Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Test Position ... more... addenum...

Author: Dan Ellwein

Date: 11:31:34 04/28/04

Go up one level in this thread


On April 28, 2004 at 13:51:08, John Merlino wrote:

>On April 27, 2004 at 09:28:37, Dan Ellwein wrote:
>
>>On April 27, 2004 at 08:09:24, Dan Ellwein wrote:
>>
>>>On April 27, 2004 at 01:53:35, John Merlino wrote:
>>>
>>>>On April 26, 2004 at 21:48:57, Dan Ellwein wrote:
>>>>
>>>>>On April 25, 2004 at 18:22:46, Johan de Koning wrote:
>>>>>
>>>>>>On April 24, 2004 at 18:21:31, Johan de Koning wrote:
>>>>>>
>>>>>>>On April 24, 2004 at 15:28:13, John Merlino wrote:
>>>>>>>
>>>>>>>>On April 23, 2004 at 20:05:51, Johan de Koning wrote:
>>>>>>>>
>>>>>>>>>On April 23, 2004 at 16:52:42, John Merlino wrote:
>>>>>>>>>
>>>>>>>>>>This position is from The King v Diep in the recent ICT4.
>>>>>>>>>>
>>>>>>>>>>[D]r3k2r/pp1bppbp/2n3p1/8/8/2N1B1P1/PP2PPBP/2RK3R w kq - 0 13
>>>>>>>>>>
>>>>>>>>>>The Chessmaster 9000 database has two games with this position, played by top
>>>>>>>>>>GMs (Andersson v Leko in 1996 and Smirin v Alterman in 1998). In both cases,
>>>>>>>>>>White played f4 and won.
>>>>>>>>>>
>>>>>>>>>>The King, however, chose Kc2 and is soon forced over to a3.
>>>>>>>>>
>>>>>>>>>Actually I cheered when TK switched to Kc2 at the last minute (make that
>>>>>>>>>the last second :-). Both a3 and a1/b1 look better then e1, superficially.
>>>>>>>>>Little did I know what lied ahead though. :-(
>>>>>>>>>
>>>>>>>>>... Johan
>>>>>>>>
>>>>>>>>Did TK switch from f4 or Ke1?
>>>>>>>
>>>>>>>I don't recall seeing f4 in any of the PVs.
>>>>>>>(Will correct tomorrow, if needed.)
>>>>>>
>>>>>>So much for my memory.
>>>>>>It switched from f4 to Kc2.
>>>>>>In 3-best mode it lists Kc3, f4, and h4 within a few cents.
>>>>>>(on slower hardware)
>>>>>>252.33  45M289 03/12 +0.07 Kc2,Bf5+ Kb3,0-0 Rhd1,Rfd8 Rxd8+,Rxd8 Bxc6,bxc6 ...
>>>>>>305.83  54M957 03/12 -0.01 f4,Rc8 Kd2,0-0 Rhd1,Bf5 Ke1,e5 fxe5,Bxe5 Kf2,Rfe8
>>>>>>383.66  68M885 03/12 +0.03 h4,Rd8 Ke1,h5 Kf1,0-0 Kg1,Be6 Kh2,Rfe8 Rhe1,Rd7
>>>>>>
>>>>>>... Johan
>>>>>
>>>>>Hi Johan
>>>>>
>>>>>just for the record CM9_OffSet likes Kc2 in this position...
>>>>>
>>>>>P3/850
>>>>>
>>>>>Time	Depth	Score	Positions	Moves
>>>>>----	----	----	----		----
>>>>>0:00	1/3	0.48	4930		1.Nb5 Rd8 2.Nc7+ Kf8
>>>>>0:00	1/3	0.49	5409		1.f4 O-O 2.Rf1 Rfc8
>>>>>0:00	1/4	0.49	8181		1.f4 O-O 2.Rf1 Rfc8
>>>>>0:00	1/4	0.55	17208		1.Nd5 Rd8 2.b3 Kf8
>>>>>0:00	1/5	0.29	36913		1.Nd5 Rd8 2.Kc2 Bf5+ 3.Kb3 Nd4+
>>>>>					4.Bxd4 Bxd4
>>>>>0:00	1/5	0.53	47262		1.Nb5 Rd8 2.Kc2 O-O 3.Nxa7 Bf5+
>>>>>					4.Kb3
>>>>>0:01	1/6	0.68	90706		1.Nb5 Rd8 2.Ke1 a5 3.f4 Bxb2
>>>>>0:02	1/7	0.29	219503		1.Nb5 O-O 2.Nxa7 Nxa7 3.Bxb7 Rad8
>>>>>					4.Ke1 Bxb2
>>>>>0:03	1/7	0.42	329111		1.Nd5 Rd8 2.Bd2 Kf8 3.Bc3 Nd4 4.Bb4
>>>>>					Nf5
>>>>>0:07	1/7	0.60	632608		1.Kc2 Bf5+ 2.Be4 Bxe4+ 3.Nxe4 f5
>>>>>					4.Nc5 O-O-O 5.Kb1 e5
>>>>>0:10	1/8	0.66	900983		1.Kc2 O-O 2.Kb1 Rfc8 3.Rhd1 Bf5+
>>>>>					4.Be4 Bg4 5.f4 e6
>>>>>0:22	1/8	0.70	1997546		1.Na4 Bf5 2.Ke1 Rc8 3.b3 Bd7 4.Nc5
>>>>>					Bb2 5.Rd1
>>>>>0:30	1/9	0.68	2770742		1.Na4 Bf5 2.Rg1 Rc8 3.Bxc6+ bxc6
>>>>>					4.Ke1 Rc7 5.g4 Be6
>>>>>0:38	1/9	0.82	3570621		1.Kc2 O-O 2.Rhd1 Bf5+ 3.Be4 Bg4
>>>>>					4.Kb1 Rfc8 5.f3 Be6 6.Nd5
>>>>>0:51	1/10	0.78	4873139		1.Kc2 O-O 2.Rhd1 Bf5+ 3.Be4 Bh3
>>>>>					4.Kb1 Rfc8 5.f4 Be6 6.Nb5
>>>>>2:16	1/11	0.87	12968586	1.Kc2 O-O-O 2.Kb1 Kb8 3.Rhd1 Rhe8
>>>>>					4.f4 Nb4 5.Ne4 Be6 6.Rxd8+ Rxd8
>>>>>5:07	1/12	0.95	29377143	1.Kc2 O-O 2.Rhd1 Be6 3.Kb1 Rfc8
>>>>>					4.Nd5 Bf5+ 5.Ka1 Kf8 6.Nf4 Kg8
>>>>>					7.Bxc6 bxc6
>>>>>13:41	2/13	1.04	78101462	1.Kc2 O-O 2.Rhd1 Bg4 3.Kb1 Rfc8
>>>>>					4.Bc5 Kf8 5.f4 Bf5+ 6.Be4 b6 7.Bxf5
>>>>>					gxf5 8.Be3 e6
>>>>>best regards,
>>>>>
>>>>>Dan
>>>>
>>>>An eval of >1.0 shows how very odd this personality is. It might be useful for
>>>>finding sacrifices, but for general analysis I wouldn't trust it at all. The
>>>>final PV doesn't even show Diep's obvious follow-up of 1...Bf5+.
>>>>
>>>>jm
>>>
>>>yes John agree...
>>>
>>>this personality shines best in positions rich in tactics...
>>>
>>>thanks for replying
>>>
>>>Dan
>>
>>okay John
>>
>>just out of curiosity I changed 2 parameters:
>>
>>Attack/Defense=0 (instead of-40)
>>
>>Material/Positional=15 (instead of 0)
>>
>>and here's the output...
>>
>>P3/850
>>
>>Time	Depth	Score	Positions	Moves
>>0:00	1/3	0.43	3670		1.f4 O-O-O 2.Ke1 Be6 3.Bxc6 bxc6
>>					4.Bxa7
>>0:00	1/4	0.45	7321		1.f4 O-O-O 2.Ke1 Kb8
>>0:00	1/5	0.30	21264		1.f4 O-O-O 2.Ke1 Kb8 3.Bc5
>>0:00	1/5	0.31	55634		1.Kc2 Bf5+ 2.Kb3 Be6+ 3.Ka3 O-O
>>					4.Rhd1 Rac8
>>0:01	1/6	0.31	100542		1.Kc2 Bf5+ 2.Kb3 Rc8 3.Ka3 O-O
>>					4.Rhd1 Be6
>>0:01	1/6	0.35	120785		1.f4 Be6 2.Ke1 O-O 3.Kf2 Rac8 4.Rhd1
>>
>>0:03	1/7	0.35	273944		1.f4 Be6 2.Ke1 O-O 3.Kf2 Rac8 4.Rhd1
>>
>>0:07	1/8	0.38	641179		1.f4 O-O 2.Ke1 Rfd8 3.Kf2 Be6 4.Rhd1
>>					Bxc3 5.bxc3 Bxa2
>>0:19	1/9	0.30	1631776		1.f4 O-O 2.Ke1 Rfd8 3.Kf2 Be6 4.Rhd1
>>					Rxd1 5.Rxd1 Rc8
>>0:25	1/9	0.36	2204417		1.Kc2 Bf5+ 2.Ne4 O-O-O 3.Kb1 Bd4
>>					4.Rhf1 Kb8 5.Bf4+ Be5 6.Bxe5+ Nxe5
>>					7.Rfd1 Rxd1 8.Rxd1
>>1:01	1/10	0.36	5498413		1.Kc2 Bf5+ 2.Kb3 O-O 3.Rhd1 h6
>>					4.Nd5 e6 5.Nc7 Rac8 6.Bxc6 bxc6
>>					7.Rxc6
>>2:45	1/11	0.37	15134721	1.Kc2 Bf5+ 2.Kb3 O-O 3.Rhd1 Rfd8
>>					4.Ka3 e6 5.Rxd8+ Rxd8 6.Bxc6 bxc6
>>					7.Bxa7 Bf8+ 8.Kb3 e5
>>5:37	1/12	0.28	31729258	1.Kc2 Bf5+ 2.Kb3 O-O 3.Rhd1 Rfd8
>>					4.Nd5 Bg4 5.f3 Be6 6.f4 Bg4 7.Rc2
>>					Bf5
>>13:12	1/12	0.31	72228683	1.h4 h5 2.Kc2 Bf5+ 3.Kb3 O-O 4.Rhd1
>>					Rac8 5.Ka3 Rfe8 6.b3 Red8 7.Rxd8+
>>					Rxd8
>>
>>here Bf5+ is considered and the pv seems more reasonable...
>>
>>best regards,
>>
>>Dan
>
>This is no surprise to me. Taking away the personality's fairly strong desire to
>attack, and taking away (some of) its highly skewed material value settings
>(where a piece is equal to two pawns, IIRC) would make it play more reasonably.
>
>But that's not the point of the personality in the first place.... :-)
>
>jm

agree... i really do like the original settings when it comes to attacking
possibilities.. :)

i've been playing some games with attack/defense = -15 (instead of -40) and so
far been having pretty good results against the default CM9

need to do a lot more testing though (as time permits) at different time
controls and more games to get a better idea

Dan



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.