Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Any chance of gandalf 6 being in the top 5 on ssdf? NT.

Author: Uri Blass

Date: 09:34:05 11/21/04

Go up one level in this thread


On November 21, 2004 at 12:01:51, Peter Skinner wrote:

>On November 21, 2004 at 11:34:24, Albert Silver wrote:
>
>>On November 21, 2004 at 11:17:31, Peter Skinner wrote:
>>
>>>On November 21, 2004 at 10:10:27, Albert Silver wrote:
>>>
>>>>My own estimate is about 86 points. It is a very serious estimate, much like
>>>>yours, and is based on no games and no knowledge of the engine, also like your
>>>>estimate. BTW, there's a new player in my club who isn't rated yet, but I
>>>>estimate at about 1900. What's your estimate? Just to see if it is similar.
>>>>
>>>
>>>Well mine _is_ based on data I have seen on Playchess, and "advertising" of
>>>engines.
>>>
>>>I remember Ruffian 2.1.0 being advertised as 100 elo stronger than 1.0.5..
>>>
>>>Gandalf 5.0 being advertised as "Much stronger than 4.32H"..
>>>
>>>Fritz 8 being advertised as 50 elo stronger than Fritz 7....
>>>
>>>I am just a sceptic based on recent advertising, and the results I have seen.
>>>
>>>As for the player in your chess club, by the looks of him, I would guess he is
>>>rated more in the range of 0-1899.. Just off what you figure :)
>>>
>>>Peter
>>
>>I understand your skepticism, and appreciate your frustrating past experience,
>>but it still doesn't put you in a position to speculate on its rating, which is
>>what I objected too.
>>
>>                                            Albert
>
>Well technically I am in a position to speculate on it's rating. That is why it
>is called a speculation.
>
>Instead of just agreeing and saying it _will_ be 100 elo stronger I looked at
>games played online at FICS and Playchess, and saw the engine still has some of
>the problems 4.32H and 5.1 had.

No surprise to read that it has some of the problems of the old versions and it
proves nothing about the question if it is 100 elo stronger or not 100 elo
stronger.

Movei improved by more than 400 since the first version but still has some of
the weakness of the first version like not using tablebases or not evaluating
backward pawns.

Uri



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.