[Computer-go] Operators for Frisbee Go Simulation

Gonçalo Mendes Ferreira gonmf at sapo.pt
Thu Apr 14 10:00:33 PDT 2016


>> In testing I've also noted that a komi of 7.0 or 7.5 is no longer
>> reasonable in Frisbee Go...
>
> Tell us more.
> In particular "correct" komi might depend on the parameter p.

I got these win rates at the first turn of matches, in 9x9 with half a
million playouts. These values tend to be slightly optimistic.

     accuracy p
komi  0.5  0.2
7.5   31%  22%
3.5   43%  36%
1.5   48%  45%
1.0   49%  47%
0.5   51%  49%
0.0   52%  51%

I hope the formatting turns out okay. This is still not taking into
account John's input on move legality.

Gonçalo

On 14/04/2016 17:32, "Ingo Althöfer" wrote:
> Hi Goncalo,
> 
> 
>> I also had another restriction in my program that I forgot to mention:
>> 6. Both players must be using the same probability p.
> 
> this will be the case in Leiden.
> 
> 
>> In testing I've also noted that a komi of 7.0 or 7.5 is no longer
>> reasonable in Frisbee Go...
> 
> Tell us more. 
> In particular "correct" komi might depend on the parameter p.
> And a question to all who intend to participate:
>  Should we have integral komi or dot-5-style?
> 
> Or should we let the bots do komi bidding in the tournament?
> 
> 
> Cheers,
> Ingo.
> _______________________________________________
> Computer-go mailing list
> Computer-go at computer-go.org
> http://computer-go.org/mailman/listinfo/computer-go
> 



More information about the Computer-go mailing list