Peter Drake wrote:
On Aug 1, 2008, at 8:08 AM, Mark Boon wrote:
The neighbours of the last move come in the picture because usually
it's only the last stone played that can be escaping a ladder and it's
the neighbours of the last move that could have been put into atari.
Nothing to do with
On Aug 2, 2008, at 4:31 AM, Gunnar Farnebäck [EMAIL PROTECTED]
wrote:
It's often a good idea to bias capturing moves in the playouts,
regardless whether it's a ladder or not. This would result in those
stones being captured in most simulations.
What method do people use for finding capture
On Sat, Aug 2, 2008 at 9:43 AM, Jason House [EMAIL PROTECTED] wrote:
On Aug 2, 2008, at 4:31 AM, Gunnar Farnebäck [EMAIL PROTECTED] wrote:
It's often a good idea to bias capturing moves in the playouts,
regardless whether it's a ladder or not. This would result in those
stones being captured
Depending on your implementation, it may be faster to re-derive the
liberty list when you need it. For example, if your playout move
suggester only suggests capturing the last stone played, you don't
need to do all of the work to update liberty counts for any other
chains.
Thanks a lot
I keep liberty counts.
David
-Original Message-
From: [EMAIL PROTECTED] [mailto:computer-go-
[EMAIL PROTECTED] On Behalf Of Jason House
Sent: Saturday, August 02, 2008 6:43 AM
To: computer-go
Subject: Re: [computer-go] Ladders and UCT again
On Aug 2, 2008, at 4:31 AM, Gunnar
: [computer-go] Ladders and UCT again
On Aug 2, 2008, at 4:31 AM, Gunnar Farnebck [EMAIL PROTECTED]
wrote:
It's often a good idea to bias capturing moves in the playouts,
regardless whether it's a ladder or not. This would result in those
stones being captured in most simulations.
What method
On Sat, Aug 2, 2008 at 16:06, Álvaro Begué [EMAIL PROTECTED] wrote:
On Sat, Aug 2, 2008 at 9:43 AM, Jason House [EMAIL PROTECTED] wrote:
On Aug 2, 2008, at 4:31 AM, Gunnar Farnebäck [EMAIL PROTECTED] wrote:
It's often a good idea to bias capturing moves in the playouts,
regardless whether
-
[EMAIL PROTECTED] On Behalf Of Jason House
Sent: Saturday, August 02, 2008 6:43 AM
To: computer-go
Subject: Re: [computer-go] Ladders and UCT again
On Aug 2, 2008, at 4:31 AM, Gunnar Farneb ck [EMAIL PROTECTED]
wrote:
It's often a good idea to bias capturing moves in the playouts
stronger. All of these numbers are for
9x9.
David
-Original Message-
From: [EMAIL PROTECTED] [mailto:computer-go-
[EMAIL PROTECTED] On Behalf Of Lukasz Lew
Sent: Saturday, August 02, 2008 3:50 PM
To: computer-go
Subject: Re: [computer-go] Ladders and UCT again
Can you describe your
] [mailto:computer-go-
[EMAIL PROTECTED] On Behalf Of Jason House
Sent: Saturday, August 02, 2008 6:43 AM
To: computer-go
Subject: Re: [computer-go] Ladders and UCT again
On Aug 2, 2008, at 4:31 AM, Gunnar Farneb ck [EMAIL PROTECTED]
wrote:
It's often a good idea to bias capturing moves
On 31-jul-08, at 21:00, Peter Drake wrote:
What you're missing (or so it seems to me) is that it's not to
prevent from running a ladder that is caught.
Really? My motivation has been to prevent my program from
embarrassingly running in just those situations. Is there something
other
On Aug 1, 2008, at 8:08 AM, Mark Boon wrote:
The neighbours of the last move come in the picture because usually
it's only the last stone played that can be escaping a ladder and
it's the neighbours of the last move that could have been put into
atari. Nothing to do with the additional
From: Peter Drake [EMAIL PROTECTED]
It seems too expensive to search every point on the board for ladders. What
to do?
Perhaps it would be worthwhile to preserve state from board to board - there
are unlikely to be more than 2 or 3 ladders on any given board. When a stone is
played,
On 1-aug-08, at 14:15, Peter Drake wrote:
On Aug 1, 2008, at 8:08 AM, Mark Boon wrote:
The neighbours of the last move come in the picture because
usually it's only the last stone played that can be escaping a
ladder and it's the neighbours of the last move that could have
been put into
Ladder read should be helpful in the open space. But when the surrounding
area is not full of space, the ladder play could hurt the eyes.
for example below in the corner. b is better than a. By ladder reading, it
would play at a, and the defender answer at b, leaves the attacker no eye in
the
I know we had this conversation recently, but I just can't seem to get
my head around writing a ladder reader. What, exactly, does the ladder
reader do?
Our approach was to read out ladders involving the last stone played.
In the playout (beyond the tree), if the attacker can capture by
On Thu, 2008-07-31 at 15:50 -0700, Peter Drake wrote:
I know we had this conversation recently, but I just can't seem to get
my head around writing a ladder reader. What, exactly, does the ladder
reader do?
Our approach was to read out ladders involving the last stone played.
In the
On 31-jul-08, at 19:50, Peter Drake wrote:
I know we had this conversation recently, but I just can't seem to
get my head around writing a ladder reader. What, exactly, does the
ladder reader do?
Our approach was to read out ladders involving the last stone
played. In the playout
I did a ladder reader once. Basically it's an alpha beta search where
you focus on one group only, and that group has limited liberties. If
it's strictly ladder reading, you only consider attacks that reduce the
liberty count of a specific string to 1 (atari moves in other words) and
defenses
On Jul 31, 2008, at 4:24 PM, Mark Boon wrote:
On 31-jul-08, at 19:50, Peter Drake wrote:
I know we had this conversation recently, but I just can't seem to
get my head around writing a ladder reader. What, exactly, does the
ladder reader do?
Our approach was to read out ladders
From: Peter Drake [EMAIL PROTECTED]
Our approach was to read out ladders involving the last stone played.
In the playout (beyond the tree), if the attacker can capture by
continuing a ladder, the attacker plays that move. If the defender can
escape by running, the defender plays that move.
Okay, let me see if I can sum this all up.
Let 2, capture, attacker stand for defending chain has 2
liberties, it will be captured if the ladder is played out, and it is
the attacker's turn.
Use the following rules to suggest moves:
1, capture, defender = defender plays ladder breaker,
On 17-jun-08, at 22:26, Peter Drake wrote:
Without the 10% random moves, would every playout from a given leaf
be identical?
I have not exprimented with introducing randomness. I always play
ladder-capturing and ladder-escaping moves.
But that does not imply at all that all playouts
Am 17.06.2008, 06:54 Uhr, schrieb Russ Williams [EMAIL PROTECTED]:
On Mon, Jun 16, 2008 at 11:42 PM, WSK [EMAIL PROTECTED]
wrote:
Am 16.06.2008, 22:20 Uhr, schrieb Peter Drake [EMAIL PROTECTED]:
UCT
The idea of MonteCarlo and UCT contain the elemination of special game
rules
( include
On Mon, Jun 16, 2008 at 4:42 PM, WSK [EMAIL PROTECTED] wrote:
program oh_iam_sooo_quick_and_dirty;
get_input_move(input);
.
.
.
do
{
set_randomseed(); //sic! :)
do
{
color = white;
pass = 0;
if(pass == play_random_move(color); // moves or
which is a function which returns either
success fail or unknown in order to feed this information into
the playout process.
--- On Mon, 6/16/08, Peter Drake [EMAIL PROTECTED] wrote:
From: Peter Drake [EMAIL PROTECTED]
Subject: Re: [computer-go] Ladders and UCT
To: computer-go computer-go
In Sunday's tournament, Orego lost a game embarrassingly by playing
out a lost ladder. I know how to write a ladder checker in general,
but I'm not sure how to incorporate such a thing into UCT. What are
other people doing?
Peter Drake
http://www.lclark.edu/~drake/
You need to move away from light playouts and play ladder-capturing
and ladder-escaping moves during playout. It will then still
occasionally play out a ladder, but only if it thinks it's far behind.
On 16-jun-08, at 17:20, Peter Drake wrote:
In Sunday's tournament, Orego lost a game
On Mon, Jun 16, 2008 at 11:42 PM, WSK [EMAIL PROTECTED] wrote:
Am 16.06.2008, 22:20 Uhr, schrieb Peter Drake [EMAIL PROTECTED]:
UCT
The idea of MonteCarlo and UCT contain the elemination of special game rules
( include experienced moves like ladders ) and used the real ability of
computers:
29 matches
Mail list logo
