Author: Uri Blass
Date: 01:50:23 03/25/02
Go up one level in this thread
On March 24, 2002 at 22:51:24, Robert Hyatt wrote:
>On March 24, 2002 at 15:21:49, Dieter Buerssner wrote:
>
>>On March 24, 2002 at 14:33:49, Uri Blass wrote:
>>
>>>On March 24, 2002 at 12:54:29, pavel wrote:
>>>
>>>>34... Nc2+ {(Rxg5) -2.65/11 518} 35. Kd1 {-1.34/12 15}
>>>
>>>I do not understand the reason that shredder used only 15 seconds for 35.Kd1
>>>when ponder is off.
>>>There is no obvious best move and 35.Kf1 is also a move to consider.
>>
>>I think, it is a typical time usage of Shredder for "easy moves". I have been
>>surprised by this, too in some games I looked at.
>>
>>When I start Shredder to analyse the 2 best moves in this position, I get:
>>
>>[German piece symbols]
>> 12 3:53 -1.48 35.Kd1 Se3+ 36.fxe3 fxe3 37.Lxe3 dxe3 38.Dxe3 Le8 39.Ta1 Lh5
>>40.Lxb7 (10.091.277) 43.2
>> 12 5:22 -2.94 35.Kf1 Se3+ 36.Lxe3 fxe3 37.f3 c6 38.Le4 Txg5 39.Dc8+ Kg7
>>40.Lxg6 Txg6 (13.963.951) 43.2
>>
>>This is the depth reported in the PGN. So, it looks not unlikely, that this 1.5
>>pawn difference can trigger an "easy move time usage" (although I think, that it
>>will be done in a different manner technically).
>>
>>Regards,
>>Dieter
>
>
>I think that such a definition of "easy move" is absolutely insane. Too many
>positions that look like one move is 1.5 better for the first N iterations,
>only to fail low later. This has happened to me in an ACM event in the past.
>
>"easy" ought to be a recapture that is forced, with no fail lows allowed as it
>might not be best to actually play the recapture at deep depths...
>
>Seems like a bug (or design flaw) to me...
I can add that Kd1 seems to be the worst move based on yace's analysis
(I gave it only one option but it changes it's mind to give scores
for all the possible legal moves) and I susupect that every top program
including shredder will prefer Kf1 if you give it enough time.
New position
[D]7k/1pp1q2p/6b1/p2BprP1/P1Pp1p2/3P3Q/1rnBPP2/2R1K2R w - - 0 1
Analysis by Yace 0.99.56:
1.Kd1 Qa3 2.c5 Nb4 3.Be4 f3 4.exf3 Nxd3 5.Qh6 Rxf3 6.Bxf3 Qb3+ 7.Ke2 Nxc1+ 8.Ke1
-+ (-3.88) Depth: 12 00:14:25 237024kN
1.Rxc2 Rxc2 2.f3 Rb2
-+ (-3.87) Depth: 12 00:17:33 293594kN
1.Rxc2 Rxc2 2.f3 Ra2 3.Rh2 Qxg5 4.Rg2 Qh5 5.Qg4 Qh1+ 6.Rg1 Qxg1+ 7.Qxg1
-+ (-3.76) Depth: 12 00:20:02 336712kN
1.Kf1 Ne3+ 2.Bxe3 fxe3 3.f3 Rxf3+ 4.exf3 Bxd3+ 5.Kg1 Qc5
-+ (-3.75) Depth: 12 00:22:19 375376kN
1.Kf1 Ne3+ 2.Bxe3 fxe3 3.Bf3 exf2 4.Qg4 Rxg5 5.Qc8+ Kg7 6.Rd1 Bf5 7.Qb8 b6
8.Kxf2 Kf6
-+ (-3.28) Depth: 12 00:26:11 439215kN
(Blass, Tel-aviv 25.03.2002)
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.