Author: Christophe Theron
Date: 23:48:07 01/31/04
Go up one level in this thread
On January 31, 2004 at 05:42:09, Kurt Utzinger wrote:
>On January 31, 2004 at 05:21:06, Mike S. wrote:
>
>>On January 30, 2004 at 17:51:32, Kurt Utzinger wrote:
>>
>>>The position below is from a 90m+30s game between
>>>Shredder8 and ChessTiger15 in which White easily
>>>won with 45.Qf6 and an won pawn ending.
>>
>>Another one, where White even gives a bishop to being exchange queens and
>>transpose into a won pawn endgame (KPP-KPP only). It occured in a game Crafty
>>17.14 - Delfi 4.2, 40/10m P3/700 MHz. IIRC, only 4-piece tbs. were available.
>>
>>[D]8/8/2q1k3/1p2B3/3Q4/PKP5/1P6/8 w - - 0 86
>>
>>86.Qd6+!
>>
>>I don't know if it's the fastest way to mate (I guess so), but also Qd6+ is
>>human-like IMO, because after the very simple
>>
>>86. Qd6+ Qxd6 87. Bxd6 Kxd6 88. a4
>>
>>it's very easy for White. So no need to deal with "endless" counter-checks from
>>a black queen...
>>
>>Analysis by Crafty 19.06, P3/700 MHz:
>>
>>86.Kb4 Qb7 87.Qc5 Kf5 88.Qxb5 Qa8 89.Bd6+ Kg6 90.Qd3+ Kf6 91.Qd4+ Kf7 92.a4
>> +- (7.84) Depth: 1/7 00:00:00
>> +- (9.50) Depth: 10/25 00:00:10 3613kN, tb=9
>>86.Qd6+ Qxd6 87.Bxd6 Kxd6 88.Kb4 Kc6 89.a4 Kb6 90.b3 Kc6 91.a5 Kd6 92.Kxb5 Kd5
>>93.a6
>> +- (10.58) Depth: 10/25 00:00:11 3747kN, tb=13
>> +- (11.44) Depth: 13/32 00:00:22 8183kN, tb=223
>>
>>Regards,
>>Mike Scheidl
>
>
> Hi Mike
> It's quite remarkable that Crafty 19.06 found the most
> easiest and human-like win with Qd6+. Most engines are
> simply unable to do so [exception = Deep Sjeng 1.5 UCI].
> Kurt
This time Chess Tiger X has another opinion. It finds Qd6+ very quickly on my
P3M-933MHz (6Mb HT, no tablebase):
N10 0.93s Qd6+ Qxd6 Bxd6 Kxd6 c4 Kc6 cxb5+ Kxb5 a4+ Ka5 Ka3 +8.12
(solution found in 0.93 second)
I expect the final version to find it much faster actually.
Christophe
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.