The well at the end of the title is implied. And computers still can't play
19x19 Go anywhere near the master level.
Ingo Althöfer wrote:
Dear Bob Hearn,
it is not what you have been looking for, but nevertheless
I want to ask you if the title of your talk
Games Computers Can't Play is still
On Nov 18, 2008, at 7:43 AM, Michael Williams [EMAIL PROTECTED]
wrote:
The well at the end of the title is implied. And computers still
can't play 19x19 Go anywhere near the master level.
I'm not very familiar with go terms, but I think kyu means student and
dan means master.
It may
On 18-nov-08, at 11:25, Jason House wrote:
On Nov 18, 2008, at 7:43 AM, Michael Williams
[EMAIL PROTECTED] wrote:
The well at the end of the title is implied. And computers
still can't play 19x19 Go anywhere near the master level.
I'm not very familiar with go terms, but I think kyu
When Myungwan Kim guesstimated that Mogo might be dan-level, he certainly was
speaking of amateur levels, not pro. He was playing Mogo with 9 stones. If I
recall correctly, he later beat Mogo in a seven-stone game.
A pro can not offer another pro seven stones. The difference between pro shodan
Many Faces gained 5 ranks when I added MCTS to it (with about 7 months of
full time work), so I have to agree that Monte Carlo changed our world.
David
-Original Message-
From: [EMAIL PROTECTED] [mailto:computer-go-
[EMAIL PROTECTED] On Behalf Of Michael Williams
Sent: Tuesday,
On Tue, 2008-11-18 at 08:28 -0800, David Fotland wrote:
Many Faces gained 5 ranks when I added MCTS to it (with about 7 months
of
full time work), so I have to agree that Monte Carlo changed our
world.
I remember that you were not a true believer at first :-)
- Don
signature.asc
From: Don Dailey [EMAIL PROTECTED]
On Tue, 2008-11-18 at 08:28 -0800, David Fotland wrote:
Many Faces gained 5 ranks when I added MCTS to it (with about 7 months
of full time work), so I have to agree that Monte Carlo changed our
world.
I remember that you were not a true believer
On 18-nov-08, at 14:32, Don Dailey wrote:
On Tue, 2008-11-18 at 08:28 -0800, David Fotland wrote:
Many Faces gained 5 ranks when I added MCTS to it (with about 7
months
of
full time work), so I have to agree that Monte Carlo changed our
world.
I remember that you were not a true believer
On Mon, Nov 17, 2008 at 11:13 PM, Michael Williams
[EMAIL PROTECTED] wrote:
No one ever alleged that pure AMAF or pure MC was infinitely scalable.
My point is that in many cases, they doesn't even keep all of their
benefits, after some number of trials have been run. So, running 10k
playouts
Weston Markham wrote:
On Mon, Nov 17, 2008 at 11:13 PM, Michael Williams
[EMAIL PROTECTED] wrote:
No one ever alleged that pure AMAF or pure MC was infinitely scalable.
My point is that in many cases, they doesn't even keep all of their
benefits, after some number of trials have been run.
I was a skeptic until the 2007 UEC cup. Then it was obvious that MCTS was
stronger than the traditional programs.
-Original Message-
From: [EMAIL PROTECTED] [mailto:computer-go-
[EMAIL PROTECTED] On Behalf Of Mark Boon
Sent: Tuesday, November 18, 2008 8:56 AM
To: [EMAIL PROTECTED];
I think it was the surprisingly useful combination of UCT with Monte-Carlo
that got the attention of the 'old school' Go programmers.
I would say Monte-Carlo + Tree Search rather than Monte-Carlo + UCT. You
can have a very strong program without UCT.
You can't without the incremental tree +
On Tue, 2008-11-18 at 12:02 -0500, Michael Williams wrote:
Weston Markham wrote:
On Mon, Nov 17, 2008 at 11:13 PM, Michael Williams
[EMAIL PROTECTED] wrote:
No one ever alleged that pure AMAF or pure MC was infinitely scalable.
My point is that in many cases, they doesn't even keep all
On Nov 17, 2008, at 11:34 PM, Ingo Althöfer wrote:
Dear Bob Hearn,
it is not what you have been looking for, but nevertheless
I want to ask you if the title of your talk
Games Computers Can't Play is still up-to-date.
I would accept something like
Games Computers Could not play well before
I had never heard of that. A Google search turned up this list of interesting
Go variants: http://www.usgo.org/resources/downloads/deviantgo.pdf
Bob Hearn wrote:
On Nov 17, 2008, at 11:34 PM, Ingo Althöfer wrote:
Dear Bob Hearn,
it is not what you have been looking for, but nevertheless
On Tue, Nov 18, 2008 at 12:02 PM, Michael Williams
[EMAIL PROTECTED] wrote:
It doesn't make any sense to me from a theoretical perspective. Do you have
empirical evidence?
I used to have data on this, from a program that I think was very
nearly identical to Don's reference spec. When I get a
On Tue, 2008-11-18 at 14:17 -0500, Weston Markham wrote:
On Tue, Nov 18, 2008 at 12:02 PM, Michael Williams
[EMAIL PROTECTED] wrote:
It doesn't make any sense to me from a theoretical perspective. Do you have
empirical evidence?
I used to have data on this, from a program that I think
It depends very much on what exactly you mean by amateur master level. Is it
a level that compares to amateur master level in chess?
And what is amateur master level in chess? USCF master, FIDE master or
international master?
Some time ago I participated in a discussion about comparing chess
On Tue, 2008-11-18 at 19:42 +, Oliver Lewis wrote:
It doesn't make any sense to me from a theoretical
perspective. Do you have empirical evidence?
I agree that empirical evidence is required, but theoretically, if MC
converges to something that is not perfect play,
From: Oliver Lewis [EMAIL PROTECTED]
On 11/18/08, Michael Williams [EMAIL PROTECTED] wrote:
It doesn't make any sense to me from a theoretical perspective. Do you have
empirical evidence?
I agree that empirical evidence is required, but theoretically,
Dear Bob,
thanks for your explanations. Now I see clearer.
First, the title is deliberately provocative.
Accepted.
Also, though, the talk is not just about go: some of it is about
formally undecidable games, that computers provably can't play
well (and of course, that humans can't
Hello,
one of the basic problems of go newbies
is their tendency to place the next stone
near to the latest stone of the opponent.
Sometimes this is called the 2-inch heuristic
of beginners.
What do you think about a formalized variant
of Go with one-sided distance-k rule?
Let k be some
I think computers would be much better at this game (than they are at Go)
because you have vastly reduced the branching factor of the game.
Ingo Althöfer wrote:
Hello,
one of the basic problems of go newbies
is their tendency to place the next stone
near to the latest stone of the opponent.
I think a computer would play this variant well if k was small.
To make the move generation consistent, the first move should be played
as if there was a previous move to the center perhaps.
Ladders would probably still be an issue.
- Don
On Tue, 2008-11-18 at 23:20 +0100, Ingo Althöfer
Well, vastly when k is small.
The only way to find a good Komi would be testing and guesstimating.
I think MCTS would be well suited to this variant because you still have the
problem of difficulty in finding a good evaluation function and MCTS solves
that.
Computers would probably be
On Tue, 2008-11-18 at 17:31 -0500, Michael Williams wrote:
Well, vastly when k is small.
The only way to find a good Komi would be testing and guesstimating.
I think MCTS would be well suited to this variant because you still have the
problem of difficulty in finding a good evaluation
for small k, this should give a massive advantage to black.
the additional requirement that white place a stone within the
smallest cityblock distance of the last stone whenever he has
no valid move within distance k of black's last move is an even more
substantial advantage for black. i'm
I assumed the rules were symmetric, in that black also had to place his stones
within distance k to white's last move.
steve uurtamo wrote:
for small k, this should give a massive advantage to black.
the additional requirement that white place a stone within the
smallest cityblock distance
Nick Wedd: [EMAIL PROTECTED]:
My report on Sunday's KGS bot tournament is now available at
http://www.weddslist.com/kgs/past/44/index.html
Many Faces of Go was undefeated in both divisions.
Did Many Faces 1 and 2 share the same cluster or one for each?
Hideki
--
[EMAIL PROTECTED] (Kato)
one of the basic problems of go newbies
is their tendency to place the next stone
near to the latest stone of the opponent.
Sometimes this is called the 2-inch heuristic
of beginners.
How about going the other way, forcing Monte-Carlo simulations
onto a coarser grid in the hope of quickly
I don't think that would be effective because you've completely changed the
rules and hence the tactics of the game.
Claus Reinke wrote:
one of the basic problems of go newbies
is their tendency to place the next stone
near to the latest stone of the opponent.
Sometimes this is called the
I'm writing my engine from scratch and have a curious question. When my
best friend an American 1dan lvl player (who has played in Japan) taught me
the game. I love it. :)
Though as I get more in depth, and programming wise, have no idea what rule
set to follow. Not sure what I was taught
Chinese rule variants are simpler to code than Japanese. This is because the game can be played to the bitter end without affecting the final score. So you
should probably start there. There were recently some discussions on this list about how to handle Japanese scoring.
Joshua Shriver
On Tue, 2008-11-18 at 20:30 -0500, Joshua Shriver wrote:
I'm writing my engine from scratch and have a curious question. When
my best friend an American 1dan lvl player (who has played in Japan)
taught me the game. I love it. :)
Though as I get more in depth, and programming wise, have no
You've been on this list a long time, haven't you?
Yes :) I started by tinkering with Monte Carlo in VRML for a hardware
solution, but aiming for a pure C, aimed at x86 engine.
Anyway, I don't think there is much of a question that Chinese rules
are much better for getting started with
One for each. Actually they were running on a 128 core cluster. The
current code only scales to 32 cores, so only half the cluster was used.
David
-Original Message-
From: [EMAIL PROTECTED] [mailto:computer-go-
[EMAIL PROTECTED] On Behalf Of Hideki Kato
Sent: Tuesday, November 18,
36 matches
Mail list logo
