On Thu, 2007-04-05 at 09:15 +0900, Darren Cook wrote:
> This is an interesting result, on one of my favourite topics. I've
> spent
> a lot of time analyzing 9x9 games by very strong/pro players in detail
> and decided correct komi is probably either 6 or 7 points. Give me a
> few
> more years and I
Does Japanese rules make any difference in what komi to use?
- Don
On Thu, 2007-04-05 at 09:15 +0900, Darren Cook wrote:
> > Here are the results (wins for black):
> > komi 6.5:
> > 1943/3640 (53%)
> > komi 7.5
> > 1754/3640 (48%)
>
> This is an interesting result, on one of my favourite topics.
On 4/5/07, Don Dailey <[EMAIL PROTECTED]> wrote:
On Wed, 2007-04-04 at 23:35 +0200, Erik van der Werf wrote:
> Sylvain, could you run the same test on 7x7 to verify that there the
> 'correct' komi would be 9 (try 8.5 vs 9.5)?
I can already tell you what will happen, I have done similar tests.
A
> Here are the results (wins for black):
> komi 6.5:
> 1943/3640 (53%)
> komi 7.5
> 1754/3640 (48%)
This is an interesting result, on one of my favourite topics. I've spent
a lot of time analyzing 9x9 games by very strong/pro players in detail
and decided correct komi is probably either 6 or 7 poi
On Thu, 2007-04-05 at 01:06 +0200, Erik van der Werf wrote:
> You seem to be more certain on this than Sylvain :-)
> Let's see what he gets out with 'only' the 3k playouts...
Yes, but I think it will be even clearer that there is no
fair fractional komi. Even if I had not done the test
several
It also makes sense to me that the "ready" player should be
the anchor ... or either anchor if there are 2.
Cheers,
David
On 4, Apr 2007, at 2:18 AM, Heikki Levanto wrote:
On Mon, Apr 02, 2007 at 02:34:47PM -0400, Don Dailey wrote:
On Mon, 2007-04-02 at 14:26 -0400, Chris Fant wrote:
Does
On 4/4/07, Don Dailey <[EMAIL PROTECTED]> wrote:
On Wed, 2007-04-04 at 23:35 +0200, Erik van der Werf wrote:
> Traditionally the weaker player plays black and it seems reasonable
> that starting the game should not be a disadvantage. Consequently, if
> these statistics are at all admissible, my c
Could this means that the perfect play is along the line of 0.5 point win? It
may well be, considering MoGo can surprise a 6 dan on 9x9. It could be that a
large numbers of games that are perfect or near perfect are centered along the
0.5 point win. If one increase the komi by one point, basical
On Wed, 2007-04-04 at 23:35 +0200, Erik van der Werf wrote:
> Sylvain, could you run the same test on 7x7 to verify that there the
> 'correct' komi would be 9 (try 8.5 vs 9.5)?
I can already tell you what will happen, I have done similar tests.
At this point, the good UCT programs are so stron
Sylvain, could you run the same test on 7x7 to verify that there the
'correct' komi would be 9 (try 8.5 vs 9.5)? If MoGo wouldn't converge
to 9 we probably shouldn't have much confidence in the generalisation
of the above results for higher levels of play on 9x9 (or you could be
on your way to dis
It seems 7.5 is the best komi even though it's against the intuition (totally).
I'm intrigued by the fact that MC score can be accurate to 0.5 point on 9x9.
Daniel Liu
-Original Message-
From: [EMAIL PROTECTED]
To: [EMAIL PROTECTED]; computer-go@computer-go.org
Sent: Wed, 4 Ap
On Wed, 2007-04-04 at 17:56 -0400, Don Dailey wrote:
> One adjustment that could made, if we change komi to 6.5 is to
> give black an additional 20 ELO point for rating purposes. This
> makes each rating pass slightly more fair. In other words, black
> would gain slightly less from winning and l
If 7.0 is in fact the correct komi, I suppose this test also indicates
that it's easier for black to win with 6.5 komi than it is for white to
win with 7.5 komi. My test also showed that. So I suppose 7.5 is
the least unfair komi.
It is true that these experiments show that at MoGo_3k level
By the way, on the 2 minute server, here are the statistics:
Total games: 19383
Total white wins: 10024
white wins: 51.71542 percent
- Don
On Wed, 2007-04-04 at 23:35 +0200, Erik van der Werf wrote:
> Traditionally the weaker player plays black and it seems reasonable
> that st
Quoting Don Dailey <[EMAIL PROTECTED]>:
It actually surpises me that go players care about this. I thought GO
was more about the beauty of ommision and the unstated understanding
of event that don't actually have to happen to be appreciated.
I think most players weaker than about amateur 1-3
On 4/4/07, Sylvain Gelly <[EMAIL PROTECTED]> wrote:
> However, it would be accurate to play a few thousand games at 7.5 komi,
> then a few thousand and 6.5 and compare the white/black win percentage. The
> one closest to 50% would be the one to use.
If some are interested by the results, I jus
Hi Sylvain,
Thank you for running that test! It verifies my earlier results and
it shows that either 6.5 or 7.5 is as fair as you can get. Presumably,
6.5 is unfair for white and 7.5 is unfair for black.
If 7.0 is in fact the correct komi, I suppose this test also indicates
that it's easie
However, it would be accurate to play a few thousand games at 7.5 komi,
then a few thousand and 6.5 and compare the white/black win percentage. The
one closest to 50% would be the one to use.
If some are interested by the results, I just did a test.
9x9 games, MoGo with 3k simulations per move
I think I am going to to use a simple, single command
format for giving the engine information about the
game and players.
It will be:
cgos-gameinfo gameID whitePlayer blackPlayer
This keeps it simple, there is only one GTP command
to implement and no sub-commands to parse separately.
Of
If you look at the games of AnchorMan, a weak player but one that
uses the standard approach of playing for the safe win - you may
notice that it tends to play more naturally and does not suffer
as much from this problem.
That's because it has tiny incentives to avoid moving into
territory heavily
On Wed, 2007-04-04 at 10:40 -0700, terry mcintyre wrote:
> If both Monte Carlo players strive for 0.5 point wins, then almost any
> ballpark komi would lead to a 50/50 split, a sort of self-fulfilling
> prophecy?
If the komi is wrong, you would see it in the score of white vs black
over a
large nu
Would it be possible to encourage more substantial wins by tweaking the
"internal
komi" used to drive move selection in this fashion?
I tried that in two ways.
First, try to put a bigger "internal komi" at the beginning decreasing to 0
(assuming you play as black), so that the programs tries
If both Monte Carlo players strive for 0.5 point wins, then almost any ballpark
komi would lead to a 50/50 split, a sort of self-fulfilling prophecy?
What happens when a player sets a more difficult komi than the one used for
scoring? Sometimes Go players use larger komi as a sort of handicap. W
It is a good idea to write a press release with the key points laid out in the
form of an article. Quote yourself in your press release - "Go Researcher
Sylvain Gelly said yadda yadda." Many published articles are almost direct
copies of press releases. This way, you can encourage more accurate
Chrilly,
I was recently interviewed by someone in the MIT wrting department
(about computer chess) and I was very hesitant to agree.I have
some anxiety about how it will come out! However, the journalist
seemed to really want to get the facts straight and I have some
hope ...
In general,
One further important rule. One should never be ironic in interviews. The
ironie is almost always lost.
E.g. when we played against Adams the default question was "why do you not play
against Kasparov". I could not stand this question anymore and in a press
conference shortly before the match I
On Wed, 2007-04-04 at 15:25 +0200, Unknown wrote:
> According to the GTP-draft, you are supposed to use hyphens, not
> underscores to prefix your private extensions. (I don't like them
> either, would have preferred periods or semicolons.)
Thanks for pointing that out to me - I would use hyphens
On Wed, 2007-04-04 at 09:17 -0400, [EMAIL PROTECTED] wrote:
> 7.5 komi is for the 19x19 game (7.5/361). It may not be correct for
> the 9x9 (7.5/81). From games played on CGOS what is the correct
> komi?
It's my strong feeling that 7.5 is the right komi if you are to use
fractional komi to avoi
The best thing one can do is to write the articles already for the journalists.
Or to formulate at least some key sentences. But even this does not really help
to be misquoted.
If one wants to be quoted, one has to live with the misquotes. As long as there
is no really bad intension, its in my o
On Wed, 2007-04-04 at 09:17 -0400, [EMAIL PROTECTED] wrote:
> 7.5 komi is for the 19x19 game (7.5/361). It may not be correct for
> the 9x9 (7.5/81). From games played on CGOS what is the correct komi?
>
> Daniel Liu
It's not possible to figure this because most of the programs
on CGOS use the M
On Wed, 2007-04-04 at 15:25 +0200, Unknown wrote:
> either, would have preferred periods or semicolons.)
Oops. I meant colons, of course.
AvK
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-
On Tue, 2007-04-03 at 23:10 -0400, Don Dailey wrote:
> On Tue, 2007-04-03 at 23:01 -0400, Don Dailey wrote:
> >
> > cgos_player color name
> > example: cgos_player white Lazarus
> >
> > cgos_elo color elo_rating
> > example: cgos_elo white 1739?
> >
> > and since cgos does use kyu/dan,
7.5 komi is for the 19x19 game (7.5/361). It may not be correct for the 9x9
(7.5/81). From games played on CGOS what is the correct komi?
Daniel Liu
AOL now offers free email to everyone. Find out more about what's f
Hi Heikki,
I already reject using cgos_name because "name" is being used by
the GTP protocol.
I did consider whether to distinguish between the opponent and
the player who is receiving the message. But it does not seem
in the spirit of of GTP and in programming there is the similar
principle of
Thank you Don.
I did not know that, I am not used to :-).
Then I'll stop worrying for these kind of things and stop trying to give
back the truth :).
Bye,
Sylvain
2007/4/4, Don Dailey <[EMAIL PROTECTED]>:
On Wed, 2007-04-04 at 10:52 +0200, Sylvain Gelly wrote:
> You should also know that we n
On Wed, 2007-04-04 at 10:52 +0200, Sylvain Gelly wrote:
> You should also know that we never claimed that "MoGo plays 9x9 go
> near the level of a professional go player", which is of course false,
> and even if it was true should ask for many many experiments, and we
> would have never say that.
On Mon, Apr 02, 2007 at 02:34:47PM -0400, Don Dailey wrote:
>
> On Mon, 2007-04-02 at 14:26 -0400, Chris Fant wrote:
> > Does the new cgos have a standby player that fills in when an odd
> > number of engines are logged-in?
>
> I want it to have that, but it doesn't yet. But it will.
>
> I was
On 4/4/07, Sylvain Gelly <[EMAIL PROTECTED]> wrote:
Of course as experts, you should have noticed errors on this newsletter, as
e.g. MoGo developed by the inventors of UCT in hungary :-).
Yes, the text clearly showed that the guy didn't do his homework.
One (far-fetched) explanation might be b
Hello Terry,
Sylvain,
Were you aware of this challenge from the American Go Association? The
following is from the latest AGA newsletter; you can send corrections or
replies to [EMAIL PROTECTED]
Yes I was aware, Roy Laird asked me to put MoGo on KGS to play these games,
and it is what I did.
39 matches
Mail list logo
