Author: Robert Hyatt
Date: 18:17:23 08/29/99
Go up one level in this thread
On August 29, 1999 at 18:45:05, Inmann Werner wrote: >On August 29, 1999 at 15:22:41, Robert Hyatt wrote: > >>On August 29, 1999 at 13:40:00, pete wrote: >> >>>On August 29, 1999 at 13:25:05, Robert Hyatt wrote: >>> >>>>On August 29, 1999 at 13:18:23, pete wrote: >>>> >>>>>1. ( I only post this one here , but it wasn't my idea ) >>>>> >>>>>If the time management suffers that much if either ponder=on or ponder=off why >>>>>not add code that checks if ponder is enabled or not at the start of the game >>>>>and adjust your time management ? I really like this idea . >>>> >>>> >>>>But the question is, adjust it to _what_? Time management is non-trivial, and >>>>most of us do a _lot_ of testing/tuning/tweaking with it before we are happy, >>>>and some of us modify it on a regular basis as things show up. Who wants to >>>>take time to play hundreds or thousands of games with pondering off, just to >>>>get the new time allocation code properly tuned? When we really don't play like >>>>this _ever_. > >I agree... > >>>> >>><snip> >>> >>>I understand that _you_ wouldn't . But is this also true for the programmers of >>>commercial progs who have to face the fact that many users ( even professional >>>testers ) will test just like that and come to misleading conclusions ? >>>For example Chessbase even advertises with results of that kind of matches . If >>>I were a programmer for the Chessbase factory I sure would think about it. >> >> >>Ed doesn't either. And I wouldn't be surprised if everyone else doesn't spend >>a lot of time on ponder=off games either. It is simply 'unnatural' to run a >>program that way... > >yes, for you, cause you let your program play on ICC. For me (in Europe) this >is much to expensive. So I let my computer play during night under Winboard or >Fritz Gui against other Engines to test it (games are better than positions). >For Comp-Human play it might be "unnatural", but give me a person playing >hundreds of matches against my program, or give me a set of computers, so I can >use autoplay.... >For the profis and commercials it may be "unnatural", but for me (and maybe some >others?) it is better than nothing.... > >Werner maybe or maybe not. If you do all your tuning with ponder=off, I'll bet your time utilization with ponder=on will not be very good. And unfortunately, you will probably use ponder=on in the _important_ games, such as at the WMCCC and WCCC. I subscribe the the philosophy "dance with the one that brung ya..." and I spend all my time testing in the mode I am going to play important games in. I have done it the other way in the past (particularly with Cray Blitz) and without proper testing, we were _never_ satisfied with its time utilization and ended up tweaking (and sometimes breaking) it during a tournament. Not with Crafty... > >> and most of us would rather spend time tuning the program >>in the state it will play games, not in some crippled state that a user might >>use to play games. IE do we also tune for (a) tiny transposition tables; (b) >>no opening book; (c) no databases (endgame); (d) modified user parameter >>settings; (e) any other random thing a user might try??? >> >>IE I do my testing in the configuration that plays the best/strongest. Not in >>configurations that someone might use "just because it is there..."
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.