[computer-go] So many MoGo run on cgos 9x9

[Olivier Teytaud]

> There is no restriction on how many mogo bots can run.  However, there
> is not much of a point if everyone is just running the same bot unless
> they are running at different levels and we can see exactly how they are
> set up.

We have launched 4 mogos, and I explain here what is tested:
- mogo1thread and mogo4threads are explicit :-)
- goala1 and goala4 are the same mogos as those above, except that some
   memory management trick detailed below.

There are therefore two tests:

*multithreading

As far as I know, the influence of the computational power has
often been emphasized in UCT, and in particular that the
improvement with multi-threaded monte-carlo planning is almost
linear (at least for small numbers of processors) in the sense
that
    4 threads ==> as strong as the 1 thread with 4x more time.

We want to see to which extent this is still true after the many
improvements which have been made in e.g. mogo (amaf, etc). We
can also compare to Moongo, which is a mogo running on a smaller
hardware but is seemingly not much weaker.

*memory management: it's a bit more tricky to explain, in short
   we want to see if by reinforcing the memory management (avoiding
   that the memory limit is almost reached) we can improve
   the thinking while the other player is to play.

Please feel free of requesting the removal of some mogos. According to
Don's email, I have launched these mogos and posted an explanation of what 
is tested, but of course I'll remove some mogos if someone considers that 
there are too many of them.

Following this idea of the "public" nature of experiments in cgos,
I am very interested in greenpeep ("playouts guided by 
patterns extracted from offline self-play", according to 
http://senseis.xmp.net/?ComputerGoServer#toc33), I would be very 
grateful if someone could provide links/infos about it, it is seemingly 
quite innovative as it introduces an original way of learning across 
games (an efficient coevolution in Monte-Carlo planning would be
great!).

Best regards,
Olivier