Author: Robert Hyatt
Date: 07:57:47 06/28/99
Go up one level in this thread
On June 28, 1999 at 09:40:02, blass uri wrote: > >On June 28, 1999 at 09:22:21, Robert Hyatt wrote: > >>On June 28, 1999 at 07:52:15, blass uri wrote: >> >>> >>>On June 28, 1999 at 02:35:31, Ed Schröder wrote: >>> >>> >>>>The main problem of these engine-engine matches on one Pc is that >>>>because the lack of the permanent brain the time control rules of the >>>>program in question is not in harmony anymore which may result in >>>>very bad play. >>>> >>>>Playing without permanent brain requires *another* set of rules for >>>>time control. Programs who play engine-engine matches on one >>>>Pc (a) should realize it is playing such a match and then (b) use a >>>>totally different (and well tuned) time control. >>>> >>>>The big question is if these engines are aware of this. If engine_X >>>>knows and engine_Y does not know then engine_X has a big >>>>advantage (50-100 elo?!) and results are meaningless in the >>>>sense that results can completely different if you run them on >>>>2 Pc's (which remains the only accurate way BTW). >>> >>>I am surprised by the 50-100 elo difference. >>>I think that the ssdf results show that the difference between pentium200 and >>>pentium90 is in this range. >>> >>>I think that not using the permanent brain may cause problem only in the moves >>>that are close to the time control(If I suppose 2 hours/40 moves) and the >>>results are similiar to the case that instead of playing all the game on >>>pentium200 you play the first moves on slightly better hardware and the last 10 >>>moves on pentium 100. >>> >>>I expect a difference of 20-40 elo and not 50-100 rating in this case. >>> >>>Uri >> >> >>This just isn't the way it works. And the reason is that no one tests like >>this. (programmers). Any more than we test with no hash tables (or way too >>small hash tables) or any other 'crippled' mode. >> >>The way to play is to play the way the programs were tested and developed. >>Just because something 'works' doesn't mean that it 'works well'... > >I agree it is not perfect but the way I test the new version of Junior is in >engine-engine games and I tell Amir Ban about the games and what are the >problems of Junior. >I am sure that he learns from these games. > >I have not 2 computers with the same speed. > >I do not think that it is exactly the same as computer-computer games but I do >not think that there is a big difference. > >Uri There is _lots_ to be learned by such games. But in nearly every one I am sent, I find a point where 10 more seconds would have made a difference. And then I wonder "If I had 10 more seconds, would my opponent have found anything else with 10 more seconds?" IE in each game time plays an important role at some point. In other games, positional mis-evaluation is seen and can certainly be fixed. But there is a difference between playing games to find problems and playing games to compare the strength of two programs. THose are _not_ the same thing at all.
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.