Life In 19x19

Life in 19x19. Go, Weiqi, Baduk... Thats the life.
https://lifein19x19.com/

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

https://lifein19x19.com/viewtopic.php?t=3960

Page 6 of 11

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

Posted: Sat Jun 25, 2011 6:42 am
by lightvector
Loons wrote:@Emeraldemon
Sorry if I missed the answer already, but is Fuego capable of laddering stones? (Given its decisions come from random playouts...)


I don't know about Fuego, but most MCTS bots have special-case code for ladders. But this is one of those things that is extremely easy to get wrong when implementing a bot, or get it to work only in some cases and not others, etc.

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

Posted: Mon Jun 27, 2011 9:04 am
by emeraldemon
Click Here To Show Diagram Code
[go]$$Bcm31
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . O . . |
$$ | . . . . O . . . . . . . . . . X O . . |
$$ | . . X . . . . . . . . . . . 5 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: 51.1%

Favored sequence:
Click Here To Show Diagram Code
[go]$$Bcm35
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . 2 O . . |
$$ | . . . . O . . . . . . . . . . X O . . |
$$ | . . X . . . . . . . . . . . 1 O X . . |
$$ | . . . . . . . . . . . . . . . 3 X . . |
$$ | . . . X . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . 6 0 . . . |
$$ | . . . . . . . . . . . . . 4 5 O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X 7 9 X O . . |
$$ | . . . . . O . O . . . . X O X 8 O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]




Here are the top moves it considered:

Click Here To Show Diagram Code
[go]$$Bcm31 Black's considerations
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . O . . |
$$ | . . . . O . . . . . . . . . . X O . . |
$$ | . . X . . . . . . . . . . . a O X . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . X . . . . . , . . c . . , . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . . X O . . |
$$ | . . . . . O . O . . . . X O X . O . . |
$$ | . . . . . . . . . d . . . . . b . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


a: 73%
b: 12%
c: 7%
d: 3%

@Loons: The documentation is pretty sparse, it doesn't say anything about ladders one way or the other. Grepping through the source I can find some code that seems to involve ladders (there's a CheckLadders function anyway), but I don't really feel like figuring out what it does right now. I spend enough time reading c++ for my real job :-|

Just to satisfy my own curiosity, I played Q11 and asked Fuego to think again, and it immediately imagined this:

Click Here To Show Diagram Code
[go]$$Bcm1
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . O . . |
$$ | . . . . O . . . . . . . . . . X O . . |
$$ | . . X . . . . . . . . . . . X O X . . |
$$ | . . . . . . . . . . . . . X O O X . . |
$$ | . . . X . . . . . , . . X O O X . . . |
$$ | . . . . . . . . . . . X O O X . X . . |
$$ | . . . . . . . . . . X O O X . . . . . |
$$ | . . . . . . . . . X O O X . . . . . . |
$$ | . . . . . . . . X O O X . . . O . . . |
$$ | . . . . X X . X . O X . . . X O . . . |
$$ | . . X , X O . . . O 1 . X . . X O . . |
$$ | . . . . . O . O . . . . X O X . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


After about a minute it decided it didn't like that sequence, and started trying other moves for black. So whether or not there's special ladder code, it's clearly capable of reading this ladder. For whatever reason, it just thought pulling out that stone is bad for white. It didn't think very long about it, only 0.6% of playouts on Q11. Notice that c in the considerations diagram seems like a ladder-breaker move, so I do think it was thinking some about that idea.

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

Posted: Mon Jun 27, 2011 9:11 am
by Chew Terr
Click Here To Show Diagram Code
[go]$$Bcm31
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . O . . |
$$ | . . . . O . . . . . . . . . . X O . . |
$$ | . . X . . . . . . . . . . . X O X . . |
$$ | . . . . . . . . . . . . . . . 6 X . . |
$$ | . . . X . . . . . , . . . . . , . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . . X O . . |
$$ | . . . . . O . O . . . . X O X . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]



Dunno if the cut was actually worthwhile, but I'm fairly convinced I won't die locally.

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

Posted: Mon Jun 27, 2011 8:39 pm
by emeraldemon
Click Here To Show Diagram Code
[go]$$Bcm31
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . O . . |
$$ | . . . . O . . . . . . . . . . X O . . |
$$ | . . X . . . . . . . . . . . X O X . . |
$$ | . . . . . . . . . . . . . . . O X . . |
$$ | . . . X . . . . . , . . . . . 7 . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . . X O . . |
$$ | . . . . . O . O . . . . X O X . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


value: 50.7%
A slight decrease.

Click Here To Show Diagram Code
[go]$$Bcm37
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O 0 . . |
$$ | . . . . . . . . . . . . . . . 8 5 6 . |
$$ | . . X . . . . . . . . . . . . 3 O . . |
$$ | . . . . O . . . . . . . . . . X O 9 . |
$$ | . . X . . . . . . . . . . . X O X . . |
$$ | . . . . . . . . . . . . . . 2 O X . . |
$$ | . . . X . . . . . , . . . . . 1 . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . 7 X O . . |
$$ | . . . . . O . O . . . . X O X 4 O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


As expected, Fuego isn't going to play out the ladder. They say "beginners play atari", but I think this is a good one, since it helps fix black's shape. I don't have time for too much analysis unfortunately.

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

Posted: Mon Jun 27, 2011 11:30 pm
by ez4u
emeraldemon wrote:...
... Here are the top moves it considered:

Click Here To Show Diagram Code
[go]$$Bcm31 Black's considerations
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . O . . |
$$ | . . . . O . . . . . . . . . . X O . . |
$$ | . . X . . . . . . . . . . . a O X . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . X . . . . . , . . c . . , . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . . X O . . |
$$ | . . . . . O . O . . . . X O X . O . . |
$$ | . . . . . . . . . d . . . . . b . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


a: 73%
b: 12%
c: 7%
d: 3%
...


For :b35: did it consider any plays around :b1:, which would create the alternative ladder shown? If so, how did it expect White to react?
Click Here To Show Diagram Code
[go]$$Bc "d" connects at 5
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . 2 1 . . . . O . . . |
$$ | . . . . . . . . . . . . . . . 4 . . . |
$$ | . . X . . . . . . . . . . g 9 3 O . . |
$$ | . . . . O . . . . . . . . f 8 X O . . |
$$ | . . X . . . . . . . . . e b 5 O X . . |
$$ | . . . . . . . . . . . . . a 0 6 X . . |
$$ | . . . X . . . . . , . . . . 7 c . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . . X O . . |
$$ | . . . . . O . O . . . . X O X . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

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

Posted: Tue Jun 28, 2011 2:17 am
by flOvermind
emeraldemon wrote:
value: 50.7%
A slight decrease.

Click Here To Show Diagram Code
[go]$$Bcm37
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O 0 . . |
$$ | . . . . . . . . . . . . . . . 8 5 6 . |
$$ | . . X . . . . . . . . . . . . 3 O . . |
$$ | . . . . O . . . . . . . . . . X O 9 . |
$$ | . . X . . . . . . . . . . . X O X . . |
$$ | . . . . . . . . . . . . . . 2 O X . . |
$$ | . . . X . . . . . , . . . . . 1 . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . 7 X O . . |
$$ | . . . . . O . O . . . . X O X 4 O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


As expected, Fuego isn't going to play out the ladder. They say "beginners play atari", but I think this is a good one, since it helps fix black's shape. I don't have time for too much analysis unfortunately.


Is it just me, or is there a lot of wishful thinking in the Fuego's reasoning? Why on earth would white play 40? Especially when Fuego itself thinks it doesn't need to be answered immediately...

That's not the first time I see that kind of move in the bot analysis. That may hint at a weakness in the bot, but I'm not quite sure what that weakness would be, and how (or even if) you can exploit it ;)

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

Posted: Tue Jun 28, 2011 7:47 am
by Chew Terr
Click Here To Show Diagram Code
[go]$$Bcm31
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . O . . |
$$ | . . . . O . . . . . . . . . . X O . . |
$$ | . . X . . . . . . . . . . . X O X . . |
$$ | . . . . . . . . . . . . . . 8 O X . . |
$$ | . . . X . . . . . , . . . . . X . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . . X O . . |
$$ | . . . . . O . O . . . . X O X . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]



Okay, I suppose this buys Fuego's right-side group better shape. The only alternative I see is playing the ladder almost all the way out to get several forcing moves against my bottom middle group, but it doesn't feel worthwhile. So I'm assuming that black will not push more than one more time from each side before fixing the weak side's cut points and resuming the non-ladder game proper.

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

Posted: Thu Jun 30, 2011 9:49 am
by emeraldemon
Click Here To Show Diagram Code
[go]$$Bcm31
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . . O . . |
$$ | . . . . O . . . . . . . . . . X O . . |
$$ | . . X . . . . . . . . . . . X O X . . |
$$ | . . . . . . . . . . . . . . O O X . . |
$$ | . . . X . . . . . , . . . 9 . X . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . . X O . . |
$$ | . . . . . O . O . . . . X O X . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


Fuego took 28:04 to decide this move, which is the longest its ever taken to decide a move.
Value dropped again to 49.8% .

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


Moves by consideration time:
a - 51%
b - 44%
c - 3%

The time spent on c was mostly at the beginning I think, when it was still considering playing out the ladder. The reason it took so long was that Fuego couldn't decide whether or not to push at b or go directly for a.

Predicted sequence:

Click Here To Show Diagram Code
[go]$$Bcm39
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . X . . . . . . . . . . . . 6 O . . |
$$ | . . . . O . . . . . . . . . 2 X O . . |
$$ | . . X . . . . . 9 . . . . 3 X O X . . |
$$ | . . . . . . . . . . . . . 5 O O X . . |
$$ | . . . X . . . . . , . . . 1 . X . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . 7 . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . . X O . . |
$$ | . . . . . O . O . . 4 . X O X . O . . |
$$ | . . . . . . . . . . . 8 . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]

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

Posted: Thu Jun 30, 2011 9:51 am
by emeraldemon
ez4u wrote:
For :b35: did it consider any plays around :b1:, which would create the alternative ladder shown? If so, how did it expect White to react?
Click Here To Show Diagram Code
[go]$$Bc "d" connects at 5
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . 2 1 . . . . O . . . |
$$ | . . . . . . . . . . . . . . . 4 . . . |
$$ | . . X . . . . . . . . . . g 9 3 O . . |
$$ | . . . . O . . . . . . . . f 8 X O . . |
$$ | . . X . . . . . . . . . e b 5 O X . . |
$$ | . . . . . . . . . . . . . a 0 6 X . . |
$$ | . . . X . . . . . , . . . . 7 c . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . . X O . . |
$$ | . . . . . O . O . . . . X O X . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]




No it hasn't liked any moves around the top (except some plays in the upper left corner). I didn't see that latter either, it's a pretty sneaky one :tmbup:

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

Posted: Thu Jun 30, 2011 9:55 am
by emeraldemon
flOvermind wrote:
emeraldemon wrote:
value: 50.7%
A slight decrease.

Click Here To Show Diagram Code
[go]$$Bcm37
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . , . . . . . O 0 . . |
$$ | . . . . . . . . . . . . . . . 8 5 6 . |
$$ | . . X . . . . . . . . . . . . 3 O . . |
$$ | . . . . O . . . . . . . . . . X O 9 . |
$$ | . . X . . . . . . . . . . . X O X . . |
$$ | . . . . . . . . . . . . . . 2 O X . . |
$$ | . . . X . . . . . , . . . . . 1 . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . 7 X O . . |
$$ | . . . . . O . O . . . . X O X 4 O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]


As expected, Fuego isn't going to play out the ladder. They say "beginners play atari", but I think this is a good one, since it helps fix black's shape. I don't have time for too much analysis unfortunately.


Is it just me, or is there a lot of wishful thinking in the Fuego's reasoning? Why on earth would white play 40? Especially when Fuego itself thinks it doesn't need to be answered immediately...

That's not the first time I see that kind of move in the bot analysis. That may hint at a weakness in the bot, but I'm not quite sure what that weakness would be, and how (or even if) you can exploit it ;)


Well, Fuego isn't capable of wishful thinking in the way humans are: it uses the exact same algorithm do choose both white and black moves, so it can't hope for the opponent to play poorly, the way a human can. That said, it can certainly miss tesuji and play poorly in the tree and in the playouts. The hope is that it will make mistakes in roughly equal amount for both sides, so that the board evaluation is correct.

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

Posted: Thu Jun 30, 2011 5:04 pm
by ez4u
emeraldemon wrote:
ez4u wrote:
For :b35: did it consider any plays around :b1:, which would create the alternative ladder shown? If so, how did it expect White to react?
Click Here To Show Diagram Code
[go]$$Bc "d" connects at 5
$$ ---------------------------------------
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . O . . . . . . . . . |
$$ | . . . O . O . . . 2 1 . . . . O . . . |
$$ | . . . . . . . . . . . . . . . 4 . . . |
$$ | . . X . . . . . . . . . . g 9 3 O . . |
$$ | . . . . O . . . . . . . . f 8 X O . . |
$$ | . . X . . . . . . . . . e b 5 O X . . |
$$ | . . . . . . . . . . . . . a 0 6 X . . |
$$ | . . . X . . . . . , . . . . 7 c . . . |
$$ | . . . . . . . . . . . . . . . . X . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . O . . . |
$$ | . . . . X X . X . . . . . . X O . . . |
$$ | . . X , X O . . . O . . X . . X O . . |
$$ | . . . . . O . O . . . . X O X . O . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ | . . . . . . . . . . . . . . . . . . . |
$$ ---------------------------------------[/go]




No it hasn't liked any moves around the top (except some plays in the upper left corner). I didn't see that latter either, it's a pretty sneaky one :tmbup:


Generally how many alternative "next move" candidates does fuego consider? It obviously spends most of its time on a select few, but how many candidates receive at least some analysis in each turn?

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

Posted: Thu Jun 30, 2011 5:50 pm
by emeraldemon
ez4u:

Here's an image showing how many playouts each point on the board received. The larger numbers get cut off, for example the game move was visited 3114430 times. But this should give the basic idea.

playouts.png
playouts.png (405.14 KiB) Viewed 9784 times

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

Posted: Thu Jun 30, 2011 6:48 pm
by ez4u
@emeraldemon

Can you see the meaning of the 23 playouts that great majority of points receive? Is there some special pattern? Obviously 23 playouts is not going to go very far! :)

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

Posted: Fri Jul 01, 2011 1:45 am
by perceval
@ez4u:

my guess would be that its a default minimal value, which seems rather low:
around 20-25 is the threshold where a bernouilli distribution begin to be approximated by a normal distribution so roughly:
(here we consider that a given move give a probaility p that a random playout after the move will give a won game for B, and we want to approximate p by playing a limited number of playout )
The error is of order 0.5/sqrt(23) which is approx 0.1. This is probably the reason why they picked 23 as the default value .This seems like a rather big error bar.

It might be interesting to check if this min number of playout can be tune up and if it would affect the result.

Maybe after those min 23 playouts fuego excludes all moves with a p less than 0.4 ? BUT considering the number of candidate moves it is probable than at least one as been wrongly estimated by more than 10%. (of course fuego implementors know their jobs so we can hope they have taken that into account)

Here the final evaluation for the best move is within 0.01 of 0.5 so it would be useful to have this kind of error bars on the estimation of all points as well.


having a 1% error interval means around 2500 playout.... which is a lot (to gain a factor of 10 in accuracy you need 100 times more playout). But it needs to be compared to the in depth exploration of the many variation for the preferred move: you have roughly 300 plays looking "not interesting" so exploring them all at 2500 playout would be 750.000 play out which is still far less than playout for the top move (around 3000000 according to esmeraldemon ).

So maybe fuego will run at roughly the same speed with this value tuned up to 2500 ?
Not sure, though, that it would bring a better playing strength.

Does esmeralddemon has the time and willingness to give it a a shot for 1 move ?

Edited out the most embarassing typos

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

Posted: Fri Jul 01, 2011 1:47 am
by topazg
perceval wrote:@ez4u:

my guess would be that its a default minimal value, which seems rather low:
around 20-25 is the threshold where a bernouilli distribution begin to be approximated by a normal distribution so roughly:
(here we consider that a given move give a probaility p that a randaom playout after the move will give a won game for B, and we want to approximate p by playing a limited number of playout

the error is of order 0.5/sqrt(23) which is approx 0.1. This seems is a rather big error bar.
it might be interesting to check if this min number of playout can be tune up and if it would affect the result.
having a 1% error interval means around 2500 playout.... which is a lot (to gain a factor of 10 accuracy you need 100 more playout). but it needs to be compared to the in depth exploration of the may vars: you have roughtly 300 plays looking "not interesting" so exploring them all at 2500 playout would be 750.000 play out which is still far less than playout for the top move.

So maybe fuego will run at roughtly the same speed with this value tuned up ?
Not sure, though, that it would bring a better playing strenght.

Does esmeralddemon as the time and willingness to give it a a shot for 1 move ?


If we're talking about sufficient data points to improve play beyond chance, then I suspect 1 move won't tell us much ;)
All times are UTC-07:00
Page 6 of 11

Powered by phpBB® Forum Software © phpBB Limited