I extended the mogo vs gnugo test 1 more level. To remind you, these
are 9x9 games. mogo level 01 is 64 playouts and each subsequent level
is a doubling, i.e. 32 * (2 ** LEV)(32 times 2 raised to the LEV
power)
After level 10, it looks like a fall-off, but Mogo at 13 showed a nice
increas
What should the komi be for 13x13 Go?
- Don
___
computer-go mailing list
computer-go@computer-go.org
http://www.computer-go.org/mailman/listinfo/computer-go/
trong program authors in private and many
people do not want to distribute a binary.
- Don
[EMAIL PROTECTED] wrote:
> > -Original Message-
> > From: Don Dailey <[EMAIL PROTECTED]>
> > To: computer-go
> > Sent: Mon, 18 Feb 2008 1:45 pm
> > Subject:
advantage, but not a "true"
advantage.
- Don
>
>
> Don Dailey wrote:
>> Hi David,
>>
>> Any opinion either of us have on this is only speculation.
>> Nevertheless, in any kind of science there tends to be unproven
>> conjectures that
target as boards get
> sufficiently huge, but it seems quite possible that the exact number
> of playable points on the board can result in some minor differences
> in score even as board sizes get very large, and it seems like it
> would take a rigorous proof for one to abandon that
The funny thing about MC scoring is that if white DOES win, then it
doesn't matter what black plays, everything loses!
That would mean that in a scalability study, where colors are
alternated, really strong versions of Mogo would score poorly against
the weaker programs.
In a serious compe
The mogo study completed. I played 1000 games per level doubling with
mogo vs Gnugo, starting at 64 play-outs (Mogo_01).
Gnugo-3.7.11 is played at level 8, as in the big study.
Rank Name Elo+- games score oppo. draws
1 Mogo_11 2361 56 48 1000 96% 18000%
Here is the entire tree, where I drop nodes if they have less than 500
samples. These are games between 1700+ players who are within 100 ELO
of each others rating.
E5 49.1% 19630
| C4 49.6% 5894
| | C5 49.9% 1558
| | | B5 54.7% 788
| | | | C6 48.4% 566
| | C6 5
>From CGOS data it looks like WHITE has the better winning chances with
komi set at 7.5.
Take this all with a grain of salt because data like this can be
misleading. The weaker moves, for instance, may be better than they
appear due to the possibility that the poorer results are caused by
weaker
| G2 75.0% 4
| B4 100.0% 3
| C4 100.0% 3
| H3 100.0% 3
| C1 0.0% 2
| H2 100.0% 2
| F1 0.0% 2
| G5 100.0% 2
| J3 100.0% 1
| A4 0.0% 1
| A2 0.0% 1
| B2 100.0% 1
| D2 100.0% 1
Don Dailey wrote:
> Here are the statistics on all the CGOS games through Janu
Here are the statistics on all the CGOS games through January 2008 with
the winning percentages. This is from the point of the view of the
player making the move.
The most popular move is E5 and it also is the only move with a positive
score. I consider all games regardless of how strong or ho
0.0% 1
| | F4 100.0% 1
| | G4 0.0% 1
| | D3 100.0% 1
| | C4 0.0% 1
| D5 66.7% 3
| | D4 66.7% 3
Jason House wrote:
> That's 41% of the 206 games that begin with E5 {G5,C5,E7,E3}
>
> On Feb 12, 2008 1:38 PM, Don Dailey <[EMAIL PROTECTED]
>
x27;s 41% of the 206 games that begin with E5 {G5,C5,E7,E3}
>
> On Feb 12, 2008 1:38 PM, Don Dailey <[EMAIL PROTECTED]
> <mailto:[EMAIL PROTECTED]>> wrote:
>
> This pattern doesn't appear to often:
>
> 5 E5 G5 G4
> 6 E5 C5 C4
> 8
E5 H5
2 B E5 H5
1 B E5 H6
- Don
Jason House wrote:
> That's 41% of the 206 games that begin with E5 {G5,C5,E7,E3}
>
> On Feb 12, 2008 1:38 PM, Don Dailey <[EMAIL PROTECTED]
> <mailto:[EMAIL PROTECTED]>> wrote:
>
> This pattern doesn't appe
This pattern doesn't appear to often:
5 E5 G5 G4
6 E5 C5 C4
8 E5 C5 C6
12 E5 E7 D7
12 E5 E7 F7
13 E5 G5 G6
14 E5 E3 D3
15 E5 E3 F3
(assuming I transposed these correctly)
- Don
Olivier Teytaud wrote:
>> 46 E5 C5
>
> By the way, mogoRelease with m
16 E5 D4
19 E5 F6
19 E5 G3
21 E5 D6
22 E5 F4
23 E5 C3
25 E5 C7
26 E5 G7
36 E5 G6
41 E5 D3
41 E5 F3
43 E5 C4
46 E5 C5
47 E5 D7
48 E5 G4
50 E5 F7
50 E5 G5
51 E5 C6
51 E5 E7
59 E5 E3
- Don
Magnus
Christoph Birk wrote:
> On Feb 11, 2008, at 9:39 PM, Don Dailey wrote:
>> My feelings on this seem to match at least one source:
>>
>> Look here:http://senseis.xmp.net/?Komi
>>
>> Here is an excerpt:
>>
>> It is widely believed that the cor
David Schneider-Joseph wrote:
> On Feb 11, 2008, at 8:42 PM, Don Dailey wrote:
>
>> David Schneider-Joseph wrote:
>>> On that topic - might it be possible that the notion of a "proper
>>> komi", derived as it is from "the hand of God" (perf
er than pros, so how can it give better
>> information about proper komi?
>>
>> On Feb 11, 2008 6:09 PM, Christoph Birk <[EMAIL PROTECTED]
>> <mailto:[EMAIL PROTECTED]>> wrote:
>>
>> On Mon, 11 Feb 2008, Don Dailey wrote:
>> > I don'
owing
about 0.547 in the score.
I don't believe what Alford says about 9.5 being the correct komi for
9x9.Where does that information come from?
- Don
>
> On Feb 11, 2008 6:09 PM, Christoph Birk <[EMAIL PROTECTED]
> <mailto:[EMAIL PROTECTED]>> wrote:
>
>
Christoph Birk wrote:
> On Mon, 11 Feb 2008, Don Dailey wrote:
>> Is your question whether 7.0 or 8.0 is the best komi? Or do you
>> suspect a different 1/2 komi value is best?
>
> I wonder what the true komi is ... I don't know (nobody knows?) if
> it's fr
Christoph Birk wrote:
> On Mon, 11 Feb 2008, Olivier Teytaud wrote:
>> With 20 minute games, some people succeed in winning games
>> against the release 3 of MoGo. But for
>> X-hours-per-move, I don't know.
>
> What are the self-play results (white vs. black) for "hour-long"
> games of Mogo?
> I
Thanks Olivier,
I will take care that the source code is not distributed to anyone else.
- Don
Olivier Teytaud wrote:
> For people requesting mogoRelease3 without the bug for long
> computation times due to a float instead of a double:
>
> http://www.lri.fr/~teytaud/mogo (32 bits version, with
I like AIGO too.It has more features than Ogo and a better user
interface with nicer graphics in my opinion and I bought a copy.
In response to Ben's post about Ogo not being significantly stronger, I
will present what I have found in my tests.
1. I never tested 19x19, I can't say whether
> That makes sense, considering past programs. But Don Dailey is apparently on
> the trail of a faster and better player for the Palm.
> On the one hand, I wonder if the Palm architecture has reached end-of-life.
> But on the other hand, discovering how to improve Go progr
Ian Osgood wrote:
> On Feb 7, 2008, at 8:43 AM, Don Dailey wrote:
>
>> I'm thinking about upgrading my palm OS Go program and I'm looking for
>> ways to strengthen it without adding a lot of code or memory.
>> Basically I want the strongest possible program tha
I'm thinking about upgrading my palm OS Go program and I'm looking for
ways to strengthen it without adding a lot of code or memory.
Basically I want the strongest possible program that is truly "light" in
every way.
PDA's of course are MUCH slower than PC's, and memory is at a premium
alth
gt;
> Mark
>
> On 6-feb-08, at 18:21, Don Dailey wrote:
>
>>
>>
>> Jason House wrote:
>>> Just curious if anyone knows if this is possible. cgosView has a mac
>>> (universal) binary that I'd expect to be runnable. I just don't know
>>>
I asked that question wrong - what is the resolution of the screen?
Does it scroll to simulate higher resolution?
- Don
Jason House wrote:
>
> On Feb 6, 2008 3:21 PM, Don Dailey <[EMAIL PROTECTED]
> <mailto:[EMAIL PROTECTED]>> wrote:
>
> What is the screen siz
Jason House wrote:
> Just curious if anyone knows if this is possible. cgosView has a mac
> (universal) binary that I'd expect to be runnable. I just don't know
> if the iPhone is restricted or has any other special requirements. I
> believe it runs Mac OSXI may have an iPhone in my future and
If anyone is interested, I adapted the bayeselo program to GO.
bayeselo can take pgn chess files and rate the players.
The problem with bayeselo for go is:
1. Rémi's program is interactive - not command line parameter driven.
So it's not easy to use inside of scripts.
2. It doesn't un
Here is an update from the new 1000 game test using gungo at level 8
instead of 10.
Rank Name Elo+- games score oppo. draws
1 Gnugo-3.7.11 1800 34 30 2186 97% 11370%
2 Mogo_03 1507 48 56 186 16% 18000%
3 Mogo_02 1202 43 51 1000
Darren Cook wrote:
Unfortunately, I used level 10 in the gnugo only games but in the big
study we use level 8. ...
>
>
>> ... it's a major pain running those games and it ties up my
>> machine.
>>
>
> Hi Don,
> I just know your reply is going to make me slap my he
Rémi Coulom wrote:
> 荒木伸夫 wrote:
>> I have considered this, and I think that this may be caused by wrong
>> training model. In my master thesis, I mentioned that the
>> relationship between
>> top 1 accuracy of move prediction and the strength of Monte-Carlo
>> is not simple (I increased the nu
Christoph Birk wrote:
>> Unfortunately, I used level 10 in the gnugo only games but in the big
>> study we use level 8. It's my understanding there is little difference
>> between these 2 but we can probably assume Mogo might be a little better
>> than indicated relative to the big scalability
Christoph Birk wrote:
>> Unfortunately, I used level 10 in the gnugo only games but in the big
>> study we use level 8. It's my understanding there is little difference
>> between these 2 but we can probably assume Mogo might be a little better
>> than indicated relative to the big scalability
David Fotland wrote:
>
> Can you elaborate on what is in a node, and what you mean by expand?
> I assume you have simple node, where each node represents a position
> and the single move to get there. Then when you find a node with no
> children and expand it, you allocate up to 81 new nodes,
As promised, to answer Rémi, I did a study with mogo vs Gnu at various
levels. There is NO self play involved, Gnugo-3.7.11 is the only
opponent for progressively higher rated version of Mogo.
Here are the raw results so far:
Rank Name Elo+- games score oppo. draws
1 Mogo_
nd perhaps even make the
code look cleaner.
If I ever get around to implementing it myself, I'll let you know.
- Don
> Joel
>
> On Tue, Feb 5, 2008 at 2:44 AM, Don Dailey <[EMAIL PROTECTED]> wrote:
>
>> Joel Veness wrote:
>> > Hi Nick,
>> &g
Joel Veness wrote:
> Hi Nick,
>
> Goanna (agog) timed out annoyingly in that game against GNU.
>
> I have since implemented a rule: "if after some number of samples you
> have a winning probability that is very close to 1.0, just make the
> best move right away". There is no need to spend so long
The mini study so far ... games against gnugo-3.7.11
PLAYERTIME/GME RATING GAMES WIN%
--- - ---
Mogo_01 0.10 1002.4189 0.53 Mogo at 64 play-outs
Mogo_02 0.14 1197.0156 2.56 Mogo at 128 play-outs
Mog
nt amount of data for
these levels.
gnugo is gnugo-3.7.11 at level 10
- Don
Don Dailey wrote:
> Rémi Coulom wrote:
>
>> I believe the main problem is that the Elo-rating model is wrong for
>> bots. The phenomenon with Mogo is probably the same as Crazy Stone: if
>>
Rémi Coulom wrote:
> I believe the main problem is that the Elo-rating model is wrong for
> bots. The phenomenon with Mogo is probably the same as Crazy Stone: if
> there are enough strong MC bots playing to shield the top MC programs
> from playing against GNU, then they'll get a high rating bec
But FatMan is still 1800!I wonder if FatMan improved causing the
deflation? :-)
- Don
Hideki Kato wrote:
> Hmm, mogo-pr-1core is also getting lower rating these days. It had
> been over 2500, 2525 at max I remember, but is 2476 today.
>
> -Hideki
>
> Yamato: <[EMAIL PROTECTED]>:
>
>> Gi
It's a shame a priority queue can't be used. But after each
simulation, all sibling change together.
- Don
Harri Salakoski wrote:
> Hi such question that do you typically traverse all child objects or
> is there faster way to select explored node child object.
> I have concluded that it is not
Hi Chuck,
Thank you for your interesting suggestions.I have previously
considered a system where the distribution is based on how many
contestants. For instance if there are hundreds of players you would
want to generate best of 5 or 6 or more, but if there were only 3 or 4
you might want be
> Here's another approach: "Range Voting"
>
> http://rangevoting.org/rangeVborda.html
>
> The author of this particular page makes much of the pitfalls of strategic
> voting, which should not matter to a set of emotionally disinterested,
> independent agent routines. But range voting has one fur
>> The author of this particular page makes much of the pitfalls of strategic
>> voting, which should not matter to a set of emotionally disinterested,
>> independent agent routines. But range voting has one further advantage over
>> borda voting: expressiveness. If an agent is given 99 votes t
terry mcintyre wrote:
>
> Here's another approach: "Range Voting"
>
>
> http://rangevoting.org/rangeVborda.html
>
That particular article doesn't come across as being very scholarly.
I would much prefer to see a good quality paper on it.
Take this with a grain of salt, but here is my ow
My main point is that voting is a pandora's box. You make it sound
like anything goes but most tries and intuitions will be broken.
So I still suggest you start with something that is known to have good
theoretical properties and use that as your base-line. Voting theory
is not ad-hoc, it h
FYI :We just added a program called "bigMogo_16" to the scalability
study.This is basically a control to see if mogo performs
significantly better when more nodes are allocated per our earlier
discussions.
bigMogo_16 is Mogo_16 with 1,600,000 allocated instead of the default
400,000, an
voting theory and the pitfalls they have and you
may be surprised.
Terry Mcintyre's last message was insightful, some kind of voting
system can greatly simplify the task of trying to normalize so many
disparate systems of valuing moves.
- Don
>
>
> Cheers,
> David
>
>
&g
Jason House wrote:
>
>
> On Jan 29, 2008 10:16 AM, Jason House <[EMAIL PROTECTED]
> <mailto:[EMAIL PROTECTED]>> wrote:
>
>
>
> On Jan 29, 2008 10:01 AM, Don Dailey <[EMAIL PROTECTED]
> <mailto:[EMAIL PROTECTED]>> wrote:
>
>
terry mcintyre wrote:
> UCT is based on a theory of a multi-armed bandit, with uncertain knowledge
> about which "arms" would be most productive. Is it possible to graft various
> sources of knowledge into a sort of meta-bandit algorithm?
>
> As to fusing top-level knowledge with random playout
Andy wrote:
> See below I created a table that shows the transformation from KGS
> ratings to the Elo that CGOS uses. I set 6k=1800 because I believe
> that is what GNU 3.7.10 is under both systems. Does anyone have more
> data points for bots that play on both systems?
>
> Also is there an "al
> those choices, so simple algebraic combinations of the values
> does not in general work, even though it did when we were using
> only GNU Go engines. But one thing is reasonable obvious: if a
> bunch of completely different engines pick the same move, the
> confidence that it is
There is much to think about with Jason's and Michaels ideas.I favor
a more integrated approach than Michael suggests because I think it
would be very difficult to essentially have 2 different programs playing
the same game (ever play non-consultation doubles in chess or go? It's
fun but the
Sylvain,
These 2 parameters are confusing to me. --collectorLimitTreeSize
sounds like it limits the tree size to whatever your setting are, but
so does --limitTreeSize. What is the difference between a tree and
a collector tree?
I assume the collector is a garbage collector of some sor
rist hashes? just curious.
>
> :)
>
> s.
>
> - Original Message
> From: Don Dailey <[EMAIL PROTECTED]>
> To: computer-go
> Sent: Thursday, January 31, 2008 3:33:36 PM
> Subject: Re: [computer-go] 9x9 study rolloff
>
>
>
> Janzert wrote:
>
Janzert wrote:
> I haven't seen anyone else mention this, although I may have missed it
> in one of the previous discussions.
>
> I find it pretty amazing that both Mogo and Fatman are leveling off at
> exactly, or almost exactly, the same number of playouts (i.e. Fatman lvl
> 14 == Mogo lvl 18 =
ve black an edge. Black responds (sorry for anthropomorphism)
> by making distracting moves elsewhere on the board, deferring the
> situation to later plys. It's a bit like a ko fight with a huge supply
> of kos.
>
> - Dave Hillis
>
> -Original Message-
> From: Do
e rated game on KGS since Dec 2, 2007.
>
> On Jan 31, 2008 1:49 PM, Gian-Carlo Pascutto <[EMAIL PROTECTED]
> <mailto:[EMAIL PROTECTED]>> wrote:
>
> Don Dailey wrote:
>
> > I don't know how David figures 1000 ELO, but I would expect the
> >
That's a crazy looking graph!It looks like CrazyStone was close to
1d last spring, then a version change (perhaps) dropped it back to 2k.
Then suddenly in Dec (a new version?) it jumped suddenly to 1d then
gradually increased to 1.5d
- Don
Gian-Carlo Pascutto wrote:
> Do
ELO ratings don't have to be absolute, just self consistent. So if you
beat someone 7.2% of the time, that means you are about 440 ELO
stronger than him.
However, I don't understand what the K-factor has to do with anything.
scaling it up or down doesn't change anything. It's common prac
terry mcintyre wrote:
>
> I may have misunderstood, so please clarify a point.
>
> Let's say the game hinges on solving a life-and-death problem - if you find
> the right move, you win the game; if not, you lose. Many such problems - as
> found in real games - are extremely order-dependent; the
David Fotland wrote:
> UCT with light playouts that just avoid filling eyes is scalable, but much
> weaker than the strongest programs at 19x19 go.
>
> The strong programs have incorporated significant go knowledge, to direct
> the search to promising lines (usually local), to order moves to try,
Message-
>> From: [EMAIL PROTECTED] [mailto:computer-go-
>> [EMAIL PROTECTED] On Behalf Of Don Dailey
>> Sent: Thursday, January 31, 2008 8:33 AM
>> To: computer-go
>> Subject: Re: [computer-go] 9x9 Study
>>
>> I would like to see a very strong players
I would like to see a very strong players analysis of some of the games
of Mogo at the high levels in the study, but I am very leery of
subjective human analysis. Even though I would like to see it, I
would take what I heard with a grain of salt.
I hate to keep bringing this up, but t
>> You probably don't understand how UCT works. UCT balances exploration
>> with exploitation. The UCT tree WILL explore B1, but will explore it
>> with low frequency.That is unless the tree actually throws out 1
>> point eye moves (in which case it is not properly scalable and broken in
>
We want a version simply for the study - it may not make a performance
difference and will probably hurt the performance for a given time
level, so I would suggest it not to be the primary version.
- Don
Sylvain Gelly wrote:
> No problem for me. I did not want to multiply the number of versio
terry mcintyre wrote:
> Earlier Don Dailey asked how much of a difference it would make, if UCT
> programs understood nakade plays.
>
But actually they already understand nakade play. It was a
misconception that they don't, and I at first believed it because I
didn't
te
> education. It has been discovered that the best way to insure implicit
> obedience is to commence tyranny in the nursery.”
>
> Benjamin Disraeli, Speech in the House of Commons [June 15, 1874]
>
> - Original Message
>
>> From: Don Dailey <[EMA
credi 30 janvier 2008, Don Dailey a écrit :
>
>> I must not understand the problem. My program has no trouble with
>> nakade unless you are talking about some special case position.My
>> program immediately places the stone on the magic square to protect it's
The scalability didn't consider THAT factor. Ha Ha Ha, got you on that
one! I know that's pretty silly, but it illustrates the
frustration of this conversation to me.
- Don
Gian-Carlo Pascutto wrote:
> Don Dailey wrote:
>>> I am concerned that the current study is, as J
>
> Regardless of the exact example, _if_ pruning rules exclude a move,
> then an engine will never play it. That means that for that
> situation, they're not scalable. That may be a big if but will
> definitely affect some bot implementations. Progressive widening and
> soft-pruning rules prob
Heikki Levanto wrote:
> On Wed, Jan 30, 2008 at 03:23:35PM -0500, Don Dailey wrote:
>
>> Having said that, I am interested in this. Is there something that
>> totally prevents the program from EVER seeing the best move?I don't
>> mean something that takes a
are anecdotes of really strong players who didn't have to work at
it, but they are anecdotes, not realities.You can be sure than any
incredibly strong player put in the time. Some I'm sure had to put in
more than others, but none of them are strangers to focused intense study
Vlad Dumitrescu wrote:
> Hi Don,
>
> On Jan 30, 2008 9:02 PM, Don Dailey <[EMAIL PROTECTED]> wrote:
>
>> According to Sensei's Library, nakade is:
>> It refers to a situation in which a group has a single large
>> internal, enclosed space th
nteraction of the pass rules and
the eye rule in the play-outs, but I'm not a very strong go player so I
would have to think about it.
- Don
Gian-Carlo Pascutto wrote:
> Don Dailey wrote:
>
>> So I think this is nakade.
>
> Yes. Leela 0.2.x would get it wrong [
but that's not relevant - all play-out
strategies introduce some bias.Uniformly random play-outs delay the
understanding of positions too for instance.
- Don
Gian-Carlo Pascutto wrote:
> Don Dailey wrote:
>
>> Yes, the tree generates pass moves and with 2 passes the game is
tly 3 points. My
example shows 4 empty points in a big eye but they have even bigger
examples.
So I think this is nakade.
- Don
Jason House wrote:
>
>
> On Jan 30, 2008 2:48 PM, Don Dailey <[EMAIL PROTECTED]
> <mailto:[EMAIL PROTECTED]>> wrote:
>
> So a
nd what is called a paternal government, there is found state
> education. It has been discovered that the best way to insure implicit
> obedience is to commence tyranny in the nursery.”
>
> Benjamin Disraeli, Speech in the House of Commons [June 15, 1874]
>
> - Original Mes
Gian-Carlo Pascutto wrote:
> Don Dailey wrote:
>> I must not understand the problem. My program has no trouble with
>> nakade unless you are talking about some special case position.My
>> program immediately places the stone on the magic square to protect it's
he nursery.”
>
> Benjamin Disraeli, Speech in the House of Commons [June 15, 1874]
>
> - Original Message
> From: Don Dailey <[EMAIL PROTECTED]>
> To: computer-go
> Sent: Wednesday, January 30, 2008 11:10:01 AM
> Subject: Re: [computer-go] 19x19 Study - pr
Is nakade actually a problem in mogo? Are there positions it could
never solve or is merely a general weakness.
I thought the search corrected such problems eventually.
- Don
Gian-Carlo Pascutto wrote:
> Don Dailey wrote:
>
>> If a nakade fixed version of mogo (that is tr
I changed bayeselo to use the prior command as Rémi suggested I could do.
It raised the ELO rating of the highest rated well established player by
about 60 ELO!
I set prior to 0.1
http://cgos.boardspace.net/study/
- Don
Rémi Coulom wrote:
> Don Dailey wrote:
>> They seem under
>
> I am concerned that the current study is, as Jacques has so ably described, a
> study of a restricted game where nakade and certain other moves are
> considered to be illegal; this restricted game approaches the game of Go, but
> the programs have certain blind spots which humans can and do
Hi Olivier,
Yes, that would be great. Please do. Also, is there a Mac version
of this? We have the possiblity of using a huge cluster of Mac
machines if we have a working binary. We could probably get you a
temporary account to build such a thing if you don't already have it.
- Don
That information is in the autotester, but it wouldn't be accurate on
the main page since many computers of various types and with various
loads are being used in the study.
I suppose one could argue that the total conglomerate average would be
reasonably accurate according to the law of averag
Jacques Basaldúa wrote:
> Dave Hillis wrote:
>
> > I've noticed this in games on KGS; a lot of people lose games
> > with generous time limits because they, rashly, try to keep up
> > with my dumb but very fast bot and make blunders.
>
> What Don says about humans scaling applies to humans making
because mature players
won't fall for this. I would implement this by just forcing the move
to take at least 10 seconds.
- Don
[EMAIL PROTECTED] wrote:
>
> > From: Don Dailey <[EMAIL PROTECTED]>
> > ...
> > > Rémi Coulom wrote:
> > > ...
> > >
I wish I knew how that translates to win expectancy (ELO rating.)Is
3 kyu at this level a pretty significant improvement?
- Don
Hiroshi Yamashita wrote:
>> Instead of playing UCT bot vs UCT bot, I am thinking about running a
>> scaling experiment against humans on KGS. I'll probably start
vidence on this but it's worthwhile. I hope Rémi
decides to do this study.
- Don
Rémi Coulom wrote:
> Don Dailey wrote:
>> They seem under-rated to me also. Bayeselo pushes the ratings together
>> because that is apparently a valid initial assumption. With enough
Hiroshi Yamashita wrote:
>> What are the time controls for the games?
>
> Both are 10 minutes + 30 seconds byo-yomi.
>
> Hiroshi Yamashita
Good. I think that is a good way to test.
- Don
>
>
> ___
> computer-go mailing list
> computer-go@computer
What are the time controls for the games?
- Don
Hiroshi Yamashita wrote:
>> Instead of playing UCT bot vs UCT bot, I am thinking about running a
>> scaling experiment against humans on KGS. I'll probably start with
>> 2k, 8k, 16k, and 32k playouts.
>
> I have a result on KGS.
>
> AyaMC 6k (5.9k
Jeff Nowakowski wrote:
> On Tue, 2008-01-29 at 17:41 -0500, Don Dailey wrote:
>
>> This is in response to a few posts about the "self-test" effect in ELO
>> rating tests.
>>
> [...]
>
>> So my assertion is that scalability based on soun
Rémi Coulom wrote:
> Don Dailey wrote:
>> They seem under-rated to me also. Bayeselo pushes the ratings together
>> because that is apparently a valid initial assumption. With enough
>> games I believe that effect goes away.
>>
>> I could test that theory wit
a few
hundred samples we have anything to worry about.
- Don
Don Dailey wrote:
> They seem under-rated to me also. Bayeselo pushes the ratings together
> because that is apparently a valid initial assumption. With enough
> games I believe that effect goes away.
>
> I could test t
Yes, you are right. But gnugo's 1800 rating is the only real point of
reference that I have. As you get farther away from 1800 I believe
it's the case that the "true" rating can be sloppy.
- Don
Sylvain Gelly wrote:
>
> between pairs
> of programs, you can get a more and more conf
This is in response to a few posts about the "self-test" effect in ELO
rating tests.
I'll start by claiming right up front that I don't believe, for certain
types of programs, that this is something we have to worry unduly
about. I'll explain why I feel that way in a moment.
One general obser
901 - 1000 of 2315 matches
Mail list logo