Page 4 of 8

Re: Rerun of the Tromp-Taylor bet

Posted: Sat Jan 14, 2012 10:46 am
by Cassandra
yoyoma wrote:Computers are quickly getting strong at Go, but it looks like it is still very difficult to strip comments from SGF. :lol:

OPEN the SGF-file with a text-editor
DO
search "]C["
IF found THEN
expand the text until the number of "[" / "]" after the "C" is equal
type "]" '= deleting the comment
ELSE
END
ENDIF
LOOP

Re: Rerun of the Tromp-Taylor bet

Posted: Sat Jan 14, 2012 11:14 am
by daal
Cassandra wrote:
yoyoma wrote:Computers are quickly getting strong at Go, but it looks like it is still very difficult to strip comments from SGF. :lol:

OPEN the SGF-file with a text-editor
DO
search "]C["
IF found THEN
expand the text until the number of "[" / "]" after the "C" is equal
type "]" '= deleting the comment
ELSE
END
ENDIF
LOOP


I have an inkling of what you're talking about, but for me, easy is something different.

Re: Rerun of the Tromp-Taylor bet

Posted: Sat Jan 14, 2012 11:28 am
by RBerenguel

Re: Rerun of the Tromp-Taylor bet

Posted: Sat Jan 14, 2012 11:43 am
by speedchase
I just decided to write one in Java... lets see how this goes

Re: Rerun of the Tromp-Taylor bet

Posted: Sat Jan 14, 2012 11:45 am
by RBerenguel
Ahhh Java... nooooo!

Re: Rerun of the Tromp-Taylor bet

Posted: Sat Jan 14, 2012 12:03 pm
by speedchase
What? i like java.

Re: Rerun of the Tromp-Taylor bet

Posted: Sat Jan 14, 2012 12:07 pm
by RBerenguel
I guessed, it's just it's one of my "don't program again in that" languages :)

Re: Rerun of the Tromp-Taylor bet

Posted: Sat Jan 14, 2012 4:16 pm
by Rémi
RBerenguel wrote:http://senseis.xmp.net/?CommentFilter

I would rather recommend sgfc:
http://www.red-bean.com/sgf/sgfc/index.html
it can remove comments with the -yC option. You can even do it directly online:
http://www.red-bean.com/sgf/sgfc/munch.html
(enter "C" in the -y option ("delete stuff"))

Rémi

Re: Rerun of the Tromp-Taylor bet

Posted: Sat Jan 14, 2012 8:01 pm
by Mike Novack
"Beginner"?

Not with regard to computer go, eh Rémi?

Re: Rerun of the Tromp-Taylor bet

Posted: Sat Jan 14, 2012 8:35 pm
by daniel_the_smith
MultiGo has a "strip comments" feature.

When I saw the game Joaz posted I thought Tromp had passed on move 4 to give a "handicap"... :lol:

EDIT: after looking over the whole thing, maybe he *should* have passed once or twice at the beginning... :o

Re: Rerun of the Tromp-Taylor bet

Posted: Sat Jan 14, 2012 9:03 pm
by yoyoma
Game 2, comments stripped by http://www.red-bean.com/sgf/sgfc/munch.cgi

Zen19N won, score is tied 1-1 now.



Re: Rerun of the Tromp-Taylor bet

Posted: Sun Jan 15, 2012 12:04 am
by Joaz Banbeck
As before, Zen is center biased. Tromp seemed to be trying to beat it at its own game.

Re: Rerun of the Tromp-Taylor bet

Posted: Sun Jan 15, 2012 10:47 am
by badukJr
Can Zen be modified during these strings of matches? He seems to not use enough of his time. I understand being conservative about it though. If MC enters overtime, there won't be enough playouts generated to create necessary confidence level about the moves.

Tromp got lucky when Zen played some wacky moves the first game. I think it will continue to be difficult for him.

Re: Rerun of the Tromp-Taylor bet

Posted: Sun Jan 15, 2012 3:55 pm
by SpongeBob
Was it neccessary for Tromp to lose the large group at the bottom in the second game? Was it a misread on his part?

Re: Rerun of the Tromp-Taylor bet

Posted: Sun Jan 15, 2012 4:39 pm
by Joaz Banbeck
SpongeBob wrote:Was it neccessary for Tromp to lose the large group at the bottom in the second game? Was it a misread on his part?


I'v'e been wondering that too. From about move 109 to 125, he had sente several times, or could have taken small losses by tenuking to save his group. After 126, it looks dead.

Particularly noteworthy was move 111, with which he killed Zen's upper right. But his net profit there looks smaller than the loss of his dead group. So my conclusion is that, yes, he misread: he thought that he could live after 126 - probably by connecting to the northeast.
Also move 123 looks smaller than the dead group, so he must have thought that he could live.

If he lost on a misread, then my bet is still that he will win. When he does not misread, the game is lopsided.