Author: Vine Smith
Date: 00:06:57 05/24/01
Go up one level in this thread
On May 23, 2001 at 14:00:46, John Merlino wrote: >On May 22, 2001 at 15:52:15, Dann Corbit wrote: > >>I think for sure, the most fascinating game was the Chessmaster/Quark battle in >>round 1. (I imagine John Merlino wishes some other game were more compelling) >>;-) >> > >No, I merely wish that the result was different. ;-) > >Perhaps this is a good time to mention that Johan did NOT use the default >settings this time around. He DOES follow CCC and was intrigued (as I was) by >the fact that the King Safety values of most of the "superior" personalities >that are posted here are much higher than the default value of 100 -- usually >somewhere between 130 and 160. > >So, Johan decided to go with an ASYMMETRIC KS value of 100,140, meaning it is >concerned about its OWN King Safety with a value of 100 (the default) but it is >concerned about its OPPONENT'S King Safety with a value of 140. Here's Johan's >explanation of why he chose this configuration: > >----------- >Though I was tempted to use Kurt Utzinger's, I later rejected this jump in the >dark. But then you mentioned the KS pattern on CCC, and I came up with the >following rationale (no proof!). The merit of high KS is that the computer >opponent will be more likely to face lethal mating threads. On the other side, >it tends to "chickenize" TK himself, which might cramp its style in wild >positions. Against the average program TK should not need to be defensive on the >King's side, hence the asymmetrical weights. >----------- > >This could very well explain moves like 25...Re5? that Vine Smith was >questioning -- a failed attempt at a kingside attack. > >As a side note, he also used Selective Search=10, but that's ONLY because he got >an Athlon 1.3GHz machine just before the tournament. If he had used his trusty >550MHz instead, then he would have stayed with SS=6. > >jm It does look like the increased tropism and heightened assymmetry may have negatively impacted The King's play in this game. Chessmaster 8000 with default settings does not want to play this move. At 13 ply (which took 37:50 to reach on my PIII 850), it chooses the far more sensible 25...Re7, also its choice at the end of the previous two ply. 25...Re5 did not show up as a selection in any of the previous plies I could see (starting around 7). It might be useful to keep this position on file to test any new "personalities" for which superiority is claimed; those that play 25...Re5 are probably overaggressive. A question about analyzing with the King -- apparently, analysis can only be done by the default Chessmaster personality, so I can get only the default 1 MB of hash? Can I change this somehow? If I adjust the "ttable" variable in cm.ini, it just immediately gets set back to 0 by the program. Regards, Vine Smith
This page took 0.01 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.