124. Chew (3k) vs Fuego (Bot)

emeraldemon · Post by **emeraldemon** » Thu Jun 23, 2011 9:16 am

A comparison to the numbers from last move: Fuego only spent 69% of its time on the played move this time. And 45% of the playouts began with this sequence:

Click Here To Show Diagram Code: [go]$$Bcm29 $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . O . . . . . . . . . | $$ | . . . O . O . . . , . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . O . . . . . . . . . . . 4 . . | $$ | . . X . . . . . . . . . . . . O 3 . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . 2 . . . | $$ | . . . . X X . X . . . . . . 1 O . . . | $$ | . . X , X O . . . O . . X . . X O . . | $$ | . . . . . O . O . . . . X O X . O . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

So it seems that black spread its playouts a little more on this move. Maybe it's an indication that it was trickier?

Website · Post by **Chew Terr** » Thu Jun 23, 2011 12:37 pm

Click Here To Show Diagram Code: [go]$$Bcm21 $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . O . . . . . . . . . | $$ | . . . O . O . . . , . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . O . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . 0 . . . | $$ | . . . . X X . X . . . . . . X O . . . | $$ | . . X , X O . . . O . . X . . X O . . | $$ | . . . . . O . O . . . . X O X . O . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

emeraldemon · Post by **emeraldemon** » Thu Jun 23, 2011 2:15 pm

Click Here To Show Diagram Code: [go]$$Bcm31 $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . O . . . . . . . . . | $$ | . . . O . O . . . , . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . O . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . O 1 . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . X X . X . . . . . . X O . . . | $$ | . . X , X O . . . O . . X . . X O . . | $$ | . . . . . O . O . . . . X O X . O . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

I found a new trick Fuego can do: it can show a rough "territory estimate". Each point is colored based on how often it is controlled by white or black in the playouts. I don't think it hurts anything if Chew Terr sees:

: fuego_estimate.png (370.62 KiB) Viewed 8055 times

hyperpape · Post by **hyperpape** » Thu Jun 23, 2011 4:07 pm

emeraldemon wrote:
value: 50.5%

As predicted last time, not too much to say.

emeraldemon · Post by **emeraldemon** » Thu Jun 23, 2011 4:43 pm

hyperpape wrote:
I assume that implies that it gave the same sequence of moves as followup.

hyperpape · Post by **hyperpape** » Thu Jun 23, 2011 4:49 pm

emeraldemon · Post by **emeraldemon** » Thu Jun 23, 2011 5:03 pm

hyperpape wrote:
I imagine that this game is a fair bit of work for you, but I ask because I'm curious about how consistent Fuego's evaluations are. Three questions seem interesting to me: [list=1][*] If you run the same position twice, how likely are you to get the same first move? [*] If you run the same position multiple times, how consistent is the tail of the sequence? [*] As you add time, what happens to the tail of the sequence? Does it start to converge so that the later moves match what Fuego would play if you just handed it that position?

hyperpape wrote:While I'm talking about my curiousity, did you have any trouble compiling Fuego? I'm using Ubuntu 10.10 and got some errors when I tried, perhaps access to certain directories? I rarely compile anything myself, and it was awhile ago, so my memory is quite hazy.

I did have some issues compiling; I don't remember the exact details, but Fuego requires a specific version of boost, and I think I had a slightly newer version of one of the libraries. In the end I think I changed a flag somewhere in a boost header. If you don't have boost installed, that would do it. If you want to try again, I can help debug as best I can. (I assume Ubuntu has boost through apt-get, but I'm not sure what it would require).

Post by **Joaz Banbeck** » Thu Jun 23, 2011 8:08 pm

Dusk Eagle · Post by **Dusk Eagle** » Thu Jun 23, 2011 8:27 pm

topazg · Post by **topazg** » Fri Jun 24, 2011 1:26 am

Website · Post by **Chew Terr** » Fri Jun 24, 2011 8:00 am

Click Here To Show Diagram Code: [go]$$Bcm31 $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . O . . . . . . . . . | $$ | . . . O . O . . . , . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . O . . . . . . . . . . . 2 . . | $$ | . . X . . . . . . . . . . . . O X . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . X X . X . . . . . . X O . . . | $$ | . . X , X O . . . O . . X . . X O . . | $$ | . . . . . O . O . . . . X O X . O . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

I was SORELY tempted to activate the sequence I posted about last time (psst nobody tell Fuego about it), but that would be really ignoring what I was trying to make. So I'll respond locally. If my opponent takes gote to fix his shape in the bottom right, then the mistake will have punished itself. The reason I'm taking the hane, which originally felt too aggressive, is that the ladder is mine. It's worth risking ladder-breakers and such because it lets me be more aggressive against the black group on the right.

emeraldemon · Post by **emeraldemon** » Fri Jun 24, 2011 9:35 am

Click Here To Show Diagram Code: [go]$$Bcm31 $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . O . . . . . . . . . | $$ | . . . O . O . . . , . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . O . . . . . . . . . . 3 O . . | $$ | . . X . . . . . . . . . . . . O X . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . X X . X . . . . . . X O . . . | $$ | . . X , X O . . . O . . X . . X O . . | $$ | . . . . . O . O . . . . X O X . O . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

value: 51.0%

In previous playouts, Fuego was trying the hane underneath, so maybe it just recently decided this was a good move? It was initially even more positive about this move, rating it at 51.4%, but it eventually lowered down after more playouts.

Click Here To Show Diagram Code: [go]$$Bcm33 $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . O . . . . . . . . . | $$ | . . . O . O . . . , . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . 6 2 . . | $$ | . . . . O . . . . . . . . . . 1 O . . | $$ | . . X . . . . . . . . . . . 5 O X . . | $$ | . . . . . . . . . . . . . . . 7 X . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . 4 . . | $$ | . . . . . . . . . . . . . . . O . 3 . | $$ | . . . . X X . X . . . . . . X O . . . | $$ | . . X , X O . . . O . . X 9 . X O . . | $$ | . . . . . O . O . . . . X O X . O . . | $$ | . . . . . . . . . . . . . 8 . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

A few notes about other comments:

Click Here To Show Diagram Code: [go]$$Bcm31 $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . O . . . . . . . . . | $$ | . . . O . O . . . , . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . . . . | $$ | . . . . O . . . . . . . . . . 3 O . . | $$ | . . X . . . . . . . . . . . . O X . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . d O . . . | $$ | . . . . X X . X . a . . . . X O . . . | $$ | . . X , X O . . . O . . X b . X O . . | $$ | . . . . . O . O . . . . X O X . O . . | $$ | . . . . . . . . . . . . . c . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

As for topazg's move at "a", it considered it a tiny bit (550 playouts), but didn't like it that much. The other question is whether or not it sees the sequence Chew wants to play. I'm not sure: It looked at "b" 4327 times, compared to 148,108 for "c" and 1,063,404 for "d". It seems strange to me that Fuego was most concerned about "d", because that looks like aji-keshi to me. And it also seems to me that pushing at "b" is better for chew than descending directly:

Click Here To Show Diagram Code: [go]$$Wcm1 What Fuego expects $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . X X . X . . . . . . X O . . . | $$ | . . X , X O . . . O . . X 2 . X O . . | $$ | . . . . . O . O . . . . X O X . O . . | $$ | . . . . . . . . . . . . . 1 . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Click Here To Show Diagram Code: [go]$$Wcm1 What Chew expects $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . X X . X . . . . a 2 X O . . . | $$ | . . X , X O . . . O . . X 1 b X O . . | $$ | . . . . . O . O . . . . X O X . O . . | $$ | . . . . . . . . . . . . . 3 . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

There's obviously the cutting point at "a" and the big endgame of "b" in Chew's diagram. One difference is that descending directly seems more sente, maybe just leaving those two stones for aji, but I still think Chew's diagram is preferable.

All that said, the cut looks quite big, and Chew will have a tough decision on whether or not to respond on the right.

topazg · Post by **topazg** » Fri Jun 24, 2011 12:04 pm

Website · Post by **Chew Terr** » Fri Jun 24, 2011 12:20 pm

Click Here To Show Diagram Code: [go]$$Bcm31 $$ --------------------------------------- $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . O . . . . . . . . . | $$ | . . . O . O . . . , . . . . . O . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . X . . . . . . . . . . . . . 4 . . | $$ | . . . . O . . . . . . . . . . X O . . | $$ | . . X . . . . . . . . . . . . O X . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . X . . . . . , . . . . . , . . . | $$ | . . . . . . . . . . . . . . . . X . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . O . . . | $$ | . . . . X X . X . . . . . . X O . . . | $$ | . . X , X O . . . O . . X . . X O . . | $$ | . . . . . O . O . . . . X O X . O . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . | $$ ---------------------------------------[/go]

Like I said, I played allowing the cut because the ladder doesn't work. If I can isolate the cutting stone, I have an opponent's weak group near a ton of potential territory for me, and the right-side black group's not looking all too stable, either.

Loons · Post by **Loons** » Sat Jun 25, 2011 4:28 am

@Emeraldemon

Life In 19x19

124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)

Re: 124. Chew (3k) vs Fuego (Bot)