.10)   Depth: 16/57   00:05:46  383179kN
>>3...Ra6 4.Qc2 Qe7 5.Qg6+ Kd8 6.Nc3 Qe8 7.Qxe8+ Kxe8 8.Ne4 Be7 9.Ng3 Bf8 10.Nf5
>>Rb6 11.Kc2 Ra6 12.Bc3 Be7 13.Nxh6
>>  ±  (1.10)   Depth: 16/57   00:07:07  469071kN
>>
>>(, MyTown 20.07.2004)
>>
>>Uri
>
>It seems to be only tranposition of moves and Be1 Ra6 Qc2 Qe7 is the same as Qc2
>Qe7(forced) Be1 Ra6
>
>Uri

It is tranposition based on Yace's plan but not based on Fritz's plan
Fritz suggests Be1 Ra6 Nc3

Uri



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.