It is currently Tue Mar 19, 2024 2:59 am

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 13 posts ] 
Author Message
Offline
 Post subject: Curious result in area scoring rule.
Post #1 Posted: Mon Feb 21, 2022 8:44 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X b . O . O . |
$$| X O O O O O O |
$$| X a . . O M M |
$$| X O O O O M M |
$$| M M M M M M M |
$$| M M M M M M M |
$$| M M M M M M M |
$$ ----------------
$$[/go]

Under japanese rule I guess a white move at "b" cannot better than a white move at "a".
But what about area scoring rules with superko rule?
Can you build in the marked area, an environment such that the best move is a white move at "b" ?
A challenge for all those interesting by curious results.

Top
 Profile  
 
Offline
 Post subject: Re: Curious result in area scoring rule.
Post #2 Posted: Mon Mar 14, 2022 11:44 am 
Beginner

Posts: 16
Liked others: 0
Was liked: 2
Rank: Tygem 2d
I tried to no avail. It is easy to find a position where b is as good as a, but not strictly better. I hoped it had to use some of these cool ideas about the forcing pass lemma but I can't fit it in this board. With larger options for a and b I think it is doable.

Do you have a solution?

Top
 Profile  
 
Offline
 Post subject: Re: Curious result in area scoring rule.
Post #3 Posted: Sat Apr 02, 2022 8:15 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Criado wrote:
I tried to no avail. It is easy to find a position where b is as good as a, but not strictly better. I hoped it had to use some of these cool ideas about the forcing pass lemma but I can't fit it in this board. With larger options for a and b I think it is doable.

Do you have a solution?


Here is my solution (I hope it works well):

Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X b . O . O . |
$$| X O O O O O O |
$$| X a . . O X . |
$$| X O O O O X X |
$$| X X O X X X X |
$$| X . X X O O O |
$$| . X X . X O . |
$$ ----------------
$$[/go]

Top
 Profile  
 
Offline
 Post subject: Re: Curious result in area scoring rule.
Post #4 Posted: Sat Apr 02, 2022 9:03 am 
Lives in sente

Posts: 896
Liked others: 22
Was liked: 168
Rank: panda 5 dan
IGS: kvasir
Please explain.

As far as I can tell white can get 4 of the marked intersections and it doesn't appear to matter which is played first. What does superko have to do with it?
Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X C C O . O . |
$$| X O O O O O O |
$$| X C C C O X C |
$$| X O O O O X X |
$$| X X O X X X X |
$$| X . X X O O O |
$$| . X X . X O . |
$$ ----------------
$$[/go]


For example
Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X 2 5 O . O . |
$$| X O O O O O O |
$$| X 3 . 1 O X 4 |
$$| X O O O O X X |
$$| X X O X X X X |
$$| X . X X O O O |
$$| . X X . X O . |
$$ ----------------
$$[/go]

Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X 1 . O . O . |
$$| X O O O O O O |
$$| X 2 3 . O X 4 |
$$| X O O O O X X |
$$| X X O X X X X |
$$| X . X X O O O |
$$| . X X . X O . |
$$ ----------------
$$[/go]

Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X 2 4 O . O . |
$$| X O O O O O O |
$$| X 1 . . O X 3 |
$$| X O O O O X X |
$$| X X O X X X X |
$$| X . X X O O O |
$$| . X X . X O . |
$$ ----------------
$$[/go]

Top
 Profile  
 
Offline
 Post subject: Re: Curious result in area scoring rule.
Post #5 Posted: Sat Apr 02, 2022 9:27 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
kvasir wrote:
Please explain.

As far as I can tell white can get 4 of the marked intersections and it doesn't appear to matter which is played first. What does superko have to do with it?
Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X C C O . O . |
$$| X O O O O O O |
$$| X C C C O X C |
$$| X O O O O X X |
$$| X X O X X X X |
$$| X . X X O O O |
$$| . X X . X O . |
$$ ----------------
$$[/go]


For example
Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X 2 5 O . O . |
$$| X O O O O O O |
$$| X 3 . 1 O X 4 |
$$| X O O O O X X |
$$| X X O X X X X |
$$| X . X X O O O |
$$| . X X . X O . |
$$ ----------------
$$[/go]

Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X 1 . O . O . |
$$| X O O O O O O |
$$| X 2 3 . O X 4 |
$$| X O O O O X X |
$$| X X O X X X X |
$$| X . X X O O O |
$$| . X X . X O . |
$$ ----------------
$$[/go]

Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X 2 4 O . O . |
$$| X O O O O O O |
$$| X 1 . . O X 3 |
$$| X O O O O X X |
$$| X X O X X X X |
$$| X . X X O O O |
$$| . X X . X O . |
$$ ----------------
$$[/go]


The point is that white can get 5 of the marked intersections by:
Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X 1 . O . O . |
$$| X O O O O O O |
$$| X 2 3 . O X 5 |
$$| X O O O O X X |
$$| X X O X X X X |
$$| X . X X O O O |
$$| . X X . X O 4 |
$$ ----------------
$$[/go]

because black would have no move after:
Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X 1 . O . O . |
$$| X O O O O O O |
$$| X 2 3 . O X 4 |
$$| X O O O O X X |
$$| X X O X X X X |
$$| X . X X O O O |
$$| . X X 5 X O . |
$$ ----------------
$$[/go]


Compare with the sequence
Click Here To Show Diagram Code
[go]$$W White to play
$$ ----------------
$$| X 2 4 O . O . |
$$| X O O O O O O |
$$| X 1 6 . O X 3 |
$$| X O O O O X X |
$$| X X O X X X X |
$$| X . X X O O O |
$$| . X X 5 X O . |
$$ ----------------
$$[/go]
where black can play the move :b6:


This post by Gérard TAILLE was liked by: le_4TC
Top
 Profile  
 
Offline
 Post subject: Re: Curious result in area scoring rule.
Post #6 Posted: Sat Apr 02, 2022 10:17 am 
Lives in sente

Posts: 896
Liked others: 22
Was liked: 168
Rank: panda 5 dan
IGS: kvasir
I see.

Why again doesn't pass lift superko?

Top
 Profile  
 
Offline
 Post subject: Re: Curious result in area scoring rule.
Post #7 Posted: Sat Apr 02, 2022 10:25 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
kvasir wrote:
I see.

Why again doesn't pass lift superko?

I do not know.
I agree with you. I have the same question.

Top
 Profile  
 
Offline
 Post subject: Re: Curious result in area scoring rule.
Post #8 Posted: Sun Apr 03, 2022 2:32 pm 
Lives in gote

Posts: 445
Liked others: 0
Was liked: 37
Gérard TAILLE wrote:
kvasir wrote:
Why again doesn't pass lift superko?
I do not know.
I agree with you. I have the same question.

Because then it wouldn't protect area scoring from all unbalanced repetition (like the non-ko line in sending3-returning1). Also the whole idea of superko is to look for position changes (instead of elapsed time as with normal ko, ie. no immediate recapture).

Top
 Profile  
 
Offline
 Post subject: Re: Curious result in area scoring rule.
Post #9 Posted: Sun Apr 03, 2022 5:40 pm 
Lives in sente

Posts: 896
Liked others: 22
Was liked: 168
Rank: panda 5 dan
IGS: kvasir
Ok. Preventing unbalanced repetition then is the reason.

Maybe there is a good definition somewhere of what a superko is and what it is not? I'd be perfectly happy with a strict definition but Sensei's library on one hand seems to allow for almost anything when it says (https://senseis.xmp.net/?Superko):
Quote:
Superko rules prohibit recurrences of an earlier board position under certain circumstances.


Pass is even discussed as something affecting superko in that article. Pass is a "circumstance", so this can be allowed with the SL definition.

Off topic, doubly off topic I guess, but elsewhere on SL it is claimed that ING rules have superko. This seems odd to me, maybe "hot stones" are the "circumstance". Next one could try to argue that the basic ko rule forbids "repetition of an earlier board position under certain circumstances", the specific circumstances have something to do with ko shapes in the position. :scratch:

Top
 Profile  
 
Offline
 Post subject: Re: Curious result in area scoring rule.
Post #10 Posted: Mon Apr 04, 2022 6:27 am 
Lives in sente

Posts: 757
Liked others: 114
Was liked: 916
Rank: maybe 2d
If you want to avoid this kind of pathology, but still break all cycles, guarantee that games are finite, and handle the "basic" situations the way that most players' intuitions think they should be handled (of course there are always still more complex rules beasts), you can use Bill Spight's innovation regarding ko bans and game end:

Use a superko rule as normal, except:
1. Passes lift all ko bans (superko only stops repetitions where neither player passed in between).
2. Regardless of what passes occur in between or not, the game ends when a player passes in a position that player previously passed in. (and this is the only game end condition now. So in normal games it means that the game usually ends with 3 consecutive passes, not 2).

#1 is of course targeted at breaking the kinds of positions in this thread.

Intuition for #2 is to say that even if ko bans are lifted, we can still end the game in the case of cycles involving enough passes, because passes usually indicate one or both players have nothing better to do.

In normal cases with 3 passes, the "hand talk" is like:
P1 passes: "I have nothing I want to do *except* maybe I'm just waiting to recapture a ko and have no other non-suicidal moves"
P2 passes: "I have nothing I want to do, not even with ko bans lifted"
P1 passes: "I have nothing I want to do, not even with ko bans lifted - okay the game should end since neither player has useful things to do".

In cases with unbalanced repetition where only 1 player is passing while the other player cycles, the "hand talk" is like:
P1 passes: "I have nothing I want to do *except* maybe I'm just waiting to recapture a ko and have no other non-suicidal moves"
P2 plays some moves that force P1 to respond, but they only lead back to the same position where P1 passes originally.
P1 passes: "I still have nothing I want to do, and if the best that you, P2, could do even with all ko bans lifted and a free extra turn was merely to cycle back to the same position where I can pass yet again, then obviously you're just wasting time, the game should end now."


This post by lightvector was liked by 2 people: le_4TC, luigi
Top
 Profile  
 
Offline
 Post subject: Re: Curious result in area scoring rule.
Post #11 Posted: Mon Apr 04, 2022 10:21 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
lightvector wrote:
If you want to avoid this kind of pathology, but still break all cycles, guarantee that games are finite, and handle the "basic" situations the way that most players' intuitions think they should be handled (of course there are always still more complex rules beasts), you can use Bill Spight's innovation regarding ko bans and game end:

Use a superko rule as normal, except:
1. Passes lift all ko bans (superko only stops repetitions where neither player passed in between).
2. Regardless of what passes occur in between or not, the game ends when a player passes in a position that player previously passed in. (and this is the only game end condition now. So in normal games it means that the game usually ends with 3 consecutive passes, not 2).

#1 is of course targeted at breaking the kinds of positions in this thread.

Intuition for #2 is to say that even if ko bans are lifted, we can still end the game in the case of cycles involving enough passes, because passes usually indicate one or both players have nothing better to do.

In normal cases with 3 passes, the "hand talk" is like:
P1 passes: "I have nothing I want to do *except* maybe I'm just waiting to recapture a ko and have no other non-suicidal moves"
P2 passes: "I have nothing I want to do, not even with ko bans lifted"
P1 passes: "I have nothing I want to do, not even with ko bans lifted - okay the game should end since neither player has useful things to do".

In cases with unbalanced repetition where only 1 player is passing while the other player cycles, the "hand talk" is like:
P1 passes: "I have nothing I want to do *except* maybe I'm just waiting to recapture a ko and have no other non-suicidal moves"
P2 plays some moves that force P1 to respond, but they only lead back to the same position where P1 passes originally.
P1 passes: "I still have nothing I want to do, and if the best that you, P2, could do even with all ko bans lifted and a free extra turn was merely to cycle back to the same position where I can pass yet again, then obviously you're just wasting time, the game should end now."


By beginning your post by "If you want to avoid this kind of pathology..." you seem quite careful and I agree with that.
As far as I am concerned it is not really a pathology but simply a consequence of the rule which does not hurt me for two reasons:
1) It will never happen in practice on a 19x19 board and I have even great doubt it can happen in practice on a 9x9 board.
2) On a smaller board this result seems even interesting and remind me the zugswang in chess. I you like very small boards what really harms by taking care of such possibility?
IOW it is for me only a curiosity of the rule as suggested by the title I chose.

Top
 Profile  
 
Offline
 Post subject: Re: Curious result in area scoring rule.
Post #12 Posted: Mon Apr 04, 2022 1:59 pm 
Lives in gote

Posts: 445
Liked others: 0
Was liked: 37
There are some drawbacks with Bill's approach though. It doesn't seem to work well on 0-sided ko (reducing half point games to no-pass go by randomly stopping in one of the oscillating states). Also seems somewhat superfluous, considering three major types of repetition:

1. unbalanced repetition (not a thing in territory scoring, needs prevention for area scoring)
2. moonshine repetition (minor problem for both scoring, best prevented)
3. normal repetition (things like triple ko, no real problem for either scoring method)

If you handle type 1 by another rule, why bother with superko at all? Its remaining functionality is either miniscule (2) or outright undesired (3), at least for the Asian game. Even if used it seems better to just "minimal superko" (no repetition of the position of the last successive two passes).

Top
 Profile  
 
Offline
 Post subject: Re: Curious result in area scoring rule.
Post #13 Posted: Sat Apr 09, 2022 6:35 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Click Here To Show Diagram Code
[go]$$W White to play
$$ --------------------
$$| X a X X X O . O . |
$$| X O O O O O O O O |
$$| X b X X O M M M M |
$$| X O O O O M M M M |
$$| M M M M M M M M M |
$$| M M M M M M M M M |
$$| M M M M M M M M M |
$$| M M M M M M M M M |
$$| M M M M M M M M M |
$$ --------------------
$$[/go]
Now you have got the main idea can you resolve this new problem?
Assuming you use area counting and superko:
Can you build in the marked area an environment such that the best move is a white move at "b" ?
I confess it was not easy for me to solve this problem but eventually I managed to find a solution.

Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 13 posts ] 

All times are UTC - 8 hours [ DST ]


Who is online

Users browsing this forum: No registered users and 1 guest


You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot post attachments in this forum

Search for:
Jump to:  
Powered by phpBB © 2000, 2002, 2005, 2007 phpBB Group