I think as a test, I would go with the full handicap system, I would
just refuse to match players that need more than 9 stones. I can
always cut it back to 4 or 6 later.
Why don't we view it as an experiment to gather a lot of statistics.
I can change back to ELO later.
The only question, and will need peoples advice on this, how to set
this
up? Do you require 3 consecutive wins in a row to advance?
I'll throw this idea out:
1. If you win increase opponent kyu by 0.1, lower yours by 0.1
2. Pair more or less randomly. Round the difference in ranks of
your opponent to the nearest kyu and handicap accordingly.
3. We could also scale the rank win/loss a little by the fractional
difference in rank. For instance if you are exactly 1 kyu
different,
you get the 1 stone handicap (black stones) but if you are only .6
kyu stronger, and have to take the handicap, you scale the rank
win/loss
accordingly, in linear fashion.
Is this reasonable?
- Don
On Fri, 2006-12-22 at 12:17 -0500, House, Jason J. wrote:
>
> >I personally think small handicaps in 19x19 might be reasonable
> because
> >I think playing good moves is still a dominant factor - at least at
> the
> >levels our programs can handle. I would be reluctant to go beyond a
> >few stones. I don't know what a good number is, but I'll take a
> >somewhat
> >educated guess and say 4 stones, because it supposedly corresponds to
> >almost
> >400 ELO points - which I have come to think of as a conceptual
> >"barrier" of superiority. If you are 400 ELO superior your losses
> >will be rare and in the linear version of the ELO formula 400 is
> >considered
> >certain victory (you can't win ELO points from beating someone 400
> >points
> >weaker or more. The linear formula therefore cuts off at 350, so
> that
> >you
> >still get a little bit for beating a weak player.)
>
> Just as an FYI, KGS refuses to do rated games with more than 6 stones.
> I feel like as I get better at playing go, handicap stones mean more.
> As a 4k, playing a 10k with 6 stones is rather daunting. When I was a
> 10k playing a 16k with 6 stones, I didn't worry too much.
>
> It may be reasonable to only rate games with up to 4 stones like you
> suggest, but maybe have higher handicap games occur occasionally just
> so people can see the conceptual superior barrier at work ;)
_______________________________________________
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
