It is currently Tue Mar 19, 2024 1:21 am

All times are UTC - 8 hours [ DST ]




Post new topic Reply to topic  [ 391 posts ]  Go to page Previous  1 ... 16, 17, 18, 19, 20  Next
Author Message
Offline
 Post subject: Re: Japonese counting
Post #361 Posted: Fri Sep 24, 2021 3:29 pm 
Lives in gote

Posts: 445
Liked others: 0
Was liked: 37
gennan wrote:
In area scoring, bent 4 in the corner might only live if there are no unremovable ko threats

Sure, but the question was specifically about bent4 + unremovable threat (see also this). Bent4 alone would not pose a question (would be dead) for the mentioned rules (confirmation with normal play + moonshine ko rule, and global enable).

But the bent4+seki case, as explained above, is very hard if not impossible to score correctly in territory scoring, whatever method you choose (error in one case of the other). And I doubt rarity means much here - most rules questions are about rarities and about the difference between being correct 99.9% or 100% of cases.

Top
 Profile  
 
Offline
 Post subject: Re: Japonese counting
Post #362 Posted: Mon Sep 27, 2021 10:21 am 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
Example 23

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

In the rule, Black is alive and White is dead.

I imagine the following sequence:

Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | 4 6 5 O O X . O X . .
$$ | O X O . O X . O X X .
$$ | 3 X X O O X X X O X .
$$ | X 1 X X X O O O O X .
$$ | 2 O X O O . . . . . .
$$ | O X X O . . . . . . .
$$ | X X O O . . . . . . .
$$ | O O O . . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------[/go]


Click Here To Show Diagram Code
[go]$$B :b9: pass-for ko, :w10: at :b7:
$$ -----------------------
$$ | 7 O . O O X . O X . .
$$ | 8 X O . O X . O X X .
$$ | X X X O O X X X O X .
$$ | X X X X X O O O O X .
$$ | O . X O O . . . . . .
$$ | . X X O . . . . . . .
$$ | X X O O . . . . . . .
$$ | O O O . . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------[/go]


Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | O O . O O X . O X . .
$$ | O X O . O X . O X X .
$$ | X X X O O X X X O X .
$$ | X X X X X O O O O X .
$$ | O . X O O . . . . . .
$$ | . X X O . . . . . . .
$$ | X X O O . . . . . . .
$$ | O O O . . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------[/go]
and I do not see how black can capture white

In addition black is in difficulty in normal play because white has two ko threats at the top.

Top
 Profile  
 
Offline
 Post subject: Re: Japonese counting
Post #363 Posted: Mon Sep 27, 2021 10:51 am 
Lives in gote

Posts: 445
Liked others: 0
Was liked: 37
This kind of defence does not work. It involves W losing stones in a perpetual cycle, thus sooner or later reaching -500 score (if played in normal play). The commentary briefly refers to this, and makes it clear such self-destroying is not acceptable even in analysis. Hypothetical play is about what could happen if the game would continue (without worrying about the cost of cleanup moves into territory).

The two ko threats are a different matter, it is true that in normal play they could prevent B from capturing, but the Japanese approach is ko-threat-less L/D ("what would happen if the game would continue w/o ko threats"). Which is not fair in several simpler cases already - but this is how Japanese rules are designed (90% because of moonshine, 10% because of bent4), taking "local life" to its extreme.

Top
 Profile  
 
Offline
 Post subject: Re: Japonese counting
Post #364 Posted: Mon Sep 27, 2021 12:37 pm 
Lives in sente

Posts: 1226
Liked others: 21
Was liked: 57
Rank: 1d
jann wrote:
This kind of defence does not work. It involves W losing stones in a perpetual cycle, thus sooner or later reaching -500 score (if played in normal play). The commentary briefly refers to this, and makes it clear such self-destroying is not acceptable even in analysis. Hypothetical play is about what could happen if the game would continue (without worrying about the cost of cleanup moves into territory).

The two ko threats are a different matter, it is true that in normal play they could prevent B from capturing, but the Japanese approach is ko-threat-less L/D ("what would happen if the game would continue w/o ko threats"). Which is not fair in several simpler cases already - but this is how Japanese rules are designed (90% because of moonshine, 10% because of bent4), taking "local life" to its extreme.


I know this problem of sending three returning two. But in confirmation phase I do not see where such sequence does not appear as an acceptable defense. I see nothing in the rule text itself and in example 23 is is simply said that Black is alive and White is dead without any explanation.

Anyway your interpretation seems logical.

Top
 Profile  
 
Offline
 Post subject: Re: Japonese counting
Post #365 Posted: Mon Sep 27, 2021 3:30 pm 
Lives in gote

Posts: 445
Liked others: 0
Was liked: 37
Gérard TAILLE wrote:
I see nothing in the rule text itself and in example 23 is is simply said that Black is alive and White is dead without any explanation.

Look again. Example 23 commentary refers back to "the purpose of the game in article 1", ie. to "take more territory". This is a roundabout way of saying -500 is not allowed even in confirmation.

But is also rather obvious: if confirmation would be played with different objectives and conditions than real game, that would be a theoretical flaw since then the players could choose sequences that could never happen and make no sense in real go. This was exactly the problem in your early rules where losing big groups was without consequence (global play without global enable => change in what the players can optimally play). You need confirmation as similar to real go as possible.

Top
 Profile  
 
Offline
 Post subject: Re: Japonese counting
Post #366 Posted: Tue Sep 28, 2021 12:46 am 
Lives in sente

Posts: 896
Liked others: 22
Was liked: 168
Rank: panda 5 dan
IGS: kvasir
I think the explanation in example 23 is referring to when white adds a move :w1: during the game. In that case black can capture 9 stones and give up 1, and it is actually seki like that.

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



Without the added move it is unclear to me why white is dead despite the cycle.

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


I think there are three theories.

:w1: Throwing away stones to prevent captures doesn't count (but I don't know why).
:w2: Cycles don't count (but I don't know why).
:w3: The status of two white stones cascades (but I don't know why).
===EDIT=== :w4: It is an error by the J89 authors


We can also play with this and create this position:

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


Are the black and white stones on top dead? It is a thought experiment about the reason for white being declared dead in example 23. For example does the cycle count in this position?

Top
 Profile  
 
Online
 Post subject: Re: Japonese counting
Post #367 Posted: Tue Sep 28, 2021 1:58 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
kvasir wrote:
Without the added move it is unclear to me why white is dead despite the cycle.

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

===EDIT=== :w4: It is an error by the J89 authors

Click Here To Show Diagram Code
[go]$$W
$$ -----------------------
$$ | . P . W W B . W X . .
$$ | P B W . W B . W X X .
$$ | . B B W W B B B O X .
$$ | B . B B B O O O O X .
$$ | P P B O O . . . . . .
$$ | P B B O . . . . . . .
$$ | B B O O . . . . . . .
$$ | O O O . . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------[/go]

Click Here To Show Diagram Code
[go]$$W
$$ -----------------------
$$ | . @ . @ @ # . @ X . .
$$ | @ # @ . @ # . @ X X .
$$ | . # # @ @ # # # O X .
$$ | # . # # # O O O O X .
$$ | @ @ # O O . . . . . .
$$ | @ # # O . . . . . . .
$$ | # # O O . . . . . . .
$$ | O O O . . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------[/go]

--------------------------

There is no need for White to fill the ko-shape in the corner.

The reference to Article 1 of the legal text in the commentary to this example does not make any sense at all.

Probably it is a mismatch with L&D Example 24.

++++++++++++++++++++++++++

Click Here To Show Diagram Code
[go]$$W
$$ -----------------------
$$ | . O . O O X . O X . .
$$ | O X O . O X . O X . .
$$ | . X X O O X X Q X X .
$$ | X . X X X O Y Y O X .
$$ | O O X O O O O O O X .
$$ | O X X O . . . . X X .
$$ | X X O O . . . . . . .
$$ | O O O . . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------[/go]

With a slight modification in the upper right, Black would have a disadvantageous two-step ko for killing the entire corner during actual play.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: Japonese counting
Post #368 Posted: Tue Sep 28, 2021 6:08 am 
Lives in gote

Posts: 445
Liked others: 0
Was liked: 37
kvasir wrote:
:w1: Throwing away stones to prevent captures doesn't count (but I don't know why).

This is what the text states (throwing away infinite stones is not possible/productive in normal play thus not accepted in hypothetical play either).

Top
 Profile  
 
Online
 Post subject: Re: Japonese counting
Post #369 Posted: Tue Sep 28, 2021 7:54 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
jann wrote:
kvasir wrote:
:w1: Throwing away stones to prevent captures doesn't count (but I don't know why).

This is what the text states (throwing away infinite stones is not possible/productive in normal play thus not accepted in hypothetical play either).

There are NO status confirmation prisoners who would make their way over to the players' lids.
Status confirmation does not have any effect on the position when the game stopped, thus "territory" remains unaffected.

This is why I stated that the reference to Article 1 does not make any sense.

----------------------------------

In L&D Example 23, Black needed a very large ko-threat elsewhere on the board to capture White's corner on his own initiative during "play".

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)


Last edited by Cassandra on Tue Sep 28, 2021 8:00 am, edited 1 time in total.
Top
 Profile  
 
Offline
 Post subject: Re: Japonese counting
Post #370 Posted: Tue Sep 28, 2021 7:56 am 
Lives in sente

Posts: 896
Liked others: 22
Was liked: 168
Rank: panda 5 dan
IGS: kvasir
jann wrote:
kvasir wrote:
:w1: Throwing away stones to prevent captures doesn't count (but I don't know why).

This is what the text states (throwing away infinite stones is not possible/productive in normal play thus not accepted in hypothetical play either).



I don't really read the reference to Article 1 like that. The meaning is too ambiguous. It may still be what was meant, it is somewhat of a non sequitur. This argument is familiar enough, kind of how one might normally explain snap-backs and seki, but in j89 it is only used for this situation in example 23.

Top
 Profile  
 
Online
 Post subject: Re: Japonese counting
Post #371 Posted: Tue Sep 28, 2021 10:01 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
A flowery, but meaningless, comment and no move sequences displayed.

This is a clear indication that the authors had no real idea of a convincing justification that they could be sure would not cause any undesirable side effects.

With L&D Example 23 we are back at "cycles" again, and obviously, cycles had not been adequately analyzed in Japan either.
It is also obvious that the START of a cycle should not be prevented. Otherwise the double-ko sequences displayed in the examples made no sense.


--------------------------

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

In L&D Example 23, ...

Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | 4 O . O O X . O X . .
$$ | O X O . O X . O X X .
$$ | 3 X X O O X X X O X .
$$ | X X X X X O O O O X .
$$ | 2 . X O O . . . . . .
$$ | . X X O . . . . . . .
$$ | X X O O . . . . . . .
$$ | O O O . . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]

... is it WHITE (with :w4:) who founded the infinite cycle, ...

Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | O O 5 O O X . O X . .
$$ | O X O . O X . O X X .
$$ | X X X O O X X X O X .
$$ | X X X X X O O O O X .
$$ | O . X O O . . . . . .
$$ | . X X O . . . . . . .
$$ | X X O O . . . . . . .
$$ | O O O . . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]

Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | . 6 X O O X . O X . .
$$ | . X O . O X . O X X .
$$ | X X X O O X X X O X .
$$ | X X X X X O O O O X .
$$ | O . X O O . . . . . .
$$ | . X X O . . . . . . .
$$ | X X O O . . . . . . .
$$ | O O O . . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]

Click Here To Show Diagram Code
[go]$$B
$$ -------------------------------------------------
$$ | 7 O . O O X . O X . . | . O . O O X . O X . . |
$$ | 8 X O . O X . O X X . | 7 X O . O X . O X X . |
$$ | X X X O O X X X O X . | X X X O O X X X O X . |
$$ | X X X X X O O O O X . | X X X X X O O O O X . |
$$ | O . X O O . . . . . . | O 8 X O O . . . . . . |
$$ | . X X O . . . . . . . | . X X O . . . . . . . |
$$ | X X O O . . . . . . . | X X O O . . . . . . . |
$$ | O O O . . . . . . . . | O O O . . . . . . . . |
$$ | . . . . . . . . . . . | . . . . . . . . . . . |
$$ -------------------------------------------------[/go]

... or is it BLACK (with :b7:) who keeps the (unforced!!!) cycle going forever, just because he wants to take White's uncapturable stones at the top off the board?


--------------------------

The solution to achieve the intended result of the L&D status confirmation (all White stones declared "dead") is simple, as we have known for some time:
Put a ban on the last real move of a cycle ( :w8: here), as it already exists on the recapture into a ko-shape.

I also think it would be possible (if considered appropriate) to limit this ban to asymmetrical cycles, where one side captures more stones than the other in one pass.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: Japonese counting
Post #372 Posted: Wed Sep 29, 2021 5:29 am 
Lives in sente

Posts: 896
Liked others: 22
Was liked: 168
Rank: panda 5 dan
IGS: kvasir
Cassandra wrote:
The solution to achieve the intended result of the L&D status confirmation (all White stones declared "dead") is simple, as we have known for some time:
Put a ban on the last real move of a cycle ( :w8: here), as it already exists on the recapture into a ko-shape.


:w8: recreates the position only in the sense that it is not the same side to move and the regular ko-ban in it is different.

I just noticed this. Doesn't this make it a bit harder to define which move to ban in that particular cycle? That is if you actually ban the cycle when it is not the same player to move it seems to invite trouble. If you ban it when it is the same player to move, it is then (I think) black that can't recreate the cycle.

Top
 Profile  
 
Online
 Post subject: Re: Japonese counting
Post #373 Posted: Wed Sep 29, 2021 6:38 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
kvasir wrote:
I just noticed this. Doesn't this make it a bit harder to define which move to ban in that particular cycle? That is if you actually ban the cycle when it is not the same player to move it seems to invite trouble. If you ban it when it is the same player to move, it is then (I think) black that can't recreate the cycle.

No, it's always the same simple solution: Prohibit the last genuine move of a cycle, and there will be absolutely no problems remaining with cycles.
As a matter of course, you will have to choose the moment just before an earlier board position is repeated.

In the cycle above, it's White who starts a cycle that -- in principle -- is disadvantageous for her. She also has the last genuine move in the cycle, so she will lose...

Please refer to GT territory rules:: Honte's loops and cycles for a detailed analysis of cycles.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: Japonese counting
Post #374 Posted: Wed Sep 29, 2021 3:42 pm 
Lives in sente

Posts: 896
Liked others: 22
Was liked: 168
Rank: panda 5 dan
IGS: kvasir
Cassandra wrote:
kvasir wrote:
I just noticed this. Doesn't this make it a bit harder to define which move to ban in that particular cycle? That is if you actually ban the cycle when it is not the same player to move it seems to invite trouble. If you ban it when it is the same player to move, it is then (I think) black that can't recreate the cycle.

No, it's always the same simple solution: Prohibit the last genuine move of a cycle, and there will be absolutely no problems remaining with cycles.
As a matter of course, you will have to choose the moment just before an earlier board position is repeated.

In the cycle above, it's White who starts a cycle that -- in principle -- is disadvantageous for her. She also has the last genuine move in the cycle, so she will lose...

Please refer to GT territory rules:: Honte's loops and cycles for a detailed analysis of cycles.



I have referred the thread but since it goes directly into examples and has more than 100 posts it is likely I have not understood it.

This is example 17 from j89.

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


Quote:
The three black stones in the corner are alive. The ten white stones surrounding them are dead. The eleven surrounded white stones to the right also die through collapse of the seki.


I have a question about the following cycle in the confirmation phase.

Click Here To Show Diagram Code
[go]$$W :w3: pass-for-ko :b4: pass-for-ko
$$ -----------------------
$$ | X X X . O X 1 X O . O X . .
$$ | . O O O O X X O O O O X .
$$ | O O X X O X O 2 O X X X
$$ | . O X O X X X O O X .
$$ | O X X O X . X O X X .
$$ | X X X O X X O X X . .
$$ | . . X O O O O O . . .
$$ | . . X X . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]


Click Here To Show Diagram Code
[go]$$W :w7: pass-for-ko :b8: pass-for-ko
$$ -----------------------
$$ | X X X . O X O 6 O . O X . .
$$ | . O O O O X X O O O O X .
$$ | O O X X O X 5 X O X X X
$$ | . O X O X X X O O X .
$$ | O X X O X . X O X X .
$$ | X X X O X X O X X . .
$$ | . . X O O O O O . . .
$$ | . . X X . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]


I think :b8: is the last move in the cycle, I don't know if it is "real" or "genuine", at any rate the last board move is :b6:. If either of these moves is forbidden then black is actually capturable.

There is something in the thread that the last move is forbidden if for the cycle length (cl) the following holds:
cl = 2 + 4n (n = 0, 1, 2, 3, ...)

You showed a cycle of length 7 and said the last move is forbidden so I assume (wrongly?) that it doesn't matter what the cycle length is?

Top
 Profile  
 
Online
 Post subject: Re: Japonese counting
Post #375 Posted: Wed Sep 29, 2021 6:16 pm 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
The thread referred to discusses cycles in a world WITHOUT pass-for-ko (and with the principle of "two-eye-formations" instead of "uncapturable"). Thus, the double-ko cycle with a length of 8 (including 4 "pass") would no longer be ENFORCED. Black would interrupt it with moves in the upper left corner, killing White's group there.

A remaining cycle for the position of L89's L&D Example 17 would be an OPTIONAL one -- in principle:

Click Here To Show Diagram Code
[go]$$W :w3: pass
$$ -----------------------
$$ | X X X . O X 1 X O . O X . .
$$ | . O O O O X X O O O O X .
$$ | O O X X O X O 2 O X X X
$$ | . O X O X X X O O X .
$$ | O X X O X . X O X X .
$$ | X X X O X X O X X . .
$$ | . . X O O O O O . . .
$$ | . . X X . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]

Click Here To Show Diagram Code
[go]$$W :b6: pass
$$ -----------------------
$$ | X X X . O X O 4 O . O X . .
$$ | . O O O O X X O O O O X .
$$ | O O X X O X 5 X O X X X
$$ | . O X O X X X O O X .
$$ | O X X O X . X O X X .
$$ | X X X O X X O X X . .
$$ | . . X O O O O O . . .
$$ | . . X X . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]

The last genuine move (i.e. NO "pass") of the cycle would be a WHITE one.
White (as well as Black) would be discouraged to START this nonsensical cycle that only lengthens the confirmation sequence.

In addition, this cycle is NOT ENFORCED, so Black would be free to use his :b6: for an atari in the upper left corner, achieving the same result as with banning the last genuine move in that cycle.


++++++++++++++++++++++++++

Under J89 ...

Click Here To Show Diagram Code
[go]$$W :w3: pass-for-ko
$$ -----------------------
$$ | X X X . O X 1 X O . O X . .
$$ | . O O O O X X O O O O X .
$$ | O O X X O X O 2 O X X X
$$ | . O X O X X X O O X .
$$ | O X X O X . X O X X .
$$ | X X X O X X O X X . .
$$ | . . X O O O O O . . .
$$ | . . X X . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]

Click Here To Show Diagram Code
[go]$$W
$$ -----------------------
$$ | X X X . O X O M O . O X . .
$$ | 4 O O O O X X O O O O X .
$$ | O O X X O X T X O X X X
$$ | . O X O X X X O O X .
$$ | O X X O X . X O X X .
$$ | X X X O X X O X X . .
$$ | . . X O O O O O . . .
$$ | . . X X . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]

... Black would also attack White's corner group, as White is hindered to recapture into any ko-shape, as long as NOT ALL ko-bans have been lifted. And we can be very sure that Black will NOT pass for "his" double-ko ko-shape.


++++++++++++++++++++++++++

Yes, there is a difference between cycles with a length of
4, 8, 12, ...
and with a length of
(2,) 6, 10, 14, ...

(Without any pass-for-ko ban, it's only the latter type that can cause problems.)

Cycles can become "problematic" during status confirmation ONLY if there is a double-ko involved.

J89's "original" (i.e. as of 1989, according to James Davies' translation) version of IMMEDIATELY enabling the recapture into any ko-shape, which ko-ban has been lifted, apparently was designed for single-ko shapes only.
The unwanted side-effect was making the 8-move double-ko cycle ENFORCED (with the SAME player starting each half of it), turning it into a means for creating chôsei for other groups under consideration. Thus, the intended result of several L&D Examples could not be reached within J89(1989).


++++++++++++++++++++++++++

The apparently unwanted 6-move-cycle of J89's L&D Example 23 -- which is "sending three, returning two" -- includes one "pass" at its end; so it's the same player who plays the first, and the last, genuine move of that cycle.
Banning the last move of that cycle would resolve a problem that cannot be solved within the application of the legal text.

Under J89(<= 2007) it would be sufficient to limit this kind of ban to cycles that have a "pass" at their end.

EDIT: As J89 has "uncapturable", it would of course have the same effect to ban the repetition of the first move of the cycle.

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)


Last edited by Cassandra on Wed Sep 29, 2021 9:16 pm, edited 1 time in total.
Top
 Profile  
 
Offline
 Post subject: Re: Japonese counting
Post #376 Posted: Wed Sep 29, 2021 9:00 pm 
Lives in sente

Posts: 896
Liked others: 22
Was liked: 168
Rank: panda 5 dan
IGS: kvasir
Are you saying that in the sequence

ABCDABCD...

the move D would just be forbidden, regardless of if the position is recreated next time that player wishes to play D?

I can see how D is the correct move to forbid, when it is the the correct move to forbid, but how do you actually tell D from say E? Are we talking about showing the sequence ABCDABCD... and then backtracking to ABC and banning D? Could D then ever be played after that?

For example is ABCEFD legal? I guess not.

This has an Ing rule vibe. This is effectively same as the disturbing ko rule, right? or wrong?

Top
 Profile  
 
Online
 Post subject: Re: Japonese counting
Post #377 Posted: Wed Sep 29, 2021 9:50 pm 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
kvasir wrote:
Are you saying that in the sequence

ABCDABCD...

the move D would just be forbidden, regardless of if the position is recreated next time that player wishes to play D?

I can see how D is the correct move to forbid, when it is the the correct move to forbid, but how do you actually tell D from say E? Are we talking about showing the sequence ABCDABCD... and then backtracking to ABC and banning D? Could D then ever be played after that?

It's a bit difficult to transpose the results from the "two-eye-formation" world into the "uncapturable" world, as both worlds are divided by a one-move difference, which has a big decisive effect.

In the "two-eye-formation" world, it is possible to ban the last move of EVERY cycle.
With the cycle, the player moving last will be unable to create a "two-eye-formation", due to its infinite repetition.
Without move "D" in your example, it is likely that something of this player will "die" in the confirmation sequence. Which leads to the same result of being unable to create a "two-eye-formation".

You will easily see the difference to the "uncapturable" world.
-- In the "two-eye-formation" world, "chôsei" has the same effect as "killing" something (elsewhere).
-- In the "uncapturable" world, "chôsei" has the same effect as "living" with something (elsewhere).

Quote:
For example is ABCEFD legal? I guess not.

In the examples analysed at that time, banning referred to the repetition of the UNinterrupted course of a cycle (i.e. "D" must not be played after the next appearance of "A", "B", "C" in order).
Thus, it would be allowed to play "D" later (here after "E", "F").
(But it did not appear something like a one-time "ko-threat" "E" that had to be answered (with "F") in-between, at that time.)

However, if just another cycle was created, "D" would be banned again. Just a bit later.

Quote:
This has an Ing rule vibe. This is effectively same as the disturbing ko rule, right? or wrong?

If I understand correctly what Sensei's Library says about it (I had nothing to do with "fighting" / "disturbing" ko before), I would say "Yes".
If the last move of a cycle is a "pass", this cycle belongs to a position, where both sides capture different stones. In J89, it has the effect of preventing the L&D status confirmation from ending. J89's authors should have simply (and explicitly) forbidden the endless repetition of cycles, where both sides capture a DIFFERENT number of stones per pass (despite the fact that "captives" do not count during status confirmation).

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: Japonese counting
Post #378 Posted: Fri Oct 01, 2021 11:24 am 
Lives in sente

Posts: 896
Liked others: 22
Was liked: 168
Rank: panda 5 dan
IGS: kvasir
I struggle to understand everything you have said. This may just be my confusion but it is still unclear to me how you pick the move to ban.

I showed this sequence.

Click Here To Show Diagram Code
[go]$$W :w3: pass-for-ko :b4: pass-for-ko
$$ -----------------------
$$ | X X X . O X 1 X O . O X . .
$$ | . O O O O X X O O O O X .
$$ | O O X X O X O 2 O X X X
$$ | . O X O X X X O O X .
$$ | O X X O X . X O X X .
$$ | X X X O X X O X X . .
$$ | . . X O O O O O . . .
$$ | . . X X . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]

Click Here To Show Diagram Code
[go]$$W :w7: pass-for-ko :b8: pass-for-ko
$$ -----------------------
$$ | X X X . O X O 6 O . O X . .
$$ | . O O O O X X O O O O X .
$$ | O O X X O X 5 . O X X X
$$ | . O X O X X X O O X .
$$ | O X X O X . X O X X .
$$ | X X X O X X O X X . .
$$ | . . X O O O O O . . .
$$ | . . X X . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]


You seemed to say :w5: is the last move of the cycle. I am pretty sure that :w5: repeats a board position WITHOUT forming a cycle of moves because a move like :b6: has never been played.

The cycle of moves is :w1: to :b8:
The first repeated position is after :w5: (I am not mentioning passes because they are either ignored or we treat pass-ko bans as part of the position)

Basically, it is possible to repeat a board positions more frequently than you loop through the cycle.


For example what seems to work with my understanding of cycles and both example 17 and 23 is to ban the first board move in the second repetition of the loop (or second cycle of the loop, if I adopt your definitions), but I have no particular reason to think this works in general. Possibly one could count the number of board moves in the cycle instead and declare a ban on the player that has more board moves, but same, I have no particular reason that this would work in general.

The disturbing ko rule as stated by Ing really leaves it to intuition to decide who is the disturber and therefore who must break the loop. But I fear it may not be so simple to decide which side to punish.

Top
 Profile  
 
Online
 Post subject: Re: Japonese counting
Post #379 Posted: Sat Oct 02, 2021 1:24 am 
Lives in sente
User avatar

Posts: 1308
Liked others: 14
Was liked: 153
Rank: German 1 Kyu
kvasir wrote:
I struggle to understand everything you have said. This may just be my confusion but it is still unclear to me how you pick the move to ban.

In the world of "two-eye-formations", there is NO "pass-for-ko".

Click Here To Show Diagram Code
[go]$$W :w3: pass
$$ -----------------------
$$ | X X X . O X 1 X O . O X . .
$$ | . O O O O X X O O O O X .
$$ | O O X X O X O 2 O X X X
$$ | . O X O X X X O O X .
$$ | O X X O X . X O X X .
$$ | X X X O X X O X X . .
$$ | . . X O O O O O . . .
$$ | . . X X . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]

Click Here To Show Diagram Code
[go]$$W :b6: pass
$$ -----------------------
$$ | X X X . O X O 4 O . O X . .
$$ | . O O O O X X O O O O X .
$$ | O O X X O X 5 X O X X X
$$ | . O X O X X X O O X .
$$ | O X X O X . X O X X .
$$ | X X X O X X O X X . .
$$ | . . X O O O O O . . .
$$ | . . X X . . . . . . .
$$ | . . . . . . . . . . .
$$ -----------------------[/go]

:w5: is the last genuine move of the cycle, thus will be banned. Thereafter, Black will end the discussion by capturing White's group to the right.

-- The initiator of this double-ko cycle (which is NOT an enforced one; Black could continue in the upper left corner, instead) will lose his double-ko group for sure.
-- Thus, starting this double-ko cycle is discouraged.
-- Thus, there is NO double-ko in L&D status assessment.
-- This is the intended result.

------------------------------

The original J89 "pass-for-ko" ruling (for the "uncapturable" world) was NOT appropriate to kill any double-ko cycle (an 8-move one there).

_________________
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)

Top
 Profile  
 
Offline
 Post subject: Re: Japonese counting
Post #380 Posted: Sat Oct 02, 2021 8:12 am 
Lives in sente

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

Has white to add a move?

What is the point?

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


Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | . X X O O X T X . O .
$$ | . X O O . O X X O O .
$$ | O O O O O X X X O . .
$$ | X X X X X X O O O . .
$$ | O O O O O O O . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .[/go]
Black has taken the ko => a pass-for-ko is request for white to retake on the marked intersection

Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | . X X O O X . X . O .
$$ | 2 X O O 3 O X X O O .
$$ | O O O O O X X X O . .
$$ | X X X X X X O O O . .
$$ | O O O O O O O . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .[/go]

Black has taken a second ko => a new pass-for-ko is request

Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | . X X O O X T X . O .
$$ | O X O O X M X X O O .
$$ | O O O O O X X X O . .
$$ | X X X X X X O O O . .
$$ | O O O O O O O . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .[/go]

Here is the point: taking the definition of a ko (article 6) the triangle square is no more a ko => the corresponding pass-for-ko request disappears.
The situation is now
Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | . X X O O X . X . O .
$$ | O X O O X M X X O O .
$$ | O O O O O X X X O . .
$$ | X X X X X X O O O . .
$$ | O O O O O O O . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .[/go]
with only ONE pass-for-ko request active

The sequence continues by
Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | 4 5 X O O X . X . O .
$$ | O X O O X M X X O O .
$$ | O O O O O X X X O . .
$$ | X X X X X X O O O . .
$$ | O O O O O O O . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .[/go]


Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | O X . O O X . X . O .
$$ | O . O O X M X X O O .
$$ | O O O O O X X X O . .
$$ | X X X X X X O O O . .
$$ | O O O O O O O . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .[/go]

:w6: pass-for-ko

Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | O X . O O X . X . O .
$$ | O 7 O O X 8 X X O O .
$$ | O O O O O X X X O . .
$$ | X X X X X X O O O . .
$$ | O O O O O O O . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .[/go]


Click Here To Show Diagram Code
[go]$$B
$$ -----------------------
$$ | O X . O O X . X . O .
$$ | O X O O M O X X O O .
$$ | O O O O O X X X O . .
$$ | X X X X X X O O O . .
$$ | O O O O O O O . . . .
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .[/go]

:b9: pass-for-ko

the first ko appears again but white can take it (it is considered a new ko) => black cannot kill white in the confirmation phase => white has not to add a move in the proposed position.

IOW when do you consider a ko disappears (and the corresponding pass-for-ko request disappears at the same time) ?

Top
 Profile  
 
Display posts from previous:  Sort by  
Post new topic Reply to topic  [ 391 posts ]  Go to page Previous  1 ... 16, 17, 18, 19, 20  Next

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