Author: Robert Hyatt
Date: 19:28:58 07/14/04
Go up one level in this thread
On July 14, 2004 at 16:35:01, Omid David Tabibi wrote: >On July 14, 2004 at 16:13:20, Russell Reagan wrote: > >>On July 14, 2004 at 12:56:10, Omid David Tabibi wrote: >> >>>>I can easily reproduce all critical moves of the Falcon-Crafty game on my >>>>notebook for black. I already asked you in another message if more time would >>>>have made any difference for Falcon for two of its critical mistakes - you >>>>didn't answer yet. >> >>>Just count the number of times Crafty changed the PV after 1/4 of its thinking >>>time (including pondering), and see how many of them were critical, and >>>prevented a loss. >>> >>>You will find many of them. Here is one I remember clearly: After Falcon's >>>...Ra5 in the endgame, Crafty wanted to repeat the position at first (score: >>>0.00). It thought for a long time, and then changed the PV to other move, which >>>was winning. But you will find many other moves in the earlier stage of the >>>game, which would have resulted in a loss were they played (after thinking for >>>1/4 of the time). >> >>Your proposal was to give everyone quad Opterons, so Crafty would play the same >>moves. So talk about giving Falcon 4x the time, not giving Crafty 1/4th the >>time. >> >>You still didn't answer his question. If Falcon had used 4x the time, would it >>have avoided the mistakes and avoided losing the game? > >I haven't analyzed the whole game in detail (giving 4x thinking time), so I >can't be certain that there was a clear winning move. But having 2x the time >Falcon would have avoided that Ra5 blunder in the endgame and would at least >draw the game. > > >> >>What about the game vs. Junior? Falcon had a good position and could have at >>least got a draw. Would 4x the hardware have avoided, say, f5 and played Nf4 >>instead? > >Just as Falcon played ...f5, there was a massive score drop, meaning that had it >seen one iteration deeper it would have avoided that move. For one iteration >deeper about 2.5-3x speedup is needed, so 4x speedup would have surely avoided >the move. 1. Where is this 4x speedup coming from? An anal orifice? Because unless you are the world's greatest parallel search programmer, you ain't going to get 4x for several years. To pass 3x takes a lot of time, and most commercial programmers are not doing that yet. 2. if you have a "massive score drop" why on earth not fix your program so that it _knows_ to search deeper? Then you don't have to blame losses on hardware, you can avoid them regardless of the hardware. 3. How does a "massive score drop" mean "that one iteration deeper it would have avoided that move"??? This sounds like poor time management rather than anything else. > >The same happened against Diep, with a massive score drop coming just as a move >was played. That is simply a programming bug. Speaking for myself, had I lost _two_ games due to the same problem, I'd be working on a fix as we speak. >The only loss of Falcon which had nothing to do with hardware was >against Deep Sjeng. Falcon simply played a bad variation with horrible >statistics for black, and its desperate exchange sacrifice was no good.
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.