[Computer-go] Michi-go: Michi implemented in the Go language

[Clark B. Wierda]

I have a conversion of Petr Baudis' Michi in the Go language.

The code is on Github at https://github.com/traveller42/michi-go

I'm still cleaning a few things up, but I've fixed the worst of the errors
I created during translation. This first pass is as close to a literal
translation of the Python as I know how to do.  Some of the limitations
have been in my understanding of either Python or Golang. (One goal of this
endeavor was to learn more about both languages.)  Once the cleanup is
complete, I will declare version 1 complete.  I will maintain a version 1
branch to allow me to accept any fixes to errors that are found in this
code, or better matches to the Python idioms.

Version 2 is the target for a more idiomatic Golang implementation of this
base code.  This is where I will pursue improvements in efficiency.

Any comments are welcome.

Regards,
Clark B. Wierda
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

[Computer-go] GPW Cup 2015 result

[Hiroshi Yamashita]

Hi,

GPW Cup 2015 was held at November 6th-7th in Karuizawa, Nagano, Japan.
Result are

9x9
1st Zen  10-0
2nd Nomitan   7-3
3rd Aya   6-4

13x13
1st Aya   5-1
2nd Zen   5-1
3rd Ray   3-3

9x9, 10 minutes SD, komi 7.0, Chinese rule, double round-robin

zen  aya  mca  ray  nom  YK0   Wins  Rank
Zen11   11   11   11   11   10-0   1
Aya   0011   11   00   116-4   3
MC_ark00   0010   01   004-6   4
Ray   00   0001   00   113-7   5
Nomitan   00   11   10   11117-3   2
YK02  00   00   00   00   00 0-10  6


13x13, 20 minutues SD, komi 7.5, Chinese rule, round-robin

zen  aya  mca  ray  nom  mura wata  Wins SB Rank
Zen011111 5   10   2
Aya   1 10111 5   12   1
MC_ark00 0010 17
Ray   011 001 38   3
Nomitan   0011 00 24   6
Muramatsu 00011 0 25   5
Watanabe  001011  35   4

* YK02 is a new program by Kunihito Hoki.
* Muramatsu and Watanabe are human player, they are around 4d on KGS.
* computers use one thread on note-PC.
* SGF's komi is not correct.

SGF
http://yss-aya.com/gpw2015.zip

Regards,
Hiroshi Yamashita

___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go

Re: [Computer-go] alarming UCT behavior

[Hideki Kato]

I often found similar for Zen with enough memory.  I and Yamato's 
interpretation follows.  

Yamato tunes many parameters with relatively shorter time settings 
because a huge number of games are necessary for the tuning.  Assuming 
the prior bias and the average outcome from the simulations have to be 
balanced (using multiple parameters)  for strong play: patterns give 
good shapes but almost nothing for local fights and complicated L 
which should be solved by the simulations.  Yamato optimized these 
parameters at a shorter time setting and so the balance is broken (i.e., 
worse than optimum) at longer time settings.

Recently, Yamato tunes the parameters at as long time setting as 
possible.

Hideki

Dave Dyer: <20151106184850.13268e1...@computer-go.org>:
>
>Developing a UCT robot for a new game, I have encountered a
>surprising and alarming behavior:  the longer think time the
>robot is given, the worse the results.  That is, the same robot
>given 5 seconds per move defeats one give 30 seconds, or 180 seconds.
>
>I'm still investigating, but the proximate cause seems to be
>my limit on the size of the UCT tree.   As a memory conservation
>measure, I have a hard limit on the size of the stored tree. After
>the limit is reached, the robot continues running simulations, refining
>the outcomes based on the existing tree and random playouts below
>the leaf nodes.
>
>My intuition would be that the search would be less effective in this
>mode, but producing worse results (as measured by self-play) is 
>strongly counter intuitive.
>
>Does it apply to Go?  Maybe not, but it's at least an indicator
>that arbitrary decisions that "ought to" be ok can be very bad in
>practice.
>
>___
>Computer-go mailing list
>Computer-go@computer-go.org
>http://computer-go.org/mailman/listinfo/computer-go
-- 
Hideki Kato 
___
Computer-go mailing list
Computer-go@computer-go.org
http://computer-go.org/mailman/listinfo/computer-go