Author: Dave Gomboc
Date: 18:15:38 09/10/01
Go up one level in this thread
On September 10, 2001 at 05:31:18, Uri Blass wrote: >No >If you extend a full ply for every singular move by 0.25 pawn you will never >finish your search. Sorry, Uri, but that's simply not the case. >>>I start with the fact that most of the game moves before Bg5 does not seem to be >>>forced so the depth after Bg5 should be very small. >> >>*c5* Be4 *Ra6* Rb1 *f5* *Bc2* Rb7 *Bd8* *g6* >> >>The * moves are singular. I only see THREE that are not in >>this sequence, and in all cases Hiarcs disagreed with the move >>choices so they may very well not be considered singular by >>Hiarcs because it does not see that they are singularly better >>but DT obviously did. > >I do not believe that they are singular > >Ra6 is only slightly better than Rb8 >g6 is only slightly better than Nf7 They might be only slightly better at shallow depths, but as soon as a difference is found, the better move will be considered singular. Most implementations mark "singular" in the hash table and continue to extend it even if subsequently the gap in score narrows again, to prevent search instability. >Note that 29.Rb1 is probably a positional error based on Deep Fritz's opinion >and black gets more than 0.5 pawn advantage after that move. > >The point of starting with Bg5 was to convince people that deep thought could >not see +2 advantage but it seems that you believe that deep thought was god and >no number of moves is going to convince you that deep thought could not see it. It is quite reasonable for DT2 to reach the positions we are examining in a search during a tournament game. If you can believe that Junior's search tree can be three times longer in some lines than in others, why can't you believe that about DT2? >In that case it seems that the best way to convince you should be to start from >the game from the first move after c5. > >I was not convinced that white is losing after 32.Bg5 but I feel more sure that >I can draw a correspondence game against you and your programs from the root >position and not from the position after 32.Bg5. > >I do not like to spend a lot of computer time on this correspondence game >without seeing a clear evidence that white is losing so I prefer to see first >that you can beat Deep Fritz or even Crafty from the initial position and only >after you do it and post your game I may try to improve the line. > >Uri The starting position of chess is quite irrelevant. We're trying to determine if 27...c5 deserves a score like DT2 gave it; your position is no. If you are no longer interested in contesting that, you should strongly consider conceding the point. Dave
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.