Cry Of The Ktulu

[Graham Banks]

CRY OF THE KTULU

AMD64x2 Dual Core
Deep Shredder 11 GUI
128mb hash each
3-4-5 piece tablebases
Ponder off
Xmas2640b.bkt
40 moves in 38 minutes repeating (adapted for the CCRL)
2 cycles 38 rounds
All engines 32-bit 1CPU


Final Standings

26.5 - CM10th Xperience
25.5 - Ktulu 8.0
24.5 - WildCat 8
24.5 - Crafty 23.0
23.0 - Chessmaster 11
21.5 - Scorpio 2.05b
20.5 - CM10th Magic II
19.5 - Pseudo 0.7c
19.5 - BugChess2 1.6.3
19.0 - Ruffian 2.1.0
18.5 - Petir 4.9999
17.5 - The Baron 2.23
17.5 - Tornado 2.2
16.5 - Zappa 1.1
16.0 - Gandalf 6
16.0 - Chronos 1.9.5
15.5 - Aristarch 4.51
15.0 - Alfil 8.1.1
12.5 - Pepito 1.59
11.0 - Rotor 0.4


The complete tournament pgn (zipped) can be downloaded here:
http://kirill-kryukov.com/chess/discuss ... p?id=14684

[Jan Brouwer]

Thanks for the tournament Graham!
I must say I felt it was an honor for rotor to be included in a tournament with such strong opponents.
And the results seem to confirm this .

[Graham Banks]

Thanks for the tournament Graham!
I must say I felt it was an honor for rotor to be included in a tournament with such strong opponents.
And the results seem to confirm this .

Hi Jan,

it was a tough test for Rotor, but it wasn't disgraced.
How is progress with Rotor at present?

Cheers, Graham.

[Jan Brouwer]

Hi Jan,

it was a tough test for Rotor, but it wasn't disgraced.
How is progress with Rotor at present?

Cheers, Graham.

I have not worked on Rotor since version 0.4, still waiting for inspiration to strike again, which typically takes several months
One problem I have been thinking about is how to measure further improvements in a statistically significant way.
Perhaps I should buy a new quadcore to increase the number of test games.
Or else think of an alternative approach which does not require many test games.

Jan

[Graham Banks]

Hi Jan,

it was a tough test for Rotor, but it wasn't disgraced.
How is progress with Rotor at present?

Cheers, Graham.

I have not worked on Rotor since version 0.4, still waiting for inspiration to strike again, which typically takes several months
One problem I have been thinking about is how to measure further improvements in a statistically significant way.
Perhaps I should buy a new quadcore to increase the number of test games.
Or else think of an alternative approach which does not require many test games.

Jan

Sounds fair enough.

Cheers,
Graham.