Author: Drexel,Michael
Date: 05:13:10 12/14/02
Go up one level in this thread
On December 14, 2002 at 00:57:30, Mark Young wrote: >On December 14, 2002 at 00:10:39, Drexel,Michael wrote: > >>On December 13, 2002 at 22:33:50, Mark Young wrote: >> >>> >>>What is the point of this line? Is not white lost with best play? Can you win >>>with white if you play at 3 min a move or even at Game 30' on fast computer, not >>>the out of date P 600 that was used? >> >>Yes, white is lost in nearly every variation. >>The point is: Fritz 8 has to play the variation with highest score. He sees no >>danger because mating threats are behind his horizon. > >What value is this...dumbing down a computer program so someone can post a >"Dream game" seems silly. Are we going to be wow'ed by this....I think not. > >If Fritz 8 plays this way with standard time controls this would have some >value. > >This is nothing more then a time blunder by Fritz 8 because of the fast time >control and the very slow computer that was used, not a flaw in the program. > >Can anyone produce a game where Fritz 8 plays this way at 3 mins. a move. > I dont agree. 1. An Athlon XP2000+ running at full speed is not a slow computer. Of course there are faster comps available now for average home users. 2. A top program should be able to win after Bf7+ even with 1+1 min time control. The position is completly lost for white. 3. I produced a game where Fritz 8 plays this way at 5+2 min (Fischer mode) for the game. Im sure that it is possible to find other examples at 3 min a move but probably very time consuming. 4. Dont forget. Even the top-programmers accept blitz games today. There had been a World Blitz Championship taken place this year, won by Shredder. As long as computers play chess without artificial intellligence, it will be in principle possible to beat computers this way. Nowadays it is not easy to exploit this weaknesses due to the horizon-effect. King saftey has much improved during the last years. You have to offer the programs high material advantages. Most of the known examples are constructed. You have to try out hundreds of stupid (for humans) variations with Fritz analysing in background to find a new one.
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.