Re: [Computer-go] News on Tromp-Cook ?

[valkyria]

I think first program who really used "3x3" patterns and other  
heuristics to make the simulations heavy was Viking 5. Unfortunately  
Viking5 was based on on Viking4 which was not designed for MC-search  
at all. I also had a completely different search method. Alpha-Beta  
search with Lazy MC-evaluation. This worked as follows. Instead of a  
normal evaluation function, something like 1000 simulations was run  
for a leaf position. When moves were evaluated safely below the alpha  
value it would stop search after 50 simulation. Similar for beta  
cutoffs (this is why I called it Lazy MC-evaluation).


It worked quite well. But when Mogo came around it quickly became  
stronger than Viking 5. And thus I started programming Valkyria. See  
my log for Viking5 here:


http://senseis.xmp.net/?Viking

I never published any papers about my programs, but I think I at least  
have made it clear to people reading this list that it is possible to  
use a lot of knowledge in playouts and benefit from which I guess all  
programs do.


-Best
Magnus

Quoting Fuming Wang :


This is certainly a good time to sit back and look at what got us here. The
following key ideas have been mentioned so far: UCB, MCTS, RAVE, Pattern and
Go knowledge during MC simulation.These ideas are all essential to a strong
MC based Go program.If we want to pick the most important idea that got us
here, I would say it is the realization that adding Go Pattern and Go
Knowledge to MC simulation can significantly improve the quality of board
evaluation. This is amount to the important sampling concept in MC
integration, which is very import for Monte Carlo applications in many
fields. MC simulation with importance sampling give us for the first time a
reasonablly accurate evaluation function for Go. UCB, MCTS, RAVE are
certainly very important, however, it is still possible that new approaches
that can achieve good results with just importantly samples MC simulation.
So, I think MoGo is the most important break-through.

Happy New Year, everyone!
Fuming



On Fri, Dec 31, 2010 at 9:20 AM, Aja  wrote:


Hi Jeff,

When, do you think, did Mogo "started dominating all the KGS computer
events and CGOS, and also was the first to extend that dominance from 9x9 to
19x19."?

In Computer Olympiad 2007, Steenvreter was gold medal on 9x9. At the final
match of 19x19, it's easily to see that Mogo and Crazy Stone were close
(finally Mogo 1st and CS 2ed). But, at the end of 2007, Crazy Stone defeated
Mogo and won the UEC Cup (19x19). Afterwards, Many Faces won 9x9 and 19x19
on 2008. Zen and Erica won 2009 and 2010, both continuing Crazy Stone
thread.

Mogo's biggest contributions, so far, in my view, are
1.Applied UCT to computer Go, and such application came from the idea
"MCTS" that proposed in 2006 by Remi Coulom. Crazy Stone was using MCTS to
win 9x9 in 2006 Computer Olympiad.
2.See 3x3 patterns around the previous move.
3.RAVE (strictly speaking, it is invented by David Silver).

UCT and RAVE are for both for the tree search. I think Crazy tone's
contribution for the playout is of same/or more important, because the
quality of simulations decide the playing strength much. From this view, we
should give Crazy Stone more and more credit.

I don't mean to raise any debate. Mogo does has important contributions,
but it's not so hard to assign credit to Crazy Stone. By the way, we should
not forget Fuego and MyGoFriend. Anyway, I think SenSei's description is
out-of-date.

Aja


-- From: Jeff Nowakowski
Sent: Friday, December 31, 2010 5:43 AM
To: computer...@computer-go.org

Subject: Re: [Computer-go] News on Tromp-Cook ?

On 12/30/2010 01:58 PM, David Fotland wrote:


You should also give more credit to CrazyStone as an early strong program
that contributed many ideas, comparable to Mogo.  Remi is Aja's advisor,
so
Erica continues the CrazyStone thread.



I did mention CrazyStone, and the Sensei's page lists it first as the
program that "started the new wave of MCTS programs by winning the 9x9
gold medal at the ICGA Computer Olympiad, in 2006."  Like I said in my
first message, though, it's hard to assign credit, and I don't mean to
slight other programs.

However, MoGo was the program that really got people to sit up and take
notice, because it started dominating all the KGS computer events and
CGOS, and also was the first to extend that dominance from 9x9 to 19x19.
I believe the biggest breakthroughs were made with MoGo (building, of
course, on earlier ideas). This is easily verified by going back to the
archives and seeing how many people patterned their program after MoGo.

___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mail

Re: [Computer-go] News on Tromp-Cook ?

[Olivier Teytaud]

I cannot speak for the Mogo team, but it is clear to me that the idea
> was simply 'out there' when Mogo started and they had access to it
> through the various research communities (Machine learning, Computer
> games, etc.). I guess at least one of their supervisors most have had
> some contacts with Levente or Csaba. Levente did not discuss
> UCB-tuned, so they might have gotten that directly from Auer.
>
>
It is very clear that we had contacts with UCB-people and UCT-people, in
particular through Rémi Munos (coauthor of
BAST, a variant of bandit-based Monte-Carlo Planning).
We had also a lot of discussions with Rémi Coulom.
Olivier
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Erik van der Werf]

On Sun, Jan 2, 2011 at 1:20 AM, Aja  wrote:
> Do you mean the whole MCTS scheme combined with UCB formula
> proposed by Mogo is completely inspried by Levente's work?

I cannot speak for the Mogo team, but it is clear to me that the idea
was simply 'out there' when Mogo started and they had access to it
through the various research communities (Machine learning, Computer
games, etc.). I guess at least one of their supervisors most have had
some contacts with Levente or Csaba. Levente did not discuss
UCB-tuned, so they might have gotten that directly from Auer.

Erik
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Stefan Kaitschick]

Am 02.01.2011 11:22, schrieb Darren Cook:

In your survey, the spread for a super-human program, from those that
correctly predicted 2010 for shodan,
is from 2023 to 2150.
So even between the best predictors sofar, there was huge disagreement
when it comes topling humanity ...
I guess current knowledge of the effectiveness and scalability of MCTS
would narrow the spread, but how much?

This was the other poll I set up; with the recent attention the number
of votes jumped a lot, and (surprisingly to me) the most popular choice
is now 2015 to 2020:
   http://polldaddy.com/poll/1614045/?view=results

Darren
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go


Thx.
Funny, the "sweet spot" from 2030 to 2050 only gets 12%.
It seems that the word is mainly populated with optimists and pessimists.
Or maybe the're all optimists, they only have different hopes. :-)

Stefan
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Darren Cook]

> In your survey, the spread for a super-human program, from those that
> correctly predicted 2010 for shodan,
> is from 2023 to 2150.
> So even between the best predictors sofar, there was huge disagreement
> when it comes topling humanity ...
> I guess current knowledge of the effectiveness and scalability of MCTS
> would narrow the spread, but how much?

This was the other poll I set up; with the recent attention the number
of votes jumped a lot, and (surprisingly to me) the most popular choice
is now 2015 to 2020:
  http://polldaddy.com/poll/1614045/?view=results

Darren
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Olivier Teytaud]

Thanks a lot. The state-of-the-art part of Compuetr Go in my thesis will be
> more accurate. Do you mean the whole MCTS scheme combined with UCB formula
> proposed by Mogo is completely inspried by Levente's work? If I understand
> Remi's paper correctly, Remi can change Crazy Stone's MCTS to Mogo's one by
> 10 mins work (including just change the selection formula to UCB)...
>
>
That's very reasonnable, and I'm sure that adding a UCB term does not make
CrazyStone stronger (or by a negligible amount).

The big influence of CrazyStone on MoGo is the MCTS principle.
The real influence of MoGo on CrazyStone is in the MC part and the RAVE
part.

Best regards,
Olivier
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Olivier Teytaud]

For the olympiads, MoGo has been adapted in a hurry by the operator for the
NNGS
server - this was not that difficult, but in the hurry there
were mistakes, so that mogo did not see the clock and was playing with a bad
clock.
I have no idea of the impact of this on the game.

In 13x13 I'm not sure mogo lost a single game on the board, maybe only games
by time troubles - but I'm not sure of this, and people interested in that
can check the olympiads logs - I have not enough motivation :-)

As for UEC Cup using the Japanese rules, in my judgment, Mogo lost badly to
> Crazy Stone even with Chinese rules
> http://jsb.cs.uec.ac.jp/~igo/past/2007/sgf/1201/mogo-crazystone.sgf
> I am curious if Mogo was weaker because of not properly dealing with
> Japanese rules.
>
>
I think nobody of MoGo's team was at the UEC Cup. I know that in some UEC
Cup
we loose games whereas we had won on the board, due to net lag (a UEC cup
for which mogo was on a remote computer).
Please note that the net lag can have an impact either by making a bot loose
by time, or by making a bot play moves in 10s instead of 40s.

No idea for other games than the game lost by net lag - if it only depended
on me, mogo would not play the UEC cup, and I decided
not to take care of anything around that. UEC Cup is perfectly organized,
but I have no motivation for computer competitions, and I will not take care
of anything else than chinese rules / KGS. Other members of the
team can do whatever they want and they can reply themselves :-)

Best regards,
Olivier
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Michael Williams]

Usually "AMAF" refers to an engine that does not build a tree.


On Sat, Jan 1, 2011 at 7:20 PM, Aja  wrote:
> Hi Erik,
>
> Thanks a lot. The state-of-the-art part of Compuetr Go in my thesis will be
> more accurate. Do you mean the whole MCTS scheme combined with UCB formula
> proposed by Mogo is completely inspried by Levente's work? If I understand
> Remi's paper correctly, Remi can change Crazy Stone's MCTS to Mogo's one by
> 10 mins work (including just change the selection formula to UCB)...
>
> Aja
>
> - Original Message - From: "Erik van der Werf"
> 
> To: 
> Sent: Sunday, January 02, 2011 6:11 AM
> Subject: Re: [Computer-go] News on Tromp-Cook ?
>
>
> On Fri, Dec 31, 2010 at 1:37 PM, Jeff Nowakowski  wrote:
>>>
>>> In Computer Olympiad 2007, Steenvreter was gold medal on 9x9.
>
> ...
>>
>> Obviously it was following MoGo's lead with UCT (the tournament was held
>> in
>> June 2007, well after the remarkable success of MoGo). I don't mean to
>> discredit Steenvreter, CrazyStone, or any other program. I'm just focusing
>> on MoGo so much because it set the bar so high and got everybody chasing
>> it.
>
> Not exactly; I already knew UCT before MoGo even existed. Levente's
> ecml paper was not the first one on UCT. He had already submitted
> other work (*) where he specifically applied UCT to the domain of
> computer games. At the time I was even considering to add UCT to Magog
> and have a nice reunion at the 2006 Olympiad in Turin (my former
> colleagues Levente Kocsis and Mark Winands were also co-authors of
> Magog). Unfortunately then some conference organizer screwed up...
>
>
>>> Mogo's biggest contributions, so far, in my view, are
>>> 1.Applied UCT to computer Go, and such application came from the idea
>>> "MCTS" that proposed in 2006 by Remi Coulom.
>
> No, it came directly from Levente. Several people got access to his
> paper around the same time as Remi's paper. It is more like they
> independently proposed similar ideas.
>
> Erik
>
>
> * "Improved Monte-Carlo Search" by Levente Kocsis, Csaba Szepesvári
> and Jan Willemson
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Aja]

Hi Erik,

Thanks a lot. The state-of-the-art part of Compuetr Go in my thesis will be 
more accurate. Do you mean the whole MCTS scheme combined with UCB formula 
proposed by Mogo is completely inspried by Levente's work? If I understand 
Remi's paper correctly, Remi can change Crazy Stone's MCTS to Mogo's one by 
10 mins work (including just change the selection formula to UCB)...


Aja

- Original Message - 
From: "Erik van der Werf" 

To: 
Sent: Sunday, January 02, 2011 6:11 AM
Subject: Re: [Computer-go] News on Tromp-Cook ?


On Fri, Dec 31, 2010 at 1:37 PM, Jeff Nowakowski  wrote:

In Computer Olympiad 2007, Steenvreter was gold medal on 9x9.

...
Obviously it was following MoGo's lead with UCT (the tournament was held 
in

June 2007, well after the remarkable success of MoGo). I don't mean to
discredit Steenvreter, CrazyStone, or any other program. I'm just focusing
on MoGo so much because it set the bar so high and got everybody chasing 
it.


Not exactly; I already knew UCT before MoGo even existed. Levente's
ecml paper was not the first one on UCT. He had already submitted
other work (*) where he specifically applied UCT to the domain of
computer games. At the time I was even considering to add UCT to Magog
and have a nice reunion at the 2006 Olympiad in Turin (my former
colleagues Levente Kocsis and Mark Winands were also co-authors of
Magog). Unfortunately then some conference organizer screwed up...



Mogo's biggest contributions, so far, in my view, are
1.Applied UCT to computer Go, and such application came from the idea
"MCTS" that proposed in 2006 by Remi Coulom.


No, it came directly from Levente. Several people got access to his
paper around the same time as Remi's paper. It is more like they
independently proposed similar ideas.

Erik


* "Improved Monte-Carlo Search" by Levente Kocsis, Csaba Szepesvári
and Jan Willemson
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go 


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

Re: [Computer-go] News on Tromp-Cook ?

[Stefan Kaitschick]

In your survey, the spread for a super-human program, from those that 
correctly predicted 2010 for shodan,

is from 2023 to 2150.
So even between the best predictors sofar, there was huge disagreement 
when it comes topling humanity ...
I guess current knowledge of the effectiveness and scalability of MCTS 
would narrow the spread, but how much?


Stefan


Am 01.01.2011 21:46, schrieb Darren Cook:

Despite his loss of the bet on the surface, I congratulate Darren for
almost correctly predicting the 19x19 computer strength development!

Actually my own prediction in 1997 was hopelessly optimistic. Here is
the 1997 list:
   http://dcook.org/gobet/mail.19970901.txt

As you can see congratulations should go to Shinichi Sei, Tristan
Cazenave, David Fotland, Yung Jye Hunag, Jun Saito and Kobayashii, who
all guessed 2010 for European shodan level (*). (The first four are
computer go programmers.)

And especially congratulations to John, who 13 years ago very cleverly
guessed he'd still be strong enough in 2010, but maybe not in 2011...

By the way, my page has been updated with results, and permanent store
of the game records:
http://dcook.org/gobet/

Of special interest to people on this list is the detailed log file for
the fourth game is available for download: it shows the prime variation,
winning percentage, alternative moves considered, number of playouts,
etc. at every move. Published with David Fotland's permission.

Darren

*: Unless you disagree and think the programs are still 1-kyu. But I'm
going by the aggregate of results, including the 3d and 4d rankings that
Many Faces and Zen consistently hold on KGS.
Also John can be described as a "strong" 2-dan.




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

Re: [Computer-go] News on Tromp-Cook ?

[Erik van der Werf]

On Fri, Dec 31, 2010 at 1:37 PM, Jeff Nowakowski  wrote:
>> In Computer Olympiad 2007, Steenvreter was gold medal on 9x9.
...
> Obviously it was following MoGo's lead with UCT (the tournament was held in
> June 2007, well after the remarkable success of MoGo). I don't mean to
> discredit Steenvreter, CrazyStone, or any other program. I'm just focusing
> on MoGo so much because it set the bar so high and got everybody chasing it.

Not exactly; I already knew UCT before MoGo even existed. Levente's
ecml paper was not the first one on UCT. He had already submitted
other work (*) where he specifically applied UCT to the domain of
computer games. At the time I was even considering to add UCT to Magog
and have a nice reunion at the 2006 Olympiad in Turin (my former
colleagues Levente Kocsis and Mark Winands were also co-authors of
Magog). Unfortunately then some conference organizer screwed up...


>> Mogo's biggest contributions, so far, in my view, are
>> 1.Applied UCT to computer Go, and such application came from the idea
>> "MCTS" that proposed in 2006 by Remi Coulom.

No, it came directly from Levente. Several people got access to his
paper around the same time as Remi's paper. It is more like they
independently proposed similar ideas.

Erik


* "Improved Monte-Carlo Search" by Levente Kocsis, Csaba Szepesvári
and Jan Willemson
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Darren Cook]

> Despite his loss of the bet on the surface, I congratulate Darren for
> almost correctly predicting the 19x19 computer strength development!

Actually my own prediction in 1997 was hopelessly optimistic. Here is
the 1997 list:
  http://dcook.org/gobet/mail.19970901.txt

As you can see congratulations should go to Shinichi Sei, Tristan
Cazenave, David Fotland, Yung Jye Hunag, Jun Saito and Kobayashii, who
all guessed 2010 for European shodan level (*). (The first four are
computer go programmers.)

And especially congratulations to John, who 13 years ago very cleverly
guessed he'd still be strong enough in 2010, but maybe not in 2011...

By the way, my page has been updated with results, and permanent store
of the game records:
   http://dcook.org/gobet/

Of special interest to people on this list is the detailed log file for
the fourth game is available for download: it shows the prime variation,
winning percentage, alternative moves considered, number of playouts,
etc. at every move. Published with David Fotland's permission.

Darren

*: Unless you disagree and think the programs are still 1-kyu. But I'm
going by the aggregate of results, including the 3d and 4d rankings that
Many Faces and Zen consistently hold on KGS.
Also John can be described as a "strong" 2-dan.


-- 
Darren Cook, Software Researcher/Developer

http://dcook.org/work/ (About me and my work)
http://dcook.org/blogs.html (My blogs and articles)
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Petr Baudis]

  Hi!

On Sat, Jan 01, 2011 at 02:39:46PM +, Nick Wedd wrote:
> I don't know what happened with pasky.  I noticed him there, and
> gave him permission to chat, but I did not see him use it.  I tried
> to chat directly to him, but received no response.

  Intriguing, I received no response either and could not chat inside.
I guess it must have been some technical glitch. Seeing that there was
a clear method behind giving out the permission to chat, I have much
fewer reservations.  :-)

Petr "Pasky" Baudis
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Nick Wedd]

In message <4d1c3938.1040...@snafu.de>, Robert Jasiek  
writes
Despite his loss of the bet on the surface, I congratulate Darren for 
almost correctly predicting the 19x19 computer strength development! It 
has been an extraordinarly impressive improvement during the last 3 
years! Before 19x19 was more like 10 kyu - now during parts of a game 
ManyFaces can hold 1d to 2d level! With some more programming effort 
for holding a program's playing strength at a constant level (maybe 
also by filtering computer suggested moves by a human approach bias 
filter to discard obviously bad moves like A15 in game 3 and by making 
endgame more expert-orientated again), this strength can soon be held 
during an entire game.


Nick has said that a 2007 Hungarian RAVE paper was the theoretical 
breakthrough. Is this its URL?


http://zaphod.aml.sztaki.hu/papers/ecml06.pdf

The site appears to be down though. Is there an alternative URL?


I don't know about "RAVE" - the paper I referred to is available at
http://www.lri.fr/~sebag/Examens_2008/UCT_ecml06.pdf

Nick

ManyFaces was described as an expert system. How does it work today? 
How does it use the modern algorithmic theories?


Congratulations also to all the theorists! Without their great 
discoveries, programs would still be weak. Might somebody please give 
an overview on the relevant theories and how they work?


One thing keeps bothering me though: What does all the strength 
improvement give us humans for better understanding the game strategy? 
Almost nothing? The information contained in the current calculation 
size is not easily translated to human applicable strategic / tactical 
knowledge. Other research, which is closer to the human way of go 
understanding, by people like Berlekamp, Spight, Cazenawe or myself is 
much more useful for players but its playing strength equivalent - 
despite a few 10p knowledge exceptions - is still on the 20k level. 
Currently there is an extreme gap between computer go theory making 
computers strong, maths theory explaining go theory for human 
understanding and traditional professional go theory, which fails to 
explain well but allows eager and gifted players to succeed by means of 
unlimited investment of time and effort. What is still mostly missing 
are ways to link well to each other the three major paths towards great 
playing strength.


- How can programs learn well from professional knowledge?
- How can programs use well mathematical descriptions of human-like 
strategy?

- How can players learn well from strong programs?
- How can further mathematical descriptions of human-like strategy be 
derived from strong computer play or its underlying algorithms?


Oh, and of course congratulations to John!



--
Nick Weddn...@maproom.co.uk
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Nick Wedd]

In message 
, steve 
uurtamo  writes

in the last game, a comment i especially noted was that miai was
handled poorly by the computer player and was a seemingly effective
strategy for playing against a computer, since there were many miai
that john left for later. once two fights got close enough to affect
one another, the miai got more complicated, because john could
seemingly (locally) sacrifice groups that were still just fine. on a
positive note, MFoG seemed to do a good job at working to keep sente.

on the question of who was allowed to comment and who was not -- in
the first game, the rule as stated was that only dan-level players
would be given permission to comment. then a 2kyu player was added
into the mix, but the level of conversation was indeed restrained by
the fact that it was only 4-5 people giving most of the commentary.


As the person relaying the games (as 'BGAmatches'), I can expand 
slightly on this.


I gave "permission to chat" to
  everyone rated 1d or better whom I noticed (I'm sure I missed some)
  everyone who asked
  a few other people whom I knew were sensible

I don't know what happened with pasky.  I noticed him there, and gave 
him permission to chat, but I did not see him use it.  I tried to chat 
directly to him, but received no response.  I noticed that chid0ri was 
in another room, drew her attention to a challenge game, and gave her 
permission to chat in it (I don't think she used it).


As for my reasons for using "nazi mode"  -
In the past I have done relays of British championship matches.  These 
don't attract as many viewers as this week's challenge match did, only 
around 200.  But I have found that if I leave the chat open to all, most 
of the chat is from weak players with little or no interest in the game. 
This makes it difficult for serious kibitzers to follow the sensible 
comments in the chat.  I consider that 'gogonuts' and other strong 
kibitzers performed a very valuable service by giving their views of 
what was happening in the games, and it would have been a pity to dilute 
this with a lot of childish prattle.


Nick













s.

On Thu, Dec 30, 2010 at 3:14 AM, terry mcintyre 
 wrote:

In one of the games, "gogonuts" - who has played many games with computer
programs - opined that multiple unresolved semeai are a weakness of 
MC programs.


A strong human would reason that if the outcome of semeai A, B, and C are
individually losses for the program, then the aggregate probability 
is to lose

all three fights.
 Terry McIntyre 


Unix/Linux Systems Administration
Taking time to do it right saves having to do it twice.



- Original Message 

From: Mark Boon 
To: computer-go@dvandva.org
Sent: Wed, December 29, 2010 6:25:14 PM
Subject: Re: [Computer-go] News on Tromp-Cook ?

On Wed, Dec 29, 2010 at 12:47 PM, Jacques Basaldúa 
 wrote:

> I don't  agree the first two games were that easy.
>
> In the second game the  bot was ahead most of the game
> and failed in life and death in the top  right corner.

Yes, we disagree. The first game was finished before it had  well
begun. By move 50 or so.

The second game the computer was also  doing poorly until John got
careless about his left-side group. The computer  played that well and
took the lead. But only briefly, as it had the top-right  corner killed
immediately after. So I wouldn't say the computer was ahead  most of
the game. It was ahead for only a very brief moment.

But  opinions about games can differ of course. Generally I saw the
computer do  some impressive things and some very silly things.

     Mark
___
Computer-go mailing  list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go





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


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


--
Nick Weddn...@maproom.co.uk
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Ingo Althöfer]

 Original-Nachricht 
> Datum: Fri, 31 Dec 2010 21:26:45 +0100
> Von: Olivier Teytaud 
>
> ...
> Also, there are contributors to MCTS older than MCTS - Monte-Carlo people
> (Cazenave, Bouzy...) and people using
> tree exploration in planning (Péret Garcia is one of my favorite
> references); also, quiescent search in alpha-beta,
> iterative-deepening. 


> Finding one and only one source is a mistake.

Amen. Amen. Amen!

Best regards,
Ingo.

PS for Olivier: Prof. Dr. Rudolf Ahlswede passed away on December 18.
-- 
NEU: FreePhone - kostenlos mobil telefonieren und surfen!   
Jetzt informieren: http://www.gmx.net/de/go/freephone
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Aja]

Hi Olivier,

I am surprised that you said “mogo was not prepared for the NNGS server”. Does 
that mean Mogo will be of no problem if the Olympiad tournament was on KGS? 

In olympiad Mogo lost to Zen and mfgo, even without timeout, in my judgment. 
But Mogo used a single machine (16 cores, far from in full strength) to play 
with Erica. So that game explains not much.

As for UEC Cup using the Japanese rules, in my judgment, Mogo lost badly to 
Crazy Stone even with Chinese rules
http://jsb.cs.uec.ac.jp/~igo/past/2007/sgf/1201/mogo-crazystone.sgf
I am curious if Mogo was weaker because of not properly dealing with Japanese 
rules. 

Aja

From: Olivier Teytaud 
Sent: Saturday, January 01, 2011 3:49 AM
To: Aja ; computer-go 
Subject: Re: [Computer-go] News on Tromp-Cook ?


  In Computer Olympiad 2007, Steenvreter was gold medal on 9x9. At the final 
match of 19x19, it's easily to see that Mogo and Crazy Stone were close 
(finally Mogo 1st and CS 2ed). But, at the end of 2007, Crazy Stone defeated 
Mogo and won the UEC Cup (19x19). Afterwards, Many Faces won 9x9 and 19x19 on 
2008. Zen and Erica won 2009 and 2010, both continuing Crazy Stone thread.


MoGo has never been configured for japanese rules, which makes UEC cup not 
suitable for MoGo.
Some students wanted mogo to play this competiton anyway, I was not in favor of 
that but I don't decide :-)

For 2010 olympiads, mogo was not prepared for the NNGS server (and, if this was 
depending only on me, mogo would not have participated - we accepted to play 
whereas we knew mogo does not accept NNGS, as requested by the organizers, but 
I know that bad results can always be cited later on whenever they don't show 
anything about the strength of programs). 

Basically, I'm not very interested in competitions between computers, which 
take too much time and don't stay in history - I prefer to take care of games 
against humans. I was nonetheless operating myself mogo sometimes in 
competition (and I think there are not many people who know sufficiently the 
details of the code for operating mogo in a competition, unfortunately),
and MoGo won against Zen in TAAI, and won against Erica in the past (Erica 
probably also won games against MoGo sometimes). 
I think there's little difference between the strong programs.

This is due to us, for sure - we decided not to work on anything else thank 
KGS. 

Best regards, happy new year,
Olivier
 
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Aja]

Hi Olivier,

I fully agree with your words. We were talking exclusively about the 
breakthrough of MCTS and UCT. Or I think everyone must admit that you, Yizao, 
Remi...etc, all of us, stand in the shoulder of the “giants” to make progress. 
I was in a team of Chinese chess program for over 3 years. So, I understand 
there was many great and BIG inventions before, especially in the field of 
computer chess.

I also don’t think the few games in Olympiad can decide for sure that Zen, mfgo 
or Erica are REALLY stronger than Mogo. The difference is quite small as your 
said. We would need a lot of games to verify that. But, at least, we are sure 
that it’s really too hard for a single program, for now, to “dominating all the 
KGS computer events and CGOS, and also was the first to extend that dominance 
from 9x9 to 19x19” in a long period. What Jeff meant was that several months 
when Mogo was the solid strongest. I didn’t mean to say Mogo is weaker, but 
mean to say Mogo was not dominating since Computer Olympiad 2007. To me, Mogo’s 
dominance in that several months is not so special. Maybe it’s because I was 
working hard for a long time to overthrow Zen’s dominance, rather than Mogo’s.

Aja


From: Olivier Teytaud 
Sent: Saturday, January 01, 2011 4:26 AM
To: Aja ; computer-go 
Subject: Re: [Computer-go] News on Tromp-Cook ?

... basically I think that there would be not so much work around MCTS without 
the MC part with a strong influence
of the last move. This is probably the main contribution of MoGo, and *all* 
computer-go MCTS are influenced by this, I think.

Most importantly, MCTS is very efficient, for many things, beyond Go. Other 
games, and industrial applications.
All contributors of MCTS have this success, collectively.

Also, there are contributors to MCTS older than MCTS - Monte-Carlo people 
(Cazenave, Bouzy...) and people using
tree exploration in planning (Péret Garcia is one of my favorite references); 
also, quiescent search in alpha-beta,
iterative-deepening. Finding one and only one source is a mistake.

Best regards,
Olivier
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Olivier Teytaud]

... basically I think that there would be not so much work around MCTS
without the MC part with a strong influence
of the last move. This is probably the main contribution of MoGo, and *all*
computer-go MCTS are influenced by this, I think.

Most importantly, MCTS is very efficient, for many things, beyond Go. Other
games, and industrial applications.
All contributors of MCTS have this success, collectively.

Also, there are contributors to MCTS older than MCTS - Monte-Carlo people
(Cazenave, Bouzy...) and people using
tree exploration in planning (Péret Garcia is one of my favorite
references); also, quiescent search in alpha-beta,
iterative-deepening. Finding one and only one source is a mistake.

Best regards,
Olivier
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[terry mcintyre]

It would help immensely if the various servers would agree on an interoperable 
standard.
 Terry McIntyre 


Unix/Linux Systems Administration
Taking time to do it right saves having to do it twice.


>
>From: Olivier Teytaud 
>To: Aja ; computer-go 
>Sent: Fri, December 31, 2010 2:49:53 PM
>Subject: Re: [Computer-go] News on Tromp-Cook ?
>
>
>
>In Computer Olympiad 2007, Steenvreter was gold medal on 9x9. At the final 
>match 
>of 19x19, it's easily to see that Mogo and Crazy Stone were close (finally 
>Mogo 
>1st and CS 2ed). But, at the end of 2007, Crazy Stone defeated Mogo and won 
>the 
>UEC Cup (19x19). Afterwards, Many Faces won 9x9 and 19x19 on 2008. Zen and 
>Erica 
>won 2009 and 2010, both continuing Crazy Stone thread.
>>

MoGo has never been configured for japanese rules, which makes UEC cup not 
suitable for MoGo.
Some students wanted mogo to play this competiton anyway, I was not in favor of 
that but I don't decide :-)

For 2010 olympiads, mogo was not prepared for the NNGS server (and, if this was 
depending only on me, mogo would not have participated - we accepted to play 
whereas we knew mogo does not accept NNGS, as requested by the organizers, but 
I 
know that bad results can always be cited later on whenever they don't show 
anything about the strength of programs). 


Basically, I'm not very interested in competitions between computers, which 
take 
too much time and don't stay in history - I prefer to take care of games 
against 
humans. I was nonetheless operating myself mogo sometimes in competition (and I 
think there are not many people who know sufficiently the details of the code 
for operating mogo in a competition, unfortunately),
and MoGo won against Zen in TAAI, and won against Erica in the past (Erica 
probably also won games against MoGo sometimes). 

I think there's little difference between the strong programs.

This is due to us, for sure - we decided not to work on anything else thank 
KGS. 


Best regards, happy new year,
Olivier


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

Re: [Computer-go] News on Tromp-Cook ?

[Olivier Teytaud]

> In Computer Olympiad 2007, Steenvreter was gold medal on 9x9. At the final
> match of 19x19, it's easily to see that Mogo and Crazy Stone were close
> (finally Mogo 1st and CS 2ed). But, at the end of 2007, Crazy Stone defeated
> Mogo and won the UEC Cup (19x19). Afterwards, Many Faces won 9x9 and 19x19
> on 2008. Zen and Erica won 2009 and 2010, both continuing Crazy Stone
> thread.
>

MoGo has never been configured for japanese rules, which makes UEC cup not
suitable for MoGo.
Some students wanted mogo to play this competiton anyway, I was not in favor
of that but I don't decide :-)

For 2010 olympiads, mogo was not prepared for the NNGS server (and, if this
was depending only on me, mogo would not have participated - we accepted to
play whereas we knew mogo does not accept NNGS, as requested by the
organizers, but I know that bad results can always be cited later on
whenever they don't show anything about the strength of programs).

Basically, I'm not very interested in competitions between computers, which
take too much time and don't stay in history - I prefer to take care of
games against humans. I was nonetheless operating myself mogo sometimes in
competition (and I think there are not many people who know sufficiently the
details of the code for operating mogo in a competition, unfortunately),
and MoGo won against Zen in TAAI, and won against Erica in the past (Erica
probably also won games against MoGo sometimes).
I think there's little difference between the strong programs.

This is due to us, for sure - we decided not to work on anything else thank
KGS.

Best regards, happy new year,
Olivier
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Fuming Wang]

That was such an amazingly crazy period. I am sure some day, some one will
write a book (or two) about all of this, so that latecomers get to know all
the details.

Best Wishes!
Fuming

On Sat, Jan 1, 2011 at 12:22 AM, "Ingo Althöfer" <3-hirn-ver...@gmx.de>wrote:

> > ... but
> > championship events are relatively poor predictors of skill because of
> > their limited number of sample points. something like cgos ranking
> > over time (among those who participate) is a pretty good way to
> > compare computer go playing programs.
>
> Both types of competition do have their justification.
> It is similar to sports:
> the one with the best average 100 meter speed (taking over the summer)
> will be top in a list, and the winner of the olympic gold medal
> will have lots of prestige.
>
> At least some programmers use CGOS and KGS simply for testing
> this and that, and only at the big events (like computer olympiads)
> their seemingly strongest version (including opening book, pattern
> data bases and so on) is presented.
>
> My list was meant in a slightly different sense:
> MoGo as a Go program simply was not there in 2006,
> when Crazy Stone started to run up the charts.
> And it is typically easier (also psychologically)
> to attack some goal when someone else has shown before
> that it is in principle possible to achieve.
>
> Happy new year,
> Ingo.
>
>
> --
> GMX DSL Doppel-Flat ab 19,99 Euro/mtl.! Jetzt auch mit
> gratis Notebook-Flat! http://portal.gmx.net/de/go/dsl
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Ingo Althöfer]

> ... but
> championship events are relatively poor predictors of skill because of
> their limited number of sample points. something like cgos ranking
> over time (among those who participate) is a pretty good way to
> compare computer go playing programs.

Both types of competition do have their justification.
It is similar to sports:
the one with the best average 100 meter speed (taking over the summer)
will be top in a list, and the winner of the olympic gold medal
will have lots of prestige.

At least some programmers use CGOS and KGS simply for testing
this and that, and only at the big events (like computer olympiads)
their seemingly strongest version (including opening book, pattern
data bases and so on) is presented.

My list was meant in a slightly different sense:
MoGo as a Go program simply was not there in 2006,
when Crazy Stone started to run up the charts.
And it is typically easier (also psychologically)
to attack some goal when someone else has shown before
that it is in principle possible to achieve.

Happy new year,
Ingo.


-- 
GMX DSL Doppel-Flat ab 19,99 Euro/mtl.! Jetzt auch mit 
gratis Notebook-Flat! http://portal.gmx.net/de/go/dsl
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[steve uurtamo]

i think that don has best made this point in the past[1], but
championship events are relatively poor predictors of skill because of
their limited number of sample points. something like cgos ranking
over time (among those who participate) is a pretty good way to
compare computer go playing programs.

s.

[1] sorry if this is a misattribution.
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Ingo Althöfer]

Hello all,


> ... I really don't want to raise any debate in such a happy 
> new year time. :)


In central Europe it is still old year for many hours.
So, no  direct reason to become romantic.
 
 
Here are some facts from the ICGA
Computer Olympiads in 2005, 2006, 2007.

September 2005 Taipei
19x19 http://www.grappa.univ-lille3.fr/icga/tournament.php?id=42
9x9 http://www.grappa.univ-lille3.fr/icga/tournament.php?id=11
On 9x9, Indigo got bronze medal.
On 19x19, Indigo made shared rank 4-6 from 7 starters.
Was "Indigo" MC-based in that event? (I think so.)

May/June 2006 Turino
19x19 http://www.grappa.univ-lille3.fr/icga/tournament.php?id=3
9x9 http://www.grappa.univ-lille3.fr/icga/tournament.php?id=1
Crazy Stone won gold medal on 9x9, and ranked
second last on 19x19.
Crazy Stone was MC-based in that event.

June 2007 Amsterdam
19x19 http://www.grappa.univ-lille3.fr/icga/tournament.php?id=167
9x9 http://www.grappa.univ-lille3.fr/icga/tournament.php?id=169
On 19x19 MoGo got gold, Crazy Stone silver, both MC-based.
On 9x9 Steenvreter got gold, MoGo silver, Crazy Stone bronze.

***
My Conclusion:
Crazy Stone was the first go bot to get a gold 
medal in Computer Olympiads.
MoGo was the first go bot to get a gold medal
on 19x19 in Computer Olympiads.
Likely (when my memory serves me well), Indigo
was the first go bot based on MC to get any medal
at the Computer Olympiads.

Ingo.
-- 
GMX DSL Doppel-Flat ab 19,99 Euro/mtl.! Jetzt auch mit 
gratis Notebook-Flat! http://portal.gmx.net/de/go/dsl
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Aja]

Hi Jeff,

Maybe I have to reviewed your messages to get your point. I really don't 
want to raise any debate in such a happy new year time. :)


Aja


Crazy Stone was strong "relatively" in that period. It's the same with
Mogo. I don't understand why you speically indicate the period when Mogo
was the strongest.


I spent too much time indicating why already. If it's not enough for you, 
then there is nothing else I can say.




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

Re: [Computer-go] News on Tromp-Cook ?

[Jeff Nowakowski]

On 12/31/2010 08:12 AM, Aja wrote:


Crazy Stone was strong "relatively" in that period. It's the same with
Mogo. I don't understand why you speically indicate the period when Mogo
was the strongest.


I spent too much time indicating why already. If it's not enough for 
you, then there is nothing else I can say.


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

Re: [Computer-go] News on Tromp-Cook ?

[Aja]

Hi Jeff,


Here I quote from the computer-go archives, unless otherwise noted:


Yes, I agree Mogo was so strong in that period (2006/12/31-2009/3/17 or 
more). I think it is just similar to the situation when Crazy Stone got gold 
medal in Computer Olympiad 2006. Can't we also say this?


Oh! Crazy Stone was so strong to get gold medal! How did Remi beat all the 
traditional programs? Remi was a such a new Go programmer!


Crazy Stone was strong "relatively" in that period. It's the same with Mogo. 
I don't understand why you speically indicate the period when Mogo was the 
strongest.


Yes, and this is what Eric had to say about it: "Steenvreter uses UCT and 
has some L&D knowledge that I reused from Magog."


and: "Steenvreter was really a rush job, hacking things together until the 
last day before the tournament and no time to test properly. I was hoping 
to be able to catch up with the stronger programs, but never expected it 
to win the tournament."


Oh... Erik really deserved the gold medal. :)

Obviously it was following MoGo's lead with UCT (the tournament was held 
in June 2007, well after the remarkable success of MoGo). I don't mean to 
discredit Steenvreter, CrazyStone, or any other program. I'm just focusing 
on MoGo so much because it set the bar so high and got everybody chasing 
it.


UCT is a selection strategy of MCTS to keep the balance between exploration 
and exploitation. Crazy Stone invented MCTS and Mogo brought UCT such a 
great application to it. I think they both make BIG contributions. That's 
all I want to say.


Aja


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

Re: [Computer-go] News on Tromp-Cook ?

[Jeff Nowakowski]

On 12/30/2010 08:20 PM, Aja wrote:

Hi Jeff,

When, do you think, did Mogo "started dominating all the KGS computer
events and CGOS, and also was the first to extend that dominance from
9x9 to 19x19."?


Hello Aja,

Here I quote from the computer-go archives, unless otherwise noted:

Dec. 31, 2006: John Tromp: "I spent most of yesterday on KGS playtesting 
MoGo on 9x9 with 30 min total thinking time. The experience was quite 
unlike any other program I've played on 9x9 in the

past. [..] I feel that the shodan level go 9x9 programs have arrived..."

Jan 12, 2007: Don Dailey: "Someone needs to get their bot on CGOS and 
end Mogo's reign of terror. A version of MoGo has achieved a CGOS rating 
of well over 2300!"


Mar 04, 2007: Peter Drake: "Congratulations to MoGo on winning the KGS 
tournament held earlier today: [..] Even under borderline "blitz" 
conditions (18 minutes sudden death for 19x19), MoGo managed to beat 
conventional programs like GNU Go. [..] Of course, MoGo also beat all 
the other MC/UCT programs. How did MoGo do it?"


Nick Wedd's summary from the above KGS tournament: "Last September 
Kocsis and Csaba Szepesvári published Bandit based Monte-Carlo Planning. 
The algorithm described there was implemented in MoGo. There was some 
doubt about its effectiveness for large boards and for fast games, but 
it was clear that it worked well for small boards particularly with slow 
time limits: MoGo has won nine of the eleven KGS bot tournaments held 
since September. However, this March event used full-sized boards and 
fast time limits. Some people expected MoGo to do less well, perhaps 
losing to more conventional programs such as ManyFaces, GNU Go, and Aya. 
They were wrong. MoGo won all twelve of its games."  [CrazyStone won 6 
games in the same tournament.] 
http://www.weddslist.com/kgs/past/24/index.html


Mar 17, 2007: Don Dailey: "It's unbelievable how strong MoGo is playing.

I remember when CGOS first came up,  I expected it to be a
few years before a program could achieve 2000.0 on the CGOS
scale.   But I was quickly surpised when programs started
breaking over 1800.0.

But this is quite incredible.   MoGo_G3.4 at 2480!   If you
look at the crosstable,  it's mostly 100% wins against
everyone else.   In fact, the only non-Mogo program to
beat it, won a single game out of 15 played, and it's
a pretty strong program too."


In Computer Olympiad 2007, Steenvreter was gold medal on 9x9.


Yes, and this is what Eric had to say about it: "Steenvreter uses UCT 
and has some L&D knowledge that I reused from Magog."


and: "Steenvreter was really a rush job, hacking things together until 
the last day before the tournament and no time to test properly. I was 
hoping to be able to catch up with the stronger programs, but never 
expected it to win the tournament."


Obviously it was following MoGo's lead with UCT (the tournament was held 
in June 2007, well after the remarkable success of MoGo). I don't mean 
to discredit Steenvreter, CrazyStone, or any other program. I'm just 
focusing on MoGo so much because it set the bar so high and got 
everybody chasing it.



Mogo's biggest contributions, so far, in my view, are
1.Applied UCT to computer Go, and such application came from the idea
"MCTS" that proposed in 2006 by Remi Coulom. Crazy Stone was using MCTS
to win 9x9 in 2006 Computer Olympiad.
2.See 3x3 patterns around the previous move.
3.RAVE (strictly speaking, it is invented by David Silver).


This is essentially what I said in my first message. I did not place 
enough emphasis on CrazyStone then, even though I did reference it ("The 
MoGo team applied UCT to Go with great success, using the idea of 
building incremental trees from CrazyStone.").



UCT and RAVE are for both for the tree search. I think Crazy tone's
contribution for the playout is of same/or more important, because the
quality of simulations decide the playing strength much. From this view,
we should give Crazy Stone more and more credit.


Sure, I agree. CrazyStone was definitely a big part of the monte carlo 
tree search revolution in computer go.


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

Re: [Computer-go] News on Tromp-Cook ?

[Aja]

Hi Fuming,

The idea of improving the quality of simulation is more earlier, than Mogo’s 
paper, in the Appendix A of Remi Coulom’s CG2006 paper “Efficient Selectivity 
and Backup Operators in Monte-Carlo Tree 
Search”(http://remi.coulom.free.fr/CG2006/CG2006.pdf):

The choice of a more clever probability distribution can improve the quality of 
the Monte-Carlo estimation

I am not sure if Remi was the first one proposing this concept in Computer Go 
field, but Mogo definitely was not.

I was in Amstertam attending Computer Olympiad 2007, in the team of Chinese 
chess program “Deep Elephant”. I played with Crazy Stone, Mogo and was very 
surprised to see they beat me. Afterwards, Mogo’s paper is so easy to 
understand/implement for me that trigger me to work on Computer Go. Indeed, 
Mogo has huge contributions, especially in the popularization of MCTS. I don’t 
mean to weaken or deny it, but just want to point out Crazy Stone’s great 
contributions. In Erica, I use CrazyStone-like simulations successfully. 
Mogo-type simulation almost does not help Erica at all. 

If we want to numerate the strongest programs, we cannot forget Fuego(2010 UEC 
Cup winner) and MyGoFriend(Computer Olympiad 2010, 9x9 winner). For academic 
progress, we cannot forget Crazy Stone. For practical development usage, we 
cannot forget GnuGo and GoGui released by Fuego team. There were really too 
many contributors in the past.

Happy New Years to all.

Aja
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Mark Boon]

I can say on a personal note that it was MoGo that first made me pay
real attention to MCTS. Even though it was maybe a transitional
program that early on wasn't strong enough yet to beat traditional
programs and later got quickly superseded by stronger MCTS variations.
And I think MoGo was the first program hat was playable by people on
KGS on 9x9, so you could tell it could beat strong (relatively)
players on that board-size.

That doesn't mean necessarily that MoGo contributed the most
scientifically of course. But these things are extremely hard to
measure. I think it's also obvious that CrazyStone has contributed a
lot.

Just my $0.02

Mark
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Fuming Wang]

This is certainly a good time to sit back and look at what got us here. The
following key ideas have been mentioned so far: UCB, MCTS, RAVE, Pattern and
Go knowledge during MC simulation.These ideas are all essential to a strong
MC based Go program.If we want to pick the most important idea that got us
here, I would say it is the realization that adding Go Pattern and Go
Knowledge to MC simulation can significantly improve the quality of board
evaluation. This is amount to the important sampling concept in MC
integration, which is very import for Monte Carlo applications in many
fields. MC simulation with importance sampling give us for the first time a
reasonablly accurate evaluation function for Go. UCB, MCTS, RAVE are
certainly very important, however, it is still possible that new approaches
that can achieve good results with just importantly samples MC simulation.
So, I think MoGo is the most important break-through.

Happy New Year, everyone!
Fuming



On Fri, Dec 31, 2010 at 9:20 AM, Aja  wrote:

> Hi Jeff,
>
> When, do you think, did Mogo "started dominating all the KGS computer
> events and CGOS, and also was the first to extend that dominance from 9x9 to
> 19x19."?
>
> In Computer Olympiad 2007, Steenvreter was gold medal on 9x9. At the final
> match of 19x19, it's easily to see that Mogo and Crazy Stone were close
> (finally Mogo 1st and CS 2ed). But, at the end of 2007, Crazy Stone defeated
> Mogo and won the UEC Cup (19x19). Afterwards, Many Faces won 9x9 and 19x19
> on 2008. Zen and Erica won 2009 and 2010, both continuing Crazy Stone
> thread.
>
> Mogo's biggest contributions, so far, in my view, are
> 1.Applied UCT to computer Go, and such application came from the idea
> "MCTS" that proposed in 2006 by Remi Coulom. Crazy Stone was using MCTS to
> win 9x9 in 2006 Computer Olympiad.
> 2.See 3x3 patterns around the previous move.
> 3.RAVE (strictly speaking, it is invented by David Silver).
>
> UCT and RAVE are for both for the tree search. I think Crazy tone's
> contribution for the playout is of same/or more important, because the
> quality of simulations decide the playing strength much. From this view, we
> should give Crazy Stone more and more credit.
>
> I don't mean to raise any debate. Mogo does has important contributions,
> but it's not so hard to assign credit to Crazy Stone. By the way, we should
> not forget Fuego and MyGoFriend. Anyway, I think SenSei's description is
> out-of-date.
>
> Aja
>
>
> -原始郵件- From: Jeff Nowakowski
> Sent: Friday, December 31, 2010 5:43 AM
> To: computer...@computer-go.org
>
> Subject: Re: [Computer-go] News on Tromp-Cook ?
>
> On 12/30/2010 01:58 PM, David Fotland wrote:
>
>> You should also give more credit to CrazyStone as an early strong program
>> that contributed many ideas, comparable to Mogo.  Remi is Aja's advisor,
>> so
>> Erica continues the CrazyStone thread.
>>
>
> I did mention CrazyStone, and the Sensei's page lists it first as the
> program that "started the new wave of MCTS programs by winning the 9x9
> gold medal at the ICGA Computer Olympiad, in 2006."  Like I said in my
> first message, though, it's hard to assign credit, and I don't mean to
> slight other programs.
>
> However, MoGo was the program that really got people to sit up and take
> notice, because it started dominating all the KGS computer events and
> CGOS, and also was the first to extend that dominance from 9x9 to 19x19.
> I believe the biggest breakthroughs were made with MoGo (building, of
> course, on earlier ideas). This is easily verified by going back to the
> archives and seeing how many people patterned their program after MoGo.
>
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Aja]

Sorry, I might be wrong at RAVE. Maybe it should be: Sylvain proprosed the 
idea of RAVE and David Silver proposed a new formula for RAVE.


Aja

-原始郵件- 
From: Aja

Sent: Friday, December 31, 2010 9:20 AM
To: computer-go@dvandva.org
Subject: Re: [Computer-go] News on Tromp-Cook ?

Hi Jeff,

When, do you think, did Mogo "started dominating all the KGS computer events
and CGOS, and also was the first to extend that dominance from 9x9 to
19x19."?

In Computer Olympiad 2007, Steenvreter was gold medal on 9x9. At the final
match of 19x19, it's easily to see that Mogo and Crazy Stone were close
(finally Mogo 1st and CS 2ed). But, at the end of 2007, Crazy Stone defeated
Mogo and won the UEC Cup (19x19). Afterwards, Many Faces won 9x9 and 19x19
on 2008. Zen and Erica won 2009 and 2010, both continuing Crazy Stone
thread.

Mogo's biggest contributions, so far, in my view, are
1.Applied UCT to computer Go, and such application came from the idea "MCTS"
that proposed in 2006 by Remi Coulom. Crazy Stone was using MCTS to win 9x9
in 2006 Computer Olympiad.
2.See 3x3 patterns around the previous move.
3.RAVE (strictly speaking, it is invented by David Silver).

UCT and RAVE are for both for the tree search. I think Crazy tone's
contribution for the playout is of same/or more important, because the
quality of simulations decide the playing strength much. From this view, we
should give Crazy Stone more and more credit.

I don't mean to raise any debate. Mogo does has important contributions, but
it's not so hard to assign credit to Crazy Stone. By the way, we should not
forget Fuego and MyGoFriend. Anyway, I think SenSei's description is
out-of-date.

Aja


-原始郵件- 
From: Jeff Nowakowski

Sent: Friday, December 31, 2010 5:43 AM
To: computer...@computer-go.org
Subject: Re: [Computer-go] News on Tromp-Cook ?

On 12/30/2010 01:58 PM, David Fotland wrote:

You should also give more credit to CrazyStone as an early strong program
that contributed many ideas, comparable to Mogo.  Remi is Aja's advisor, 
so

Erica continues the CrazyStone thread.


I did mention CrazyStone, and the Sensei's page lists it first as the
program that "started the new wave of MCTS programs by winning the 9x9
gold medal at the ICGA Computer Olympiad, in 2006."  Like I said in my
first message, though, it's hard to assign credit, and I don't mean to
slight other programs.

However, MoGo was the program that really got people to sit up and take
notice, because it started dominating all the KGS computer events and
CGOS, and also was the first to extend that dominance from 9x9 to 19x19.
I believe the biggest breakthroughs were made with MoGo (building, of
course, on earlier ideas). This is easily verified by going back to the
archives and seeing how many people patterned their program after MoGo.

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

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

Re: [Computer-go] News on Tromp-Cook ?

[Aja]

Hi Jeff,

When, do you think, did Mogo "started dominating all the KGS computer events 
and CGOS, and also was the first to extend that dominance from 9x9 to 
19x19."?


In Computer Olympiad 2007, Steenvreter was gold medal on 9x9. At the final 
match of 19x19, it's easily to see that Mogo and Crazy Stone were close 
(finally Mogo 1st and CS 2ed). But, at the end of 2007, Crazy Stone defeated 
Mogo and won the UEC Cup (19x19). Afterwards, Many Faces won 9x9 and 19x19 
on 2008. Zen and Erica won 2009 and 2010, both continuing Crazy Stone 
thread.


Mogo's biggest contributions, so far, in my view, are
1.Applied UCT to computer Go, and such application came from the idea "MCTS" 
that proposed in 2006 by Remi Coulom. Crazy Stone was using MCTS to win 9x9 
in 2006 Computer Olympiad.

2.See 3x3 patterns around the previous move.
3.RAVE (strictly speaking, it is invented by David Silver).

UCT and RAVE are for both for the tree search. I think Crazy tone's 
contribution for the playout is of same/or more important, because the 
quality of simulations decide the playing strength much. From this view, we 
should give Crazy Stone more and more credit.


I don't mean to raise any debate. Mogo does has important contributions, but 
it's not so hard to assign credit to Crazy Stone. By the way, we should not 
forget Fuego and MyGoFriend. Anyway, I think SenSei's description is 
out-of-date.


Aja


-原始郵件- 
From: Jeff Nowakowski

Sent: Friday, December 31, 2010 5:43 AM
To: computer...@computer-go.org
Subject: Re: [Computer-go] News on Tromp-Cook ?

On 12/30/2010 01:58 PM, David Fotland wrote:

You should also give more credit to CrazyStone as an early strong program
that contributed many ideas, comparable to Mogo.  Remi is Aja's advisor, 
so

Erica continues the CrazyStone thread.


I did mention CrazyStone, and the Sensei's page lists it first as the
program that "started the new wave of MCTS programs by winning the 9x9
gold medal at the ICGA Computer Olympiad, in 2006."  Like I said in my
first message, though, it's hard to assign credit, and I don't mean to
slight other programs.

However, MoGo was the program that really got people to sit up and take
notice, because it started dominating all the KGS computer events and
CGOS, and also was the first to extend that dominance from 9x9 to 19x19.
I believe the biggest breakthroughs were made with MoGo (building, of
course, on earlier ideas). This is easily verified by going back to the
archives and seeing how many people patterned their program after MoGo.

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


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

Re: [Computer-go] News on Tromp-Cook ?

[Jeff Nowakowski]

On 12/30/2010 06:10 PM, Michael Williams wrote:

Perhaps the client viewer should have the ability to hide comments by
rank.  Then anyone can be allowed to post, as long as they know that
not everyone will hear them.


The problem with that is you end up with a bunch of disjointed chat 
(people replying to those whose comments you have hidden), and even then 
the dans get sucked into random conversations.


I was really happy with the dan analysis, but then it's also fun to have 
the free-for-all kibitz too, so I think two separate game windows is the 
best solution. The best part is that this is already possible with the 
KGS clone feature.


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

Re: [Computer-go] News on Tromp-Cook ?

[Adrian Petrescu]

Sounds nice in theory, but in practice it will get super-confusing really
quick as different people will have different settings, so you will see many
half-conversations from people you can hear responding to people you can't.

On Thu, Dec 30, 2010 at 6:10 PM, Michael Williams <
michaelwilliam...@gmail.com> wrote:

> Perhaps the client viewer should have the ability to hide comments by
> rank.  Then anyone can be allowed to post, as long as they know that
> not everyone will hear them.
>
>
> On Thu, Dec 30, 2010 at 11:11 AM, Jeff Nowakowski 
> wrote:
> > On 12/29/2010 05:47 PM, Jacques Basaldúa wrote:
> >>
> >> I agree with Petr that the censoring was not just strange but
> >> probably lowered the level (computer-go wise) of the remarks
> >> compared to what we are used to in KGS tournaments. Maybe, its
> >> the price of having near 500 observers, which is a great
> >> achievement we all must be happy with.
> >
> > Normally I'm not happy with censoring, but I enjoyed reading the
> high-level
> > analysis from the dan players. I was actually shocked when I first
> started
> > reading it. Normally it is drowned out in lots of unrelated kibitz. I
> don't
> > think the kyu should have been allowed in, htough. If you're going to
> > restrict it to dan players, be consistent.
> >
> > Obviously you can't make everybody happy with these kinds of decisions.
> > Maybe an alternative is to have an official clone of the game that allows
> > everybody to speak.
> >
> > I'll add my thanks to Nick and everybody else involved for doing a
> splendid
> > job.
> >
> > ___
> > Computer-go mailing list
> > Computer-go@dvandva.org
> > http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
> >
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Michael Williams]

Perhaps the client viewer should have the ability to hide comments by
rank.  Then anyone can be allowed to post, as long as they know that
not everyone will hear them.


On Thu, Dec 30, 2010 at 11:11 AM, Jeff Nowakowski  wrote:
> On 12/29/2010 05:47 PM, Jacques Basaldúa wrote:
>>
>> I agree with Petr that the censoring was not just strange but
>> probably lowered the level (computer-go wise) of the remarks
>> compared to what we are used to in KGS tournaments. Maybe, its
>> the price of having near 500 observers, which is a great
>> achievement we all must be happy with.
>
> Normally I'm not happy with censoring, but I enjoyed reading the high-level
> analysis from the dan players. I was actually shocked when I first started
> reading it. Normally it is drowned out in lots of unrelated kibitz. I don't
> think the kyu should have been allowed in, htough. If you're going to
> restrict it to dan players, be consistent.
>
> Obviously you can't make everybody happy with these kinds of decisions.
> Maybe an alternative is to have an official clone of the game that allows
> everybody to speak.
>
> I'll add my thanks to Nick and everybody else involved for doing a splendid
> job.
>
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Hideki Kato]

David Fotland: <076301cba853$99234ad0$cb69e0...@com>:
>You should also give more credit to CrazyStone as an early strong program
>that contributed many ideas, comparable to Mogo.  Remi is Aja's advisor, so
>Erica continues the CrazyStone thread.

I'd like to add both Zen and Erica use the large patterns trained by 
the method in Remi's paper, "Computing Elo Ratings of Move Patterns in 
the Game of Go."  Zen's strength comes from both Rave and the large 
patterns with handcraft, high quality (less random) simulations by 
Yamato.

Hideki

>David
>
>> -Original Message-
>> From: computer-go-boun...@dvandva.org [mailto:computer-go-
>> boun...@dvandva.org] On Behalf Of "Ingo Althöfer"
>> Sent: Thursday, December 30, 2010 8:28 AM
>> To: computer-go@dvandva.org
>> Subject: Re: [Computer-go] News on Tromp-Cook ?
>> 
>> > ...
>> > The two strongest programs since MoGo are Zen and Many Faces
>> 
>> You should not forget "Erica" by Aja Huang,
>> winning the gold medal on 19x19 in the Computer Olympiad 2010.
>> (Zen and Many Faces also participated, getting ranks 2 and 3.)
>> 
>> Ingo.
>> --
>> GMX DSL Doppel-Flat ab 19,99 Euro/mtl.! Jetzt auch mit
>> gratis Notebook-Flat! http://portal.gmx.net/de/go/dsl
>> ___
>> Computer-go mailing list
>> Computer-go@dvandva.org
>> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
>___
>Computer-go mailing list
>Computer-go@dvandva.org
>http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
-- 
Hideki Kato <mailto:hideki_ka...@ybb.ne.jp>
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Jeff Nowakowski]

On 12/30/2010 01:58 PM, David Fotland wrote:

You should also give more credit to CrazyStone as an early strong program
that contributed many ideas, comparable to Mogo.  Remi is Aja's advisor, so
Erica continues the CrazyStone thread.


I did mention CrazyStone, and the Sensei's page lists it first as the 
program that "started the new wave of MCTS programs by winning the 9x9 
gold medal at the ICGA Computer Olympiad, in 2006."  Like I said in my 
first message, though, it's hard to assign credit, and I don't mean to 
slight other programs.


However, MoGo was the program that really got people to sit up and take 
notice, because it started dominating all the KGS computer events and 
CGOS, and also was the first to extend that dominance from 9x9 to 19x19. 
I believe the biggest breakthroughs were made with MoGo (building, of 
course, on earlier ideas). This is easily verified by going back to the 
archives and seeing how many people patterned their program after MoGo.


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

Re: [Computer-go] News on Tromp-Cook ?

[Robert Jasiek]

On 30.12.2010 20:40, Petr Baudis wrote:
> if their "insight" to give large

weight to the center of the board is truthful or just something that
will gradually disappear as they attain more strength.


If they are all tuned to fit traditional databases / traditional expert 
knowledge rules, then it will disappear again. Else it is an open 
question - also for myself as a player. Against unsuspecting opponents 
up to 7d, it is very effective. Against opponents experienced with 
playing me, winning percentages are similar to traditional openings.


Since the number of corners is even, playing the first move in the 
center still guarantees half of the corners (if one still wants them). 
(According to Saijo, they are not miai though - but I forgot his reasoning.)


I still think that each 3rd line move or higher works. Regardless of 
where the first 5 moves are played - their haengma is always relevant.


First generation MC programs might be prejudiced towards the center 
because groups in the center die harder accidentally and therefore 
empirically.


--
robert jasiek
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Petr Baudis]

  Hi!

On Thu, Dec 30, 2010 at 08:48:08AM +0100, Robert Jasiek wrote:
> Congratulations also to all the theorists! Without their great
> discoveries, programs would still be weak. Might somebody please
> give an overview on the relevant theories and how they work?

  http://pasky.or.cz/~pasky/go/ has a slideshow with a state-of-art
overview of most of the relevant concepts in current Computer Go. You
will perhaps find most of the content rather shallow and just scratching
the surface, but it could be a good starting point (with some references
too).

  I will need to expand it, though, as for the closed-source programs
the details are frequently unknown. E.g. it has been news for me that
MFoG uses progressive unpruning, I personally considered it largely
abandoned - I need to add a discussion on that, and perhaps also explore
it myself. ;-)

> One thing keeps bothering me though: What does all the strength
> improvement give us humans for better understanding the game
> strategy?

  What I honestly do keep wondering if their "insight" to give large
weight to the center of the board is truthful or just something that
will gradually disappear as they attain more strength.

-- 
Petr "Pasky" Baudis
Computer science education cannot make an expert programmer any more
than studying brushes and pigment can make an expert painter. --esr
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[David Fotland]

You should also give more credit to CrazyStone as an early strong program
that contributed many ideas, comparable to Mogo.  Remi is Aja's advisor, so
Erica continues the CrazyStone thread.

David

> -Original Message-
> From: computer-go-boun...@dvandva.org [mailto:computer-go-
> boun...@dvandva.org] On Behalf Of "Ingo Althöfer"
> Sent: Thursday, December 30, 2010 8:28 AM
> To: computer-go@dvandva.org
> Subject: Re: [Computer-go] News on Tromp-Cook ?
> 
> > ...
> > The two strongest programs since MoGo are Zen and Many Faces
> 
> You should not forget "Erica" by Aja Huang,
> winning the gold medal on 19x19 in the Computer Olympiad 2010.
> (Zen and Many Faces also participated, getting ranks 2 and 3.)
> 
> Ingo.
> --
> GMX DSL Doppel-Flat ab 19,99 Euro/mtl.! Jetzt auch mit
> gratis Notebook-Flat! http://portal.gmx.net/de/go/dsl
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

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

Re: [Computer-go] News on Tromp-Cook ?

[Ingo Althöfer]

> ...
> The two strongest programs since MoGo are Zen and Many Faces

You should not forget "Erica" by Aja Huang,
winning the gold medal on 19x19 in the Computer Olympiad 2010.
(Zen and Many Faces also participated, getting ranks 2 and 3.)

Ingo.
-- 
GMX DSL Doppel-Flat ab 19,99 Euro/mtl.! Jetzt auch mit 
gratis Notebook-Flat! http://portal.gmx.net/de/go/dsl
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Jeff Nowakowski]

On 12/29/2010 05:47 PM, Jacques Basaldúa wrote:


I agree with Petr that the censoring was not just strange but
probably lowered the level (computer-go wise) of the remarks
compared to what we are used to in KGS tournaments. Maybe, its
the price of having near 500 observers, which is a great
achievement we all must be happy with.


Normally I'm not happy with censoring, but I enjoyed reading the 
high-level analysis from the dan players. I was actually shocked when I 
first started reading it. Normally it is drowned out in lots of 
unrelated kibitz. I don't think the kyu should have been allowed in, 
htough. If you're going to restrict it to dan players, be consistent.


Obviously you can't make everybody happy with these kinds of decisions. 
Maybe an alternative is to have an official clone of the game that 
allows everybody to speak.


I'll add my thanks to Nick and everybody else involved for doing a 
splendid job.


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

Re: [Computer-go] News on Tromp-Cook ?

[Jeff Nowakowski]

On 12/30/2010 02:48 AM, Robert Jasiek wrote:


Nick has said that a 2007 Hungarian RAVE paper was the theoretical
breakthrough. Is this its URL?

http://zaphod.aml.sztaki.hu/papers/ecml06.pdf


It's very hard to say exactly which paper was a definitive breakthrough. 
They all build on other work, going all the way back into the 1990s.


The paper that you link to is the first description of UCT, but it 
wasn't used for Go in the paper. You can find a copy at archive.org: 
http://web.archive.org/web/*/http://zaphod.aml.sztaki.hu/papers/ecml06.pdf


The MoGo team applied UCT to Go with great success, using the idea of 
building incremental trees from CrazyStone. Another big factor in their 
success was the hand-picked patterns and the idea of using sequences (if 
a pattern matches near the last move, play it; this gives plausible 
sequences). The landmark paper that describes these techniques, and that 
many programs used as a blueprint was the first MoGo paper: "Modification 
of UCT with Patterns in Monte-Carlo Go", 
http://hal.inria.fr/inria-00117266 .


The MoGo team kept on advancing rapidly and introduced RAVE as their 
next big improvement. Another big improvement is the use of prior 
knowledge to seed the starting winning percentage for nodes in the tree. 
There are other improvements, but I'm not going to list everything here. 
Read the papers about MoGo if you want to know more.


The two strongest programs since MoGo are Zen and Many Faces. My 
understanding is that they have mainly focused on improving Go knowledge 
into the evaluation function.



Congratulations also to all the theorists! Without their great
discoveries, programs would still be weak. Might somebody please give an
overview on the relevant theories and how they work?


I'd start by reading the Sensei's page: http://senseis.xmp.net/?MonteCarlo

You can read the MoGo papers and trace the references if you want a 
detailed understanding. You can also search this list for many past 
discussions of these topics.



One thing keeps bothering me though: What does all the strength
improvement give us humans for better understanding the game strategy?


Not much, but eventually a strong computer program can be used to test 
theories in a more automated fashion, so that's something.


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

Re: [Computer-go] News on Tromp-Cook ?

[Robert Finking]

Many thanks

On 30/12/2010 10:34, Petr Baudis wrote:

On Thu, Dec 30, 2010 at 10:26:49AM +, Robert Finking wrote:

Hi David,

Many thanks for sharing this info with us. I am not a strong Go
player and have a question which I suspect many people on this list
could answer:

On 30/12/2010 08:09, David Fotland wrote:

...
hand tuned.  There are rules for not filling eyes, not making self Atari
(unless it is a good self Atari), and avoiding other kinds of bad moves.

When is self atari a good move?

E.g. when you are filling dead eyespace:

| . O O O .
| X X X X O
| O O . X O
| O a O X .
+--

a is selfatari, but good idea for w to play.

Another examples of interesting selfataris may be one-stone throw-ins to
falsify eyes or setting up snapbacks.

See also

http://repo.or.cz/w/pachi.git/blob/HEAD:/tactics/selfatari.c

as an example of rather complex code to evaluate selfataris. The code
itself is probably not very legible, but the comments contain a lot of
examples.


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

Re: [Computer-go] News on Tromp-Cook ?

[Petr Baudis]

On Thu, Dec 30, 2010 at 10:26:49AM +, Robert Finking wrote:
> Hi David,
> 
> Many thanks for sharing this info with us. I am not a strong Go
> player and have a question which I suspect many people on this list
> could answer:
> 
> On 30/12/2010 08:09, David Fotland wrote:
> >...
> >hand tuned.  There are rules for not filling eyes, not making self Atari
> >(unless it is a good self Atari), and avoiding other kinds of bad moves.
> When is self atari a good move?

E.g. when you are filling dead eyespace:

| . O O O .
| X X X X O
| O O . X O
| O a O X .
+--

a is selfatari, but good idea for w to play.

Another examples of interesting selfataris may be one-stone throw-ins to
falsify eyes or setting up snapbacks.

See also

http://repo.or.cz/w/pachi.git/blob/HEAD:/tactics/selfatari.c

as an example of rather complex code to evaluate selfataris. The code
itself is probably not very legible, but the comments contain a lot of
examples.

-- 
Petr "Pasky" Baudis
Computer science education cannot make an expert programmer any more
than studying brushes and pigment can make an expert painter. --esr
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Robert Finking]

On 30/12/2010 08:19, David Fotland wrote:


...

The old mfgo expert system suggested reasons/goals for moves, each 
with a value. After a move was made and the position evaluated, each 
reason was checked to see if the move actually achieved the goal. For 
example if the suggestion was “this move gains 20 points by killing 
group A”, after the move was played (with quiescence search) the new 
evaluation checked to see if group A had actually died.


There was a fairly elaborate system to combine the results of the 
goals to avoid double counting.



Awesome! Again - thanks for sharing this with us.

Merry Christmas and a Happy New Year
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Robert Finking]

Hi David,

Many thanks for sharing this info with us. I am not a strong Go player 
and have a question which I suspect many people on this list could answer:


On 30/12/2010 08:09, David Fotland wrote:

...
hand tuned.  There are rules for not filling eyes, not making self Atari
(unless it is a good self Atari), and avoiding other kinds of bad moves.

When is self atari a good move?

Thanks in advance

Raffles

PS I love the word "unpruning" by the way!

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

Re: [Computer-go] News on Tromp-Cook ?

[David Fotland]

Only the last game had full use of better hardware, since it used a 64 bit
version with more total memory.  If the time used is available we should see
that Many Faces used much more time on the last game.

 

From: computer-go-boun...@dvandva.org
[mailto:computer-go-boun...@dvandva.org] On Behalf Of Petri Pitkanen
Sent: Thursday, December 30, 2010 12:08 AM
To: computer-go@dvandva.org
Subject: Re: [Computer-go] News on Tromp-Cook ?

 

In last game it did not look like he was taking it easy :) I think Lower
left B was in real trouble. With my crappy tactival skill I thought that W
could have killed 4 black stones and saved his huge corner group. That would
have been swing of 40 points minimum.

So MFoG fights well most of the time then makes easy mistakes. Usually by
making Tenuki. I have some similar feeling playing MAnyFaces  on my laptop.
Every now and then I escape local disasters because something on global
scale seems to be more important - and some time it really is.

 But still huge effort to really challenge European 2d on a even game. Not
something that could have happened 5 years ago. 

2010/12/29 Mark Boon 

On Wed, Dec 29, 2010 at 11:37 AM, Darren Cook  wrote:

> John didn't notice much difference in the better hardware (*) used on
> day 2 (about 20,000 playouts/second, instead of 10,000 I was getting on
> my notebook).

Interesting. I wasn't aware the hardware was better for day 2. But I
did have a distinct feeling the computer was playing better the last
two games. I was even wondering if John had started to take it easy
after the first two easy wins.

Mark

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

 

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

Re: [Computer-go] News on Tromp-Cook ?

[David Fotland]

Many of the tenukis are due to a simple problem that I haven't had time to
fix yet.

 

The old mfgo expert system suggested reasons/goals for moves, each with a
value.  After a move was made and the position evaluated, each reason was
checked to see if the move actually achieved the goal.  For example if the
suggestion was "this move gains 20 points by killing group A", after the
move was played (with quiescence search) the new evaluation checked to see
if group A had actually died.

 

There was a fairly elaborate system to combine the results of the goals to
avoid double counting.

 

In the MCTS search, I use the suggested reasons to bias the search, but I
don't have any way to check the results as the old search did.  So many
goals are double counted.  This causes some of the biases to be way to high,
leading to tenukis.

 

Fixing this is not difficult, but there is a lot of new code to write and
debug, since there are many possible interactions between different goals
(about 200 total goals are checked).

 

David

 

From: computer-go-boun...@dvandva.org
[mailto:computer-go-boun...@dvandva.org] On Behalf Of Petri Pitkanen
Sent: Thursday, December 30, 2010 12:08 AM
To: computer-go@dvandva.org
Subject: Re: [Computer-go] News on Tromp-Cook ?

 

In last game it did not look like he was taking it easy :) I think Lower
left B was in real trouble. With my crappy tactival skill I thought that W
could have killed 4 black stones and saved his huge corner group. That would
have been swing of 40 points minimum.

So MFoG fights well most of the time then makes easy mistakes. Usually by
making Tenuki. I have some similar feeling playing MAnyFaces  on my laptop.
Every now and then I escape local disasters because something on global
scale seems to be more important - and some time it really is.

 But still huge effort to really challenge European 2d on a even game. Not
something that could have happened 5 years ago. 

2010/12/29 Mark Boon 

On Wed, Dec 29, 2010 at 11:37 AM, Darren Cook  wrote:

> John didn't notice much difference in the better hardware (*) used on
> day 2 (about 20,000 playouts/second, instead of 10,000 I was getting on
> my notebook).

Interesting. I wasn't aware the hardware was better for day 2. But I
did have a distinct feeling the computer was playing better the last
two games. I was even wondering if John had started to take it easy
after the first two easy wins.

Mark

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

 

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

Re: [Computer-go] News on Tromp-Cook ?

[David Fotland]

A short description of Many Faces:

It's an MCTS full board searcher.  For the tree it uses the UCT formula and
RAVE, with an exploration term, and an MFGO bias.  It does progressive
unpruning up to a maximum of 30 moves per position.  The unpruning decision
is based on rave and MFGO bias.

MFGO bias is based on the move values from the old Many Faces of Go move
generator (a full board static evaluation (with local tactical search), big
pattern databases, and expert system rules).  The generator estimates points
gained by each move.

The playouts are pretty heavy, with local responses, hand tuned 3x3
patterns, and moves played with a probability distribution similar to Crazy
Stones gamma values, but without the automatic learning.  Gamma values are
hand tuned.  There are rules for not filling eyes, not making self Atari
(unless it is a good self Atari), and avoiding other kinds of bad moves.
The eye and self Atari rules are a little different from published methods.
There is no tactical look-ahead in the playouts.

The multithread search is different from other programs.

It doesn't do pondering yet.

David

> -Original Message-
> From: computer-go-boun...@dvandva.org [mailto:computer-go-
> boun...@dvandva.org] On Behalf Of Robert Jasiek
> Sent: Wednesday, December 29, 2010 11:48 PM
> To: computer-go@dvandva.org
> Subject: Re: [Computer-go] News on Tromp-Cook ?
> 
> Despite his loss of the bet on the surface, I congratulate Darren for
> almost correctly predicting the 19x19 computer strength development! It
> has been an extraordinarly impressive improvement during the last 3
> years! Before 19x19 was more like 10 kyu - now during parts of a game
> ManyFaces can hold 1d to 2d level! With some more programming effort for
> holding a program's playing strength at a constant level (maybe also by
> filtering computer suggested moves by a human approach bias filter to
> discard obviously bad moves like A15 in game 3 and by making endgame
> more expert-orientated again), this strength can soon be held during an
> entire game.
> 
> Nick has said that a 2007 Hungarian RAVE paper was the theoretical
> breakthrough. Is this its URL?
> 
> http://zaphod.aml.sztaki.hu/papers/ecml06.pdf
> 
> The site appears to be down though. Is there an alternative URL?
> 
> ManyFaces was described as an expert system. How does it work today? How
> does it use the modern algorithmic theories?
> 
> Congratulations also to all the theorists! Without their great
> discoveries, programs would still be weak. Might somebody please give an
> overview on the relevant theories and how they work?
> 
> One thing keeps bothering me though: What does all the strength
> improvement give us humans for better understanding the game strategy?
> Almost nothing? The information contained in the current calculation
> size is not easily translated to human applicable strategic / tactical
> knowledge. Other research, which is closer to the human way of go
> understanding, by people like Berlekamp, Spight, Cazenawe or myself is
> much more useful for players but its playing strength equivalent -
> despite a few 10p knowledge exceptions - is still on the 20k level.
> Currently there is an extreme gap between computer go theory making
> computers strong, maths theory explaining go theory for human
> understanding and traditional professional go theory, which fails to
> explain well but allows eager and gifted players to succeed by means of
> unlimited investment of time and effort. What is still mostly missing
> are ways to link well to each other the three major paths towards great
> playing strength.
> 
> - How can programs learn well from professional knowledge?
> - How can programs use well mathematical descriptions of human-like
> strategy?
> - How can players learn well from strong programs?
> - How can further mathematical descriptions of human-like strategy be
> derived from strong computer play or its underlying algorithms?
> 
> Oh, and of course congratulations to John!
> 
> --
> robert jasiek
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

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

Re: [Computer-go] News on Tromp-Cook ?

[Petri Pitkanen]

In last game it did not look like he was taking it easy :) I think Lower
left B was in real trouble. With my crappy tactival skill I thought that W
could have killed 4 black stones and saved his huge corner group. That would
have been swing of 40 points minimum.

So MFoG fights well most of the time then makes easy mistakes. Usually by
making Tenuki. I have some similar feeling playing MAnyFaces  on my laptop.
Every now and then I escape local disasters because something on global
scale seems to be more important - and some time it really is.

 But still huge effort to really challenge European 2d on a even game. Not
something that could have happened 5 years ago.

2010/12/29 Mark Boon 

> On Wed, Dec 29, 2010 at 11:37 AM, Darren Cook  wrote:
>
> > John didn't notice much difference in the better hardware (*) used on
> > day 2 (about 20,000 playouts/second, instead of 10,000 I was getting on
> > my notebook).
>
> Interesting. I wasn't aware the hardware was better for day 2. But I
> did have a distinct feeling the computer was playing better the last
> two games. I was even wondering if John had started to take it easy
> after the first two easy wins.
>
> Mark
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Robert Jasiek]

Despite his loss of the bet on the surface, I congratulate Darren for 
almost correctly predicting the 19x19 computer strength development! It 
has been an extraordinarly impressive improvement during the last 3 
years! Before 19x19 was more like 10 kyu - now during parts of a game 
ManyFaces can hold 1d to 2d level! With some more programming effort for 
holding a program's playing strength at a constant level (maybe also by 
filtering computer suggested moves by a human approach bias filter to 
discard obviously bad moves like A15 in game 3 and by making endgame 
more expert-orientated again), this strength can soon be held during an 
entire game.


Nick has said that a 2007 Hungarian RAVE paper was the theoretical 
breakthrough. Is this its URL?


http://zaphod.aml.sztaki.hu/papers/ecml06.pdf

The site appears to be down though. Is there an alternative URL?

ManyFaces was described as an expert system. How does it work today? How 
does it use the modern algorithmic theories?


Congratulations also to all the theorists! Without their great 
discoveries, programs would still be weak. Might somebody please give an 
overview on the relevant theories and how they work?


One thing keeps bothering me though: What does all the strength 
improvement give us humans for better understanding the game strategy? 
Almost nothing? The information contained in the current calculation 
size is not easily translated to human applicable strategic / tactical 
knowledge. Other research, which is closer to the human way of go 
understanding, by people like Berlekamp, Spight, Cazenawe or myself is 
much more useful for players but its playing strength equivalent - 
despite a few 10p knowledge exceptions - is still on the 20k level. 
Currently there is an extreme gap between computer go theory making 
computers strong, maths theory explaining go theory for human 
understanding and traditional professional go theory, which fails to 
explain well but allows eager and gifted players to succeed by means of 
unlimited investment of time and effort. What is still mostly missing 
are ways to link well to each other the three major paths towards great 
playing strength.


- How can programs learn well from professional knowledge?
- How can programs use well mathematical descriptions of human-like 
strategy?

- How can players learn well from strong programs?
- How can further mathematical descriptions of human-like strategy be 
derived from strong computer play or its underlying algorithms?


Oh, and of course congratulations to John!

--
robert jasiek
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[steve uurtamo]

in the last game, a comment i especially noted was that miai was
handled poorly by the computer player and was a seemingly effective
strategy for playing against a computer, since there were many miai
that john left for later. once two fights got close enough to affect
one another, the miai got more complicated, because john could
seemingly (locally) sacrifice groups that were still just fine. on a
positive note, MFoG seemed to do a good job at working to keep sente.

on the question of who was allowed to comment and who was not -- in
the first game, the rule as stated was that only dan-level players
would be given permission to comment. then a 2kyu player was added
into the mix, but the level of conversation was indeed restrained by
the fact that it was only 4-5 people giving most of the commentary.

s.

On Thu, Dec 30, 2010 at 3:14 AM, terry mcintyre  wrote:
> In one of the games, "gogonuts" - who has played many games with computer
> programs - opined that multiple unresolved semeai are a weakness of MC 
> programs.
>
> A strong human would reason that if the outcome of semeai A, B, and C are
> individually losses for the program, then the aggregate probability is to lose
> all three fights.
>  Terry McIntyre 
>
>
> Unix/Linux Systems Administration
> Taking time to do it right saves having to do it twice.
>
>
>
> - Original Message 
>> From: Mark Boon 
>> To: computer-go@dvandva.org
>> Sent: Wed, December 29, 2010 6:25:14 PM
>> Subject: Re: [Computer-go] News on Tromp-Cook ?
>>
>> On Wed, Dec 29, 2010 at 12:47 PM, Jacques Basaldúa  wrote:
>> > I don't  agree the first two games were that easy.
>> >
>> > In the second game the  bot was ahead most of the game
>> > and failed in life and death in the top  right corner.
>>
>> Yes, we disagree. The first game was finished before it had  well
>> begun. By move 50 or so.
>>
>> The second game the computer was also  doing poorly until John got
>> careless about his left-side group. The computer  played that well and
>> took the lead. But only briefly, as it had the top-right  corner killed
>> immediately after. So I wouldn't say the computer was ahead  most of
>> the game. It was ahead for only a very brief moment.
>>
>> But  opinions about games can differ of course. Generally I saw the
>> computer do  some impressive things and some very silly things.
>>
>>      Mark
>> ___
>> Computer-go mailing  list
>> Computer-go@dvandva.org
>> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>>
>
>
>
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[terry mcintyre]

In one of the games, "gogonuts" - who has played many games with computer 
programs - opined that multiple unresolved semeai are a weakness of MC programs.

A strong human would reason that if the outcome of semeai A, B, and C are 
individually losses for the program, then the aggregate probability is to lose 
all three fights. 
 Terry McIntyre 


Unix/Linux Systems Administration
Taking time to do it right saves having to do it twice.



- Original Message 
> From: Mark Boon 
> To: computer-go@dvandva.org
> Sent: Wed, December 29, 2010 6:25:14 PM
> Subject: Re: [Computer-go] News on Tromp-Cook ?
> 
> On Wed, Dec 29, 2010 at 12:47 PM, Jacques Basaldúa  wrote:
> > I don't  agree the first two games were that easy.
> >
> > In the second game the  bot was ahead most of the game
> > and failed in life and death in the top  right corner.
> 
> Yes, we disagree. The first game was finished before it had  well
> begun. By move 50 or so.
> 
> The second game the computer was also  doing poorly until John got
> careless about his left-side group. The computer  played that well and
> took the lead. But only briefly, as it had the top-right  corner killed
> immediately after. So I wouldn't say the computer was ahead  most of
> the game. It was ahead for only a very brief moment.
> 
> But  opinions about games can differ of course. Generally I saw the
> computer do  some impressive things and some very silly things.
> 
>  Mark
> ___
> Computer-go mailing  list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
> 


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

Re: [Computer-go] News on Tromp-Cook ?

[David Fotland]

The first three games didn't have enough memory for the time limits so
weren't full strength.  The search stopped early when memory ran out due to
a garbage collection bug.  The last game used a 64-bit build that had plenty
of memory, so it used its time better.

David

> -Original Message-
> From: computer-go-boun...@dvandva.org [mailto:computer-go-
> boun...@dvandva.org] On Behalf Of Mark Boon
> Sent: Wednesday, December 29, 2010 1:59 PM
> To: computer-go@dvandva.org
> Subject: Re: [Computer-go] News on Tromp-Cook ?
> 
> On Wed, Dec 29, 2010 at 11:37 AM, Darren Cook  wrote:
> 
> > John didn't notice much difference in the better hardware (*) used on
> > day 2 (about 20,000 playouts/second, instead of 10,000 I was getting on
> > my notebook).
> 
> Interesting. I wasn't aware the hardware was better for day 2. But I
> did have a distinct feeling the computer was playing better the last
> two games. I was even wondering if John had started to take it easy
> after the first two easy wins.
> 
> Mark
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

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

Re: [Computer-go] News on Tromp-Cook ?

[Mark Boon]

On Wed, Dec 29, 2010 at 12:47 PM, Jacques Basaldúa  wrote:
> I don't agree the first two games were that easy.
>
> In the second game the bot was ahead most of the game
> and failed in life and death in the top right corner.

Yes, we disagree. The first game was finished before it had well
begun. By move 50 or so.

The second game the computer was also doing poorly until John got
careless about his left-side group. The computer played that well and
took the lead. But only briefly, as it had the top-right corner killed
immediately after. So I wouldn't say the computer was ahead most of
the game. It was ahead for only a very brief moment.

But opinions about games can differ of course. Generally I saw the
computer do some impressive things and some very silly things.

Mark
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Mark Boon]

On Wed, Dec 29, 2010 at 11:37 AM, Darren Cook  wrote:

> John didn't notice much difference in the better hardware (*) used on
> day 2 (about 20,000 playouts/second, instead of 10,000 I was getting on
> my notebook).

Interesting. I wasn't aware the hardware was better for day 2. But I
did have a distinct feeling the computer was playing better the last
two games. I was even wondering if John had started to take it easy
after the first two easy wins.

Mark
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Darren Cook]

> Thanks for posting. David (and others) made great progress. From these
> games it doesn't quite look like 1d yet? At least not on that
> hardware. But it's getting there.

John thought it might be 1-kyu (this assumes John is 2-dan) and I think
that is reasonable: if the computer got a 2-stone handicap I think win
rate might be nearer 50%, and if the computer got 3-stones I think John
would be in trouble.

John didn't notice much difference in the better hardware (*) used on
day 2 (about 20,000 playouts/second, instead of 10,000 I was getting on
my notebook).

My feeling, watching computer stats and some of the KGS feedback, was
that the computer is generally okay in the opening; it's joseki would be
better if it didn't tenuki so much. Then in the middle game it is
sometimes stronger than John, then sometimes so weak even I could play
better moves.

John's advice to beat the computer was to:
  a) be patient and wait for it to screw up;
  b) don't make weak groups, as the computer is good at attacking them.

Darren


*: I'll write up my notes on using Amazon EC2 in a few days. If you're
actually interested, please do nag me.

-- 
Darren Cook, Software Researcher/Developer

http://dcook.org/gobet/  (Shodan Go Bet - happening now!)
http://dcook.org/work/ (About me and my work)
http://dcook.org/blogs.html (My blogs and articles)
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Mark Boon]

Thanks for posting. David (and others) made great progress. From these
games it doesn't quite look like 1d yet? At least not on that
hardware. But it's getting there.

Mark

On Wed, Dec 29, 2010 at 9:52 AM, Andy  wrote:
> Tromp won 4-0.  Here are the games with comments/clickos removed:
> Game 1 http://eidogo.com/#HV6TkFC
> Game 2 http://eidogo.com/#1gE2QFA7
> Game 3 http://eidogo.com/#2Dao1GLA
> Game 4 http://eidogo.com/#1YFMXdg
>
>
>
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Ingo Althöfer]

Thanks for the service and the links.

I recommend to have also a look at the
sgf WITH the online chat. The comments
are not only smalltalk but contain a lot
of interesting information and opinion.

http://www.gokgs.com/gameArchives.jsp?user=bgamatches
You can download the sgf by clicking on "yes" in the column "visible?".

Ingo.

 Original-Nachricht 
> Datum: Wed, 29 Dec 2010 13:52:07 -0600
> Von: Andy 
> An: computer-go@dvandva.org
> Betreff: Re: [Computer-go] News on Tromp-Cook ?

> Tromp won 4-0.  Here are the games with comments/clickos removed:
> 
> Game 1 http://eidogo.com/#HV6TkFC
> Game 2 http://eidogo.com/#1gE2QFA7
> Game 3 http://eidogo.com/#2Dao1GLA
> Game 4 http://eidogo.com/#1YFMXdg
> <http://eidogo.com/#1YFMXdg>

-- 
GMX DSL Doppel-Flat ab 19,99 Euro/mtl.! Jetzt auch mit 
gratis Notebook-Flat! http://portal.gmx.net/de/go/dsl
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Andy]

Tromp won 4-0.  Here are the games with comments/clickos removed:

Game 1 http://eidogo.com/#HV6TkFC
Game 2 http://eidogo.com/#1gE2QFA7
Game 3 http://eidogo.com/#2Dao1GLA
Game 4 http://eidogo.com/#1YFMXdg

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

Re: [Computer-go] News on Tromp-Cook ?

[Darren Cook]

Well, 3-0 down, but the program seems to be getting closer to the win in
each game. The 4th, and probably final, game starts in ten minutes.

Hopefully this will be the first game using the 64-bit engine, meaning
it can use more than 1.5Gb of memory. I'm expecting that to give a small
increase in strength, but nothing amazing.

Darren


-- 
Darren Cook, Software Researcher/Developer

http://dcook.org/gobet/  (Shodan Go Bet - happening now!)
http://dcook.org/work/ (About me and my work)
http://dcook.org/blogs.html (My blogs and articles)
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Petr Baudis]

  Hi!

On Wed, Dec 29, 2010 at 12:45:32PM +0100, "Ingo Althöfer" wrote:
> send pn to BGAmatches(=NickWedd)
> and ask him to let you talk.

  Of course I (sort of) tried that too, but it didn't work. But even so,
it is very strange.

Petr "Pasky" Baudis
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Ingo Althöfer]

PEtr,

send pn to BGAmatches(=NickWedd)
and ask him to let you talk.

Ingo.

 Original-Nachricht 
> Datum: Wed, 29 Dec 2010 12:12:32 +0100
> Von: Petr Baudis 
> An: computer-go@dvandva.org
> Betreff: Re: [Computer-go] News on Tromp-Cook ?

> On Thu, Dec 23, 2010 at 11:41:13AM +, Nick Wedd wrote:
> > In message <20101223094425.218...@gmx.net>, Ingo Althöfer
> > <3-hirn-ver...@gmx.de> writes
> > >Will the games be transmitted on KGS?
> > 
> > Yes - in the EGR, relayed by 'BGAmatches'.
> 
> Thank you for the relay. So far it seems John will win his bet. ;-)
> Will also other programs than ManyFaces try their luck?
> 
> However, it is rather frustrating that one cannot participate in the
> game discussion since chat is blocked for all but (seemingly) random
> selection of individuals. Why has this strange decision been made?
> It feels so wrong on many levels...
> 
> -- 
>   Petr "Pasky" Baudis
> Computer science education cannot make an expert programmer any more
> than studying brushes and pigment can make an expert painter. --esr
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

-- 
GMX DSL Doppel-Flat ab 19,99 Euro/mtl.! Jetzt auch mit 
gratis Notebook-Flat! http://portal.gmx.net/de/go/dsl
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Petr Baudis]

On Thu, Dec 23, 2010 at 11:41:13AM +, Nick Wedd wrote:
> In message <20101223094425.218...@gmx.net>, Ingo Althöfer
> <3-hirn-ver...@gmx.de> writes
> >Will the games be transmitted on KGS?
> 
> Yes - in the EGR, relayed by 'BGAmatches'.

Thank you for the relay. So far it seems John will win his bet. ;-)
Will also other programs than ManyFaces try their luck?

However, it is rather frustrating that one cannot participate in the
game discussion since chat is blocked for all but (seemingly) random
selection of individuals. Why has this strange decision been made?
It feels so wrong on many levels...

-- 
Petr "Pasky" Baudis
Computer science education cannot make an expert programmer any more
than studying brushes and pigment can make an expert painter. --esr
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Willemien]

John Tromp EGD.: PIN: 10598984 :

http://www.europeangodatabase.eu/EGD/Player_Card.php?&key=10598984

On Tue, Dec 28, 2010 at 10:53 AM, steve uurtamo  wrote:
> can't find any rating info for john tromp via the european ranking
> data or the AGA. :)
>
> s.
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Darren Cook]

>> It seems first game today is at 2:30pm (GMT), not 10am.

Starting at 12:10pm-ish. We had a short time-limit test match, and
decided to do away with byo-yomi (too hard to keep match clock in sync
with the computer clock) so we've increased main clock to 110m each,
sudden death.

Second game today will be 1-2 hours after the first game finishes.

Darren
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Darren Cook]

> It seems first game today is at 2:30pm (GMT), not 10am...

We might start earlier now... I'll try and post just before the game
starts, for anyone wanting to follow the games from the start.

Darren
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[steve uurtamo]

can't find any rating info for john tromp via the european ranking
data or the AGA. :)

s.

On Tue, Dec 28, 2010 at 11:31 AM, Darren Cook  wrote:
>> Looks like the games start one hour earlier than that, at 10am and 2pm:
>>    http://www.computer-go.info/h-c/gobet/index.html
>
> It seems first game today is at 2:30pm (GMT), not 10am. (Second game is
> at 6:30pm.) Both John and I were sure it was this morning, and we both
> turned up at 10am, so perhaps the change was made late (Nick isn't here
> yet to ask). The games the rest of the week will be at 10am/2pm.
>
> First game, at least, will use Many Faces. It may be running under 32
> bits, meaning memory will be limited to 1.5Gb or 2Gb. Hardware being
> used for the games today is i7-740QM, 4 cores at 1.7Ghz, used as 8
> hyperthreaded cores.
>
> Many Faces has shown good improvement on KGS since the Kanazawa
> tournament in Sep 2010:
>  http://www.gokgs.com/graphPage.jsp?user=manyfaces
>
> I spoke with John over breakfast and he said he hasn't played any
> practice games other than those against Zen and Many Faces that were
> posted here a month or two ago. He said he brought a couple of go books
> with him but only got 12 pages in before decided to sleep on the plane :-)
>
> Personally I'd have been doing some last minute cramming, but he seems
> relaxed and confident.
>
> Darren
>
>
>
>> The event is being hosted by the London Go Open, more information here:
>>   http://www.britgo.org/tournaments/2010/logc
>>
>>> Will the games be transmitted on KGS?
>>
>> That is the plan, assuming we have a spare computer for it.
>>
>> Games will be run on my notebook, at least for the first two games.
>> Assuming my email works okay, I'll try and write reports each day.
>>
>> Darren
>>
>>
>
>
> --
> Darren Cook, Software Researcher/Developer
>
> http://dcook.org/gobet/  (Shodan Go Bet - happening now!)
> http://dcook.org/work/ (About me and my work)
> http://dcook.org/blogs.html (My blogs and articles)
> ___
> Computer-go mailing list
> Computer-go@dvandva.org
> http://dvandva.org/cgi-bin/mailman/listinfo/computer-go
>
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Darren Cook]

> Looks like the games start one hour earlier than that, at 10am and 2pm:
>http://www.computer-go.info/h-c/gobet/index.html

It seems first game today is at 2:30pm (GMT), not 10am. (Second game is
at 6:30pm.) Both John and I were sure it was this morning, and we both
turned up at 10am, so perhaps the change was made late (Nick isn't here
yet to ask). The games the rest of the week will be at 10am/2pm.

First game, at least, will use Many Faces. It may be running under 32
bits, meaning memory will be limited to 1.5Gb or 2Gb. Hardware being
used for the games today is i7-740QM, 4 cores at 1.7Ghz, used as 8
hyperthreaded cores.

Many Faces has shown good improvement on KGS since the Kanazawa
tournament in Sep 2010:
  http://www.gokgs.com/graphPage.jsp?user=manyfaces

I spoke with John over breakfast and he said he hasn't played any
practice games other than those against Zen and Many Faces that were
posted here a month or two ago. He said he brought a couple of go books
with him but only got 12 pages in before decided to sleep on the plane :-)

Personally I'd have been doing some last minute cramming, but he seems
relaxed and confident.

Darren



> The event is being hosted by the London Go Open, more information here:
>   http://www.britgo.org/tournaments/2010/logc
> 
>> Will the games be transmitted on KGS?
> 
> That is the plan, assuming we have a spare computer for it.
> 
> Games will be run on my notebook, at least for the first two games.
> Assuming my email works okay, I'll try and write reports each day.
> 
> Darren
> 
> 


-- 
Darren Cook, Software Researcher/Developer

http://dcook.org/gobet/  (Shodan Go Bet - happening now!)
http://dcook.org/work/ (About me and my work)
http://dcook.org/blogs.html (My blogs and articles)
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[steve uurtamo]

>> Will the games be transmitted on KGS?
>
> Yes - in the EGR, relayed by 'BGAmatches'.
>
> Nick

great! is this still valid info? (first game in 14 minutes or so?)

s.
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Nick Wedd]

In message <20101223094425.218...@gmx.net>, Ingo Althöfer 
<3-hirn-ver...@gmx.de> writes

Hello,

December 29 is approaching, and with it the
start of the "ShodanBet showdown" in London.

Is the timetable of
http://dcook.org/gobet/details.html
still valid?



The current plan is that two games will be played each day,
at 11am and 3pm, on the Tuesday, Wednesday and Thursday
(December 28th to 30th), and then if still not decided, the
final and deciding game will be Friday, December 31st, 2010,
at 11am.



No.  There is a valid (I hope) timetable at
http://www.computer-go.info/h-c/gobet/index.html


Will the games be transmitted on KGS?


Yes - in the EGR, relayed by 'BGAmatches'.

Nick
--
Nick Weddn...@maproom.co.uk
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go

Re: [Computer-go] News on Tromp-Cook ?

[Darren Cook]

> December 29 is approaching, and with it the
> start of the "ShodanBet showdown" in London.
> 
>> The current plan is that two games will be played each day, 
>> at 11am and 3pm, on the Tuesday, Wednesday and Thursday 
>> (December 28th to 30th), and then if still not decided, the 
>> final and deciding game will be Friday, December 31st, 2010, 
>> at 11am.

Looks like the games start one hour earlier than that, at 10am and 2pm:
   http://www.computer-go.info/h-c/gobet/index.html

The event is being hosted by the London Go Open, more information here:
  http://www.britgo.org/tournaments/2010/logc

> Will the games be transmitted on KGS?

That is the plan, assuming we have a spare computer for it.

Games will be run on my notebook, at least for the first two games.
Assuming my email works okay, I'll try and write reports each day.

Darren


-- 
Darren Cook, Software Researcher/Developer

http://dcook.org/gobet/  (Shodan Go Bet - who will win?)
http://dcook.org/work/ (About me and my work)
http://dcook.org/blogs.html (My blogs and articles)
___
Computer-go mailing list
Computer-go@dvandva.org
http://dvandva.org/cgi-bin/mailman/listinfo/computer-go