Cyclone 2.0 rating after 200 games

[Jouni]

I played 200 short games against different engines (single CPU) and compared rating to Toga II 1.4 Beta5c. Result +10 points = nothing special so far.

Jouni

[kranium]

I played 200 short games against different engines (single CPU) and compared rating to Toga II 1.4 Beta5c. Result +10 points = nothing special so far.

Jouni

Hey 10 points is 10 points! I'll take it! <:
Milking a few ELO out of these programs is sometimes like getting water from a stone! <"

Would you be willing to try a head to head, i.e. cyclone vs 14beta5c?
i think the difference wil be more like +30-50...(that's what i normally see in my test games).).

Norm

[Jouni]

I have done it already:

Code: Select all

Noomen  0

                      
1   Toga II 1.4        0½111010010½½½01½½½½½½½½00½½½½010½0½½0½1½111½½10½1  25.0/50  625.00
2   Cyclone 2.0 1cpu  1½000101101½½½10½½½½½½½½11½½½½101½1½½1½0½000½½01½0   25.0/50  625.00

Jouni

[kranium]

I have done it already:

Code: Select all

Noomen  0

                      
1   Toga II 1.4        0½111010010½½½01½½½½½½½½00½½½½010½0½½0½1½111½½10½1  25.0/50  625.00
2   Cyclone 2.0 1cpu  1½000101101½½½10½½½½½½½½11½½½½101½1½½1½0½000½½01½0   25.0/50  625.00

Jouni

Thanks for the games Jouni,

CCRL has:

Cyclone 2.1 2924 +45 −45 44.1% +36.5 48.3% 145
Cyclone 1.0 2878 +32 −32 50.3% −2.4 40.4% 302

that's a 46 point increase,
so I'm pretty happy with the results so far...(need a lot more games!)

Norm

[Jouni]

Against Fritz10 stunning score!

Code: Select all

                      
1   Cyclone 2.0 1cpu   10½01½111½½½10½111½1111111111½0½½11100½½1011101½½½   34.5/50
2   Fritz 10           01½10½000½½½01½000½0000000000½1½½00011½½0100010½½½   15.5/50

But about 50% when playing Glaurung.

Jouni

[kranium]

thanks for the tests,
but as you likely know, one needs a lot more games than 50 to draw a conclusion...

[Spock]

CCRL has:

Cyclone 2.1 2924 +45 −45 44.1% +36.5 48.3% 145
Cyclone 1.0 2878 +32 −32 50.3% −2.4 40.4% 302

that's a 46 point increase,
so I'm pretty happy with the results so far...(need a lot more games!)

Norm

Current position:

Code: Select all

Cyclone 2.1                          2927  +34  -35  47.2%  +17.5   249
Cyclone 1.0                          2879  +32  -32  50.5%   -3.6   305

Still a lot more games required as you say, but the bigger problem is Cyclone 1.0 doesn't have enough games to provide a solid base for comparison

[kranium]

CCRL has:

Cyclone 2.1 2924 +45 −45 44.1% +36.5 48.3% 145
Cyclone 1.0 2878 +32 −32 50.3% −2.4 40.4% 302

that's a 46 point increase,
so I'm pretty happy with the results so far...(need a lot more games!)

Norm

Current position:

Code: Select all

Cyclone 2.1                          2927  +34  -35  47.2%  +17.5   249
Cyclone 1.0                          2879  +32  -32  50.5%   -3.6   305

Still a lot more games required as you say, but the bigger problem is Cyclone 1.0 doesn't have enough games to provide a solid base for comparison

Thanks Ray-
(am I correct in assuming you're Ray Banks, CCRL tester, and these games are official?)

another 104 games...
(well, i'm happy to see the the rating went up a little instead of falling!).

Norm

[Graham Banks]

Thanks Ray-
(am I correct in assuming you're Ray Banks, CCRL tester, and these games are official?)

Norm

Yes to both.

Brother Graham.

[kranium]

Thanks Ray-
(am I correct in assuming you're Ray Banks, CCRL tester, and these games are official?)

Norm

Yes to both.

Brother Graham.

Ah yes -
the identical last names should have been a clue, but it never dawned on me you might be related!