Magnus,
I'm getting an odd result with Mogo. I am running games at a fairly
decent level with 6.5 and 8.5 komi to see what will happen.
At 6.5 komi we get this:
* B => 289 = 0.58859
W => 202 = 0.41141*
At 8.5 komi we get this:
* B => 383 = 0.52610
W => 345 = 0.47390
*
BLACK WINS even at 8.5 komi!
When I do this test at lower levels, It is virtually a tie:
* B => 653 = 0.50038
W => 652 = 0.49962*
The low level is 512 play-outs and the high level is 16384 play-outs.
Is it possible that the correct komi should actually be 8.0? I've
always believed it to be 7.0
I did a simple test to get some indication if games are repeating, but
the test doesn't pick up board transformations nor does it know if two
or moves transpose. I'm just using simple command line tools such as
grep, sort, uniq, awk, etc. I am going to build a little utility that
checks for this at some point soon checking for all board
transformations and such. I've needed such a thing for a long time.
It looks like a few games repeat EXACTLY up to the 10th move. NONE
repeat exactly up to the 20th move. But if some repeat exactly up to
the 10th move it's possible that many others repeat by transposition
and/or transformation.
There are at least 3 possibilities:
1. Correct komi really is more than 7.
2. Since mogo seems to play pretty limited choices - it could be that
it's predisposed to playing an early white move that is fairly bad.
One bad white move could tip the balance to black.
3. Along the same lines, Mogo seems to have 1 move opening book even
when no book is specified. It is started out with what is probably the
best black first move at 9x9 and this might be enough to give a
substantial advantage for black.
Here are frequencies that it plays each response to E5 out of 761 games:
1 E5 B4
1 E5 B6
1 E5 D8
1 E5 F8
1 E5 H4
1 E5 H6
2 E5 B5
2 E5 F2
3 E5 E4
3 E5 H5
5 E5 E8
6 E5 E2
16 E5 D4
19 E5 F6
19 E5 G3
21 E5 D6
22 E5 F4
23 E5 C3
25 E5 C7
26 E5 G7
36 E5 G6
41 E5 D3
41 E5 F3
43 E5 C4
46 E5 C5
47 E5 D7
48 E5 G4
50 E5 F7
50 E5 G5
51 E5 C6
51 E5 E7
59 E5 E3
- Don
Magnus Persson wrote:
> Quoting Don Dailey <[EMAIL PROTECTED]>:
>
>
>> My feelings on this seem to match at least one source:
>>
>> Look here: http://senseis.xmp.net/?Komi
>>
>> Here is an excerpt:
>>
>> It is widely believed that the correct komi is independent of board size
>> for all but the smallest boards. For area scoring, this would give 7 for
>> 9x9+, 8 for 8x8, 7 for 7x7, 4 for 6x6, 25 for 5x5 (w cannot live), 0 for
>> 4x4, 9 for 3x3, 4 for 2x2 with a superko rule, and 0 for 1x1. (these
>> need to be verified)
>
> I corrected this sensei page to give komi 9 for 7x7 and added a link
> to the sgf file John Tromp provides with the analysis.
>
> I played a lot with Valkyria on 7x7 and although it proofs nothing it
> is really convincing that 9 is the correct komi.
>
> Similarily one find that in very simple games on 9x9, but where the
> moves are good solid shape white almost always win with 0.5 points
> with 7.5 komi. Thus if one designs and opening book for 9x9 (as I
> tried to) one should try to complicate things as black and play simple
> as white early on. The exception may be when black opens at 5,5 in the
> center. Then often white ends up trying to live with two groups which
> can be very difficult against a competent opponent.
>
> -Magnus
>
> _______________________________________________
> computer-go mailing list
> [email protected]
> http://www.computer-go.org/mailman/listinfo/computer-go/
>
_______________________________________________
computer-go mailing list
[email protected]
http://www.computer-go.org/mailman/listinfo/computer-go/
