Rival Chess

6 views
Skip to first unread message

Djordje Vidanovic

unread,
Nov 4, 1998, 3:00:00 AM11/4/98
to

To Chris Moreton:

Hello, you have a very nice page. I looked around and found one
detail that I'd like to question -- it is quite obvious that the
score agains Phalanx XVIII which is rated about 2400 must somehow be
explained, especially taking into account other results. The problem
with Phalanx is that you have to use /-reuse command in the command
line because otherwise, after a couple of games, Phalanx will start
losing miserably, in a most disgraceful manner. Dusan Dobes, its
author, could not be reached regarding this apparent bug. Therefore
I suggest you reran the match :-) as this is a far cry of what the
match should have been like.

Best regards,
Djordje

The University of Nis Computer Chess Page:
http://www.cent.co.yu/chess


Chris Moreton

unread,
Nov 4, 1998, 3:00:00 AM11/4/98
to
Thanks Djordje,

I re-ran the test using the /-reuse flag and the results on the site
http://dialspace.dial.pipex.com/town/drive/ycy39/howto5.shtml . Rival
managed to win 7 of 30 due to Phalanx's clock running out but lost the other
23 (1 on time). I intend to run tests where a clock increment is involved
as this will take the emphasis of the
engine's time management capabilities.

-Chris

>it is quite obvious that the
>score agains Phalanx XVIII which is rated about 2400 must somehow be
>explained, especially taking into account other results. The problem
>with Phalanx is that you have to use /-reuse command in the command
>line because otherwise, after a couple of games, Phalanx will start
>losing miserably, in a most disgraceful manner.

Reply all
Reply to author
Forward
0 new messages