Computer Chess Club Archives


Search

Terms

Messages

Subject: Re: Battle of the Crowns -- calibration data [really almost there!]

Author: Dann Corbit

Date: 18:30:52 06/26/00

Go up one level in this thread


On June 26, 2000 at 20:04:39, Adrien Regimbald wrote:

>Hi,
>
>>Today is the last day for an engine update.  The contest will start on Friday.
>
>
>I am wondering if it would even be fair of me to use the updated Faile 1.4 for
>this contest .. I was prepared for a release last night and am working on the
>web page to put the new version on, it is ready for release.
>
>Faile 1.4 is a _LOT_ better than 0.6 .. I'd estimate as much as 200-300 ELO
>better..  do you think it would be fair for me to use it as an "update" ?
>Because I have done extensive play testing with the new version and am very
>happy with the results I've gotten..
>
>Some recent 16 game (10 minute time controls) matches:
>
>Gromit 2   10.5
>Faile 1.4   5.5
>
>Gromt  2   11.0
>Faile 1.4   5.0
>
>Faile 1.4  11.5
>SSEChess    4.5
>
>Faile 1.4  10.5
>SSEChess 2  5.5
>
>Faile 1.4  15.5
>Faile 0.6   0.5
>
>
>I know that these results aren't extensive enough to provide any significantly
>accurate statistical results, but it is quite clear that Faile 1.4 is VERY MUCH
>STRONGER than Faile 0.6 ...
>
>Would it be possible for you to seed Faile 1.4 in with the 2 crown group?  I
>really don't think it would be very appropriate for me to use it as an "update"
>and have it play in the 1 crown group...
>
>Please let me know what you think .. I am currently at the university .. if you
>want the new version, I will send it to you when I get back in an hour or two.

At least a half-dozen other engine writers mailed me a new engine today.
The rules say that today by midnight is the deadline, so your new engine is in.

Simple as that.

However, it might get stuck in the wrong bracket (actually, it is almost surely
the case if you are really 100-200 ELO stronger).

A few test games might bump it to a higher bracket, but when there are a large
number of games already played, the calculation has a lot of "inertia" which
damps sudden changes.



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.