Commentary on the Korean 2013 Rules

Bill Spight · **#61**

Mef wrote:

Bill Spight wrote:

Thank you, Robert, for the links to the rules texts. It really helps.

Edit: Hmmm. It looks like these rules are a variant of the Ing rules.

Brief comment:

"The number of liberties of a contiguous group is the sum of liberties of each stone in the group" (p. 5).

That statement is false, and they even give an example without noticing that.

These stones together have 7 liberties. The D-16 stone has 2 liberties, the D-15 stone has 3 liberties, and the E-16 stone has 3 liberties. 2 + 3 + 3 = 8. Tilt! :shock:

I'm late to the discussion, but this doesn't necessarily need to be an error. There is no problem with that group having 8 liberties (instead of what you would traditionally think of as 7). The only distinction in the rules where liberties are pertinent are moves that fill the last liberty, and all of those still apply. Admittedly, the examples they give that describe "two compartmentalized liberties" would be in error, but upon my first reading, I don't see an issue that would crop up from this definition.

Well, then, this Black group has two liberties, eh? :mrgreen:

asura · **#62**

For the rules alone I don't see a problem if a group in atari has two liberties (or even up to four). It just would be strange to define the term "atari" this way.

Btw. I like it that "3 points without capture" is now a seki when not played out during the game. Else you should play with area-scoring

Cassandra · **#63**

Bill Spight wrote:

This is similar to Three Points without Capturing, based upon a ruling by Honinbo Shuwa (IIRC), which is worth only two points under J89. I think that Shuwa was right, and would rule this a seki (or zero points, anyway).

I agree. This J1989-example should be a Seki, so being more consistent within the rules' environment.

Quote:

Besides, are you sure that that group is alive under hypothetical play? White can throw in to make a ko, which Black takes, and then if White passes for that ko, Black simply wins the ko. If White throws in to make another ko, Black takes it, as well. Now whichever ko White passes for, Black wins.

Sorry, I am afraid, I switched between different rule sets here.

For one reason or the other (I am currently busy with the J1989-examples), I adopted Robert's supplement for enabling recapturing Ko, which assures that all J1989-examples get the dedicated results of the official rules.

asura · **#64**

Cassandra, if you don't want that a ko-pass lift all ko-bans, try out my model

It works for all examples in the commentary and also for positions like that from Matti

Cassandra · **#65**

asura wrote:

Cassandra, if you don't want that a ko-pass lift all ko-bans, try out my model

It works for all examples in the commentary and also for positions like that from Matti

asura, your rules' text is a bit too long for me ;-)

I am just working on the adaption (primarily for "status confirmation") of a suggestion for Japanese-style rules that I found in a Japanese book. I.e. considering / comparing the results of the J1989-examples.

In my opinion, this model has a very elegant solution for not explicitely defining "locality", but at the same time delimiting the spots on the board that are available for "permanent stones after capture".

However, this model does not use any special Ko-ban for "status confirmation" (but the "usual" one that is known from the "play" phase), so -- with its original application -- the J1989-examples that contain triple-Ko all end with different status. Adding Robert's suggestion helps closing this gap, I assume. So there is nothing wrong with this kind of Ko-ban.

asura · **#66**

Cassandra wrote:

asura, your rules' text is a bit too long for me ;-)

Hmm, I think it's not the model itself - maybe it's more that I've written down it 'wrong'. The core is very small, but ensuring everything is well defined made the text much longer...

Quote:

I am just working on the adaption (primarily for "status confirmation") of a suggestion for Japanese-style rules that I found in a Japanese book. I.e. considering / comparing the results of the J1989-examples.

In my opinion, this model has a very elegant solution for not explicitely defining "locality", but at the same time delimiting the spots on the board that are available for "permanent stones after capture".

What is the advantage of restricting where you are allowed to play during the analysis instead of defining if a new stone at the end of the analysis was enabled by the capture or if the new stone was just played anywhere (e.g. inside safe territory) ?
Somehow both methods seems to make use of some kind of local relationship to the chain in question?

Quote:

However, this model does not use any special Ko-ban for "status confirmation" (but the "usual" one that is known from the "play" phase

I think for Japanese-style Rules in general (not J1989) it seems necessary to use "locality" OR "ko-pass", but not both.
Actually using both in a ruleset is a bit redundant because both mainly want enforce the same: locality.
But if neither of them is used you cannot keep the bent-4 ruling when there are *anywhere* unremoveable ko threats and the rules become more like area-scoring with territory-counting. (What doesn't mean it is bad. Somehow (maybe just atm

) I just like the other a bit more - but cannot give any good reason)

Matti · **#67**

I had a pause of some years in the discussions of improving territory style rules. However if pass to recapture a ko is not sufficient to solve the problems, I suggest the following:
In the phase where groups' statuses are resolved a player may recapture a ko if he has made a pass to recapture that ko. To recapture that ko again he must make two extra passes before making the capture and to recapture a thiird time he would need to make three extra passes etc.

Bill Spight · **#68**

Here is a simple idea. If a ko is taken in hypothetical play or an encore, it may not be taken back without a pass for that ko, and a pass for that ko prevents the opponent from ever taking it again.

For instance, in a double ko seki, let Black take the first ko. White cannot pass now, so takes the second ko. Now Black passes for the second ko, and White passes for the first ko. Black takes the second ko and White takes the first ko. Now play stops in the double ko seki, as neither player can take one of its kos in this position. This also works where the double ko seki is temporary.

The fact that only a pass can act as a ko threat to allow the defender to take a ko back favors the player who takes the ko. In effect, it makes that player the komaster of that ko. But in return, the prohibition against taking that ko again means that the komaster must make use of that advantage or lose it.

Mef · **#69**

Bill Spight wrote:

Well, then, this Black group has two liberties, eh? :mrgreen:

Indeed it does! (=

Luckily that fact has no effect on how one would strategically want to play. There is no reason a group with two liberties can't be captured with only one move.

asura · **#70**

Bill Spight wrote:

For instance, in a double ko seki [...]
Now play stops in the double ko seki, as neither player can take one of its kos in this position. This also works where the double ko seki is temporary.

How can it work in this kind of position? When the play in the double ko seki stops then it becomes uncapturable.

Cassandra · **#71**

asura wrote:

Bill Spight wrote:

For instance, in a double ko seki [...]
Now play stops in the double ko seki, as neither player can take one of its kos in this position. This also works where the double ko seki is temporary.

How can it work in this kind of position? When the play in the double ko seki stops then it becomes uncapturable.

Yes, indeed.

In my opinion, as far as the "critical" Nihon Kiin Examples are concerned that contain a Triple-Ko (in what combination of Ko ever), any special Ko-rule for "status confirmation", which is designed to lift one SINGLE special Ko-ban at a time only, in principle has the same effect as if there was no special Ko-ban for "status confirmation" at all.

The so called "Seki" in these examples will never "collapse" through actual moves, this is the same as within "play".

Bill Spight · **#72**

asura wrote:

Bill Spight wrote:

For instance, in a double ko seki [...]
Now play stops in the double ko seki, as neither player can take one of its kos in this position. This also works where the double ko seki is temporary.

How can it work in this kind of position? When the play in the double ko seki stops then it becomes uncapturable.

That's a good point.

Consider this position from the J89 commentary.

Click Here To Show Diagram Code: [go]$$Bc Moonshine life? $$ +---------------------------------------+ $$ | . O O . O . O X . X O . O X . . . . . | $$ | O X X O O O O X X O O O O X . . . . . | $$ | X X . X X X O X O . O X X X . . . . . | $$ | . . . . X O X X X O O X . . . . . . . | $$ | . . . . X O X . X O X X . . . . . . . | $$ | . . . . X O X X O X X . . . . . . . . | $$ | . . . . X O O O O . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . |[/go]

The double ko seki eventually collapses under J89 rules. But not by my suggestion.

Click Here To Show Diagram Code: [go]$$Bc Moonshine life? $$ +---------------------------------------+ $$ | 1 O O . O . O X 2 X O . O X . . . . . | $$ | O X X O O O O X X O O O O X . . . . . | $$ | X X . X X X O X O 3 O X X X . . . . . | $$ | . . . . X O X X X O O X . . . . . . . | $$ | . . . . X O X . X O X X . . . . . . . | $$ | . . . . X O X X O X X . . . . . . . . | $$ | . . . . X O O O O . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . |[/go]

= pass for ko at :b3:

,

= pass for ko at :w2:

Click Here To Show Diagram Code: [go]$$Bc Moonshine life? $$ +---------------------------------------+ $$ | B O O . O . O X O 7 O . O X . . . . . | $$ | . X X O O O O X X O O O O X . . . . . | $$ | X X . X X X O X 6 X O X X X . . . . . | $$ | . . . . X O X X X O O X . . . . . . . | $$ | . . . . X O X . X O X X . . . . . . . | $$ | . . . . X O X X O X X . . . . . . . . | $$ | . . . . X O O O O . . . . . . . . . . | $$ | . . . . . . . . . . . . . . . . . . . |[/go]

Now Black can no longer capture the :w6:

ko, even after the seki collapses.

The once only rule still works, however. A player can take a particular ko only once in the same position of the whole board. That prevents interminable back and forth in the double ko, but allows the ko capture when the seki has collapsed.

asura · **#73**

Bill, I had exactly this position in mind

I know (for sure) that I've read on SL about this idea many years ago and (not sure) beleave it was exactly that position, but at least that principle.
Putting all informations together - considering that you've put the conter in the same sentence - I think you have lost exactly one bit of information, transforming your knowlege: (x does not work because y) into (x works even if y)

When considering the whole board it works. I'm not sure if the main priciple is very close to the ko-pass concept or if it is more related to cycles in general.

Matti's idea (to pass n times for the n. recapture of a ko) seems to be related to both concepts, too, because it is somehow necessary to define which ko recapture is the same (e.g. in multi stage kos). The indirect solution to provide sooner or later enough time to ignore the ko-passes as a 'side-effect' is interesting.

Matti · **#74**

asura wrote:

Matti's idea (to pass n times for the n. recapture of a ko) seems to be related to both concepts, too, because it is somehow necessary to define which ko recapture is the same (e.g. in multi stage kos). The indirect solution to provide sooner or later enough time to ignore the ko-passes as a 'side-effect' is interesting.

To make book keeping easier I might replace "n passes for ko" with "n-tuple pass for ko", which means that a player declares to make n concecutive passes.

I think assinging passes to a ko, the ko should be defined by two intersections: intersection of the stone which is played and the intersection of the stone which gets captured. We might manage the definiton with one intersection, but then we might get a new question to sort out, if we have a multi stage ko which can be played from two directions.

Commentary on the Korean 2013 Rules

Who is online