Author: Christophe Theron
Date: 23:24:50 11/09/01
Go up one level in this thread
On November 09, 2001 at 23:41:38, Thomas Mayer wrote:
>Hi Christophe,
>
>>>We have, because included in those lists are some programs that also have a
>>>rating at the SSDF (e.g. Little Goliath, SOS, Crafty) - so at least the lists
>>>are compareable.
>
>>No they are not unless in the rating lists you are talking about Crafty and the
>>other known programs have exactly the same rating as in the SSDF list.
>>I do not know what lists you are talking about, but the rating of the chess
>>servers for example are totally useless here.
>
>okay, time to learn something about the winboard world... :) Thats no
>chess-server-List, it's just a rating list gained out of hundreds of games.
>Crafty, SOS & Little Goliath are listed there with nearly the same ELO then in
>SSDF-list, at least IN the margin of error shown in the SSDF-List...
>There are many such lists around, all have Crafty around 2500... (Some use fixed
>2500 for Crafty just to have a good comparison)
>So this ratings are nearly as good as the SSDF-ratings. In some case even
>better, because there is also testing with the older programs, all programs has
>EQUAL hardware, there is much more interference between the upper and the lower
>class, each engine does not only play against 10 or so opponents.
>So there could be no discussion wether we have a rating or not for Grizzly. When
>you want to say now to me, that you believe that Grizzly is over 2200, I must
>laugh a little bit... Michael, the author, would be proud that Théron has said
>it, but would also laugh with me...
I do not know what's wrong exactly, but something's wrong.
Either it is the rating of Grizzly, or the low number of games played.
Chess Tiger for Palm is definitely over 1800, unless you are using the demo
mode.
>>>> The valid way to evaluate the strength of Chess Tiger for Palm is to let it
>>>> play against computers that have a SSDF rating.
>
>>> if that is wanted I will try to get one and let it play against Fritz 7,
>>> Junior 7, ChessTiger 14, Rebel Century 3, Shredder 5, Gandalf 4.32 etc... Do
>>> you think it will earn 2100 ??? In fact I am quite sure that it HAS around
>>> 2100 ELOs, but in comp-comp match depth is more important then against
>>> humans and those differences get to big on the list.
>
>>If you match it against too strong opponents, you'll get a mathematically high
>>margin of error. So that will not work.
>
>well, so tell me an engine that I can take because it has around 2100 on Athlon
>1333 on SSDF. Please, understand me, it is not ment against ChessTiger or
>anything, but the ratings on SSDF in comparisson to humans does not make any
>sense to me at the moment.
>Let's take another one:
>Lets look on ChessTiger 11.8, a fantastic engine those days - listed with 2382
>ELOs on SSDF list... Now let it play on it's P90 against the tops on the list,
>say everything what is listed with an Athlon 1200... Do you think ChessTiger
>11.8 would achieve 2382 ??? I doubt that. It would get much less... But: Is that
>fair ???
It is not a question of fairness. You will get almost meaningless results from a
mathematical point of view. In other terms you will get a result with a huge
error margin. It will tell you close to nothing.
That's why it's better to match a computer/program/player against players of its
class. You get results with a better accuracy.
> I don't think so, I don't see any reason why ChessTiger 11.8 gets
>weaker the last years ? There is nothing changed at all, so how should it get
>weaker on that hardware ??? So what else can we do ? Raise the rating of the
>tops ? Then they would be over 2800, maybe over 2900... But who thinks that they
>have over 2900 ?
>When the SSDF would have another base value I would not have any problem with
>that, say on this list ChessTiger 11.8 might have 5382 and the tops are around
>5900... that would be no problem, because the comparisson is still possible and
>with Arpad Elos formula we could still calculate how many percent ChessTiger
>11.8 on P90 against ChessTiger 14 on Athlon 1200 would make.
>And we would have ChessTiger on Palm around 5100... everything fine...
>
>> You need to let it play against computers/programs listed in the SSDF which
>> have a rating close to the estimated rating of Chess Tiger for Palm.
>
>But those computers/programs are not in ANY comparisson to the strong programs
>because they play like incest last some years under the top programs, no
>interference with the older progs on older hardware.
>
>Hope you get the point now what I want to say...
I generally understand what you say, but I don't agree to everything.
Christophe
This page took 0.01 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.