Author: Ralf Elvsén
Date: 06:06:00 07/18/00
Go up one level in this thread
On July 18, 2000 at 07:33:16, Mogens Larsen wrote: >On July 18, 2000 at 06:40:38, Ralf Elvsén wrote: > >>Let me make one thing clear: if a programmer explicitely says his program >>has a good ponder = OFF management I have few problems with testing the >>program in this mode (see below). > >I think it's safe to say that almost all programs have good time management with >ponder off, otherwise it would be virtually impossible to test the program >without financial support. Do you really think that the authors of the programs >on the Young Talents CD started testing their programs on a dual or quad >machine? The tweaking of the various chess program parameters did most likely >occur on one machine using ponder off, unless the author is a skilled chess >player or know someone there is. There are two kinds of program parameters: those who control time management and those who control search and evaluation. I doubt strongly the latter depends on whether the program plays with ponder ON or OFF. But I will be happy to be corrected if this is not the case. But perhaps that is not what you meant? If you mean that the parameters for ponder = OFF are good because these programmers paid attention to this, maybe this is true. I really don't know. I would prefer to hear it from them explicitely. > >>I prefer to stare at Junior at Dortmund. > >I don't think that will last very long. No, but the information is of very high quality. > >>How it compares to other programs I don't know and >>this is exactly my point. If I want to estimate performance under optimal >>conditions (ponder = ON) with results from ponder = OFF this introduces an >>uncontrolled parameter. > >Well, noone is suggesting doing that. Ponder off is adequate for estimating the >comparative strength between programs, not the actual strength of programs with >ponder on. That would be a ludicrous attempt. If "the actual strength of programs with ponder on" is one thing, what is then "the comparative strength" ? Ah, I think I understand from one of your paragraphs below. > >>So do I. At least it isn't any worse than e.g. Hiarcs' . But I only >>look at these games because I have convinced myself of this. I wouldn't >>take an unknown program, run 100 ponder = OFF games overnight and then >>think the result reflected its playing strength with ponder = ON. > >Again, noone is doing that. Some are, though not you and I never meant to imply that. > >>All I need to know is that the programmer wants to compete in the >>ponder = ON business. Then it is just another aspect of the program, >>like king safety or endgame knowledge. Its quality is affecting the >>performance of the program, together with every other piece of knowledge >>it has, and is reflected in the results. And this is what I care about. > >If you want quality then ponder on is the best option, but quality is >unimportant in engine-engine comparison. > >>You have a point when you're saying (if I understand you correctly) that >>the relative quality of ponder ON/OFF management for two programs is >>difficult to know and poses a problem. Even if both programmers says >>both modes are competitive, one program can be relatively weak in >>ponder = ON and strong in ponder = OFF. Since I think the ponder = ON >>mode is the only interesting game, this is another argument for me to >>not take ponder = OFF games so seriously without careful observations. > >I think we're interested in different things. You're interested in the quality >of the games played. Fair enough, but I'm interested in the results and since >there's no evidence to suggest that ponder OFF skews the relative strength >between programs I prefer that option. Ed Schröder run a test before which he thought (if I recall everything correctly) supported the theory that it made a difference. Objections were made by others, and of course noone really knows for sure as usual. My point is that one should try to make an estimate of the effect for the program at hand, both through observations and (if possible) with information from the programmer. Then one may draw conclusions from these games. >If I used ponder ON on one machine then >critiscism would be aimed at cpu management, so it's a no win situation. >However, if I want to be beaten to a pulp by Crafty I do use ponder ON :o). > >Best wishes... >Mogens I doubt anyone else but us is very interested in this. This must be the gazillionth thread on this topic. If you don't mind I think I'll stop here, unless you manage to provoke me ;) Ralf
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.