I mentioned a similar idea some time ago (https://www.lifein19x19.com/viewtopic.p ... 30#p259130), but I don't think captures can be ignored like this.
An anomaly or rule flaw is an unintended difference in behavior compared to some baseline. I think the first comparison should always be to continued play with normal rules. Confirmation/analysis only needs to address the removal/cleanup costs, anything else like extra ko rules are optional hacks for different problems (and source of anomalies, not surprisingly).
So in this case what would happen if play would continue? W could proceed to either accumulate 500 pts then give up the board, or make pass alive in the whole area in question. This is the baseline/reality. (see also J89 LD ex 23 commentary)
Japonese counting
-
Gérard TAILLE
- Gosei
- Posts: 1346
- Joined: Sun Aug 23, 2020 2:47 am
- Rank: 1d
- GD Posts: 0
- Has thanked: 21 times
- Been thanked: 57 times
Re: Japonese counting
Let me try to understand through the following example you mentionned in your post viewtopic.php?p=266642#p266642Cassandra wrote:It does NOT matter whether White can kill Black's stones or not during status confirmation.Gérard TAILLE wrote:Oops something goes wrong.
I you begin the confirmation phase by this black move that means that you try to kill white and we agree it is not possible.
If now white wants to prove she can kill black stones she will simply begin by: Where is your point with this position?
The deciding question is whether you can turn White's group into a "two-eyed alive" one, even if Black moves first.
Only if you can answer this question with "yes", you have achieved "INDEPENDENT life" for White's group (probably you like this wording more).
There are TWO conditions for Black stones (which remained on the board) that shall become White's prisoners after the end of the game, i.e. without really capturing these by actual play:
These stones did NOT achieve the property "two-eyed alive" during status confirmation.
These stones must be fully enclosed by (nothing else than) a White group that has achieved the property "two-eyed alive" during status confirmation.
In your example, condition
Therefore, the entire position is a seki.
Please note that you will not find any "dead" in my writing above. Just because this term would be only a descriptive one (and no specifying one) for Black stones that fulfill BOTH conditions
There is NO death DURING status confirmation, which would make its way back onto the board position at the end of "play" (i.e. at the start of the status confirmation).
Whether stones are "dead" or not is decided after the status confirmation of all (relevant) groups has been finished, and derived from the combination of all these results, as I tried to explain above.
Please to not mix "two-eyed alive" with J89's or J2003's "uncapturable" / "capturable-1".
Two questions for my understanding Cassandra:
1) Are
"two-eyed alive"2) Are
"two-eyed alive"-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: Japonese counting
In order to understand a concept you should start with asking easy questionsGérard TAILLE wrote:Let me try to understand through the following example you mentionned in your post viewtopic.php?p=266642#p266642
But nevertheless, I will explain J89 life-and-death example 4 in detail. Hopefully this will widen the understanding of the syntax.
Let's start with the status confirmation of Black's three stones in the corner.
As a matter of course, Black's group can get two eyes, if he plays first.
--------------------------
White plays first.
White cannot prevent this result, even if she plays first.
--------------------------
Variation:
Variation, same result as above.
################################
################################
Second step is the status confirmation for Black's three stones at the top.
The status of Black's three stones at the left will be the same, as the position is symmetrical.
In the sequence above that Black started,
captured White's stones at the top.Black's three stones got captured, but they could be re-established in total. They have two eyes.
--------------------------
White plays first.
In the main line above,
captured White's stones at the top.Same result as above, even if White moves first.
##########################
##########################
Your unasked question about the status of White's groups would have been answered in the posting you linked at if we ever wanted to get a result of this status confirmation for Black playing first:
White's stones got captured, and neither could they become re-established in total (for achieving the status "two-eyed alive"), nor partially, which would give them the chance to become turned into components of a seki in the concluding status assessment (in another environmant than here of course).
However, there remains that psychologically unfavorable "life" of the other White group, which was not currently under examination at all. Therefore. ...
... we are fine with letting White show what she can achieve with having the advantage of the first move.
We already know from the sequences above that White will be unable to turn her stones into a "two-eyed alive" group. No single stone of a White group under consideration will survive on the board.
--------------------------
The concluding result of the status confirmation has all Black groups "two-eyed alive", while all White groups are "NOT two-eyed alive".
White's "NOT two-eyed alive" stones are all completely surrounded by Black "two-eyed alive" groups; therefore, they can taken off the board after the end of the status confirmation as prisoners (being "dead").
I am very well aware that this result does not match that of J89, but as I explained in my linked posting, the authors have exaggerated here with their desire to generate a seki.
As I stated there, the White groups are completely surrounded by Black "two-eyed alive" groups EACH, and there can be NEVER EVER ANY interaction between these two seperated sections of the board.
Probably it had been this life-and-death example that caused Robert to also overextend his "local-2" in his J2003.
As a matter of course, one of both White's groups cannot be captured by actual play before the game stops.
But can any the double-ko groups in life-and-death examples 16 to 18 become captured by actual play before the game stops? NO!!! Nevertheless, these are declared "dead".
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)
Re: Japonese counting
You should look at the logic behind this. As I wrote above, the main reason for any confirmation phase is to prevent cleanup moves messing up the score. Allowing something that can not be done in actual play (or giving points for two groups where only one can be captured) is much worse than seki (which costs at most a few points on extra moves).Cassandra wrote:I am very well aware that this result does not match that of J89, but as I explained in my linked posting, the authors have exaggerated here with their desire to generate a seki.
A simple moonshine life is better for your point (not capturable by actual play still dead in all rules) - which is a different story.As a matter of course, one of both White's groups cannot be captured by actual play before the game stops.
But can any the double-ko groups in life-and-death examples 16 to 18 become captured by actual play before the game stops? NO!!! Nevertheless, these are declared "dead".
-
Gérard TAILLE
- Gosei
- Posts: 1346
- Joined: Sun Aug 23, 2020 2:47 am
- Rank: 1d
- GD Posts: 0
- Has thanked: 21 times
- Been thanked: 57 times
Re: Japonese counting
OK Cassandra your definition is clear at least on this example.Cassandra wrote:In order to understand a concept you should start with asking easy questionsGérard TAILLE wrote:Let me try to understand through the following example you mentionned in your post viewtopic.php?p=266642#p266642
But nevertheless, I will explain J89 life-and-death example 4 in detail. Hopefully this will widen the understanding of the syntax.
Let's start with the status confirmation of Black's three stones in the corner.
Just a additionnal confirmation concerning your definition of "two-eyed alive" on a more difficult (but well-known) example:
Is it correct to say that white (with the move) can prevent black marked stone to become "two-eyed alive", according to your defintion?
If yes does that mean that black should add a move to avoid seki?
-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: Japonese counting
You should note that "uncapturable" is a completely different pair of shoes than "two-eyed alive".jann wrote:A simple moonshine life is better for your point (not capturable by actual play still dead in all rules) - which is a different story.
The concept of "uncapturable" is a passive one.
It is sufficient to do NOTHING, in order to prove that a component of a seki is "uncapturable". YOUR OPPONENT must take the lead during the evidence.
The concept of "two-eyed alive" is an active one.
In many cases (expect those where a group is solidly and uncuttable connected), YOU will have to take action in order to create two taboo-points for your opponent inside your group, taking the lead during the evidence.
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)
-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: Japonese counting
Gérard TAILLE wrote:Just a additionnal confirmation concerning your definition of "two-eyed alive" on a more difficult (but well-known) example:
Is it correct to say that white (with the move) can prevent black marked stone to become "two-eyed alive", according to your defintion?
If yes does that mean that black should add a move to avoid seki?
I suppose that this position has been developed by Robert
Especially to prove the WORDING of "after it has been captured" rules mistaken.
But it's me here, so let's ask for the deeper meaning of the condition for turning captured stones into "two-eyed alive" ones first, which was hidden beyond the surface before (supposed) Robert found this example above.
Please remember that the primary concern of such a condition were snap-back and nakade, none of which is addressed in the position above.
This is one possible "final" position of the status confirmation for Black's three stones at the upper edge in life-and-death example 4 of J89.
Let's modify it a bit.
And let's give Black the task of creating a group with only two single eye points, in order to be sure that we apply the rule right (I know that Robert is very fond of this idea).
You will quickly notice that Black is not the very most experienced status confirmator ...
Task completed!
Ups! Black's captured stones are not reborn in total, so did Black create a seki for the final position of the game? Of course not!
Every single board point that was occupied by a captured Black stone must remain under the "control" of a "two-eyed alive" Black group.
I. e. NONE of these points must become part of a "two-eyed alive" White group.
If so, it is irrelevant whether such point is occupied by a Black stone or empty at the end of the status confirmation.
"Being an eye point" is equivalent to "being a reborn captured stone".
As you will know, this is the final position of the status confirmation for Black's marked stone at hoshi.
You will realise for sure that the board point that was occupied by Black's stone is empty. And that there is a BLACK "two-eyed alive" group around.
To keep a long story short ...
There is NO need for Black to add a move.
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)
-
Gérard TAILLE
- Gosei
- Posts: 1346
- Joined: Sun Aug 23, 2020 2:47 am
- Rank: 1d
- GD Posts: 0
- Has thanked: 21 times
- Been thanked: 57 times
Re: Japonese counting
I believe I understand Cassandra. The wording "control" seems fine : it is either a stone or one of the eyes of a "two-eyed alive" formation.Cassandra wrote: Every single board point that was occupied by a captured Black stone must remain under the "control" of a "two-eyed alive" Black group.
Now I can come back to the following position:
1) you can prove
are under control2) you can prove
are under control3) but you cannot
and can be under control at the same time.Eventually what will happen in the confirmation phase. Can black claim that both
and are under control or can she only choose one group or ...?IOW the "under control" concept is clear but it is not clear if you are allowed to add groups "under control" without precaution.
-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: Japonese counting
I am always happy when I can help. Especially at the moment, when the preparatory works for the forthcoming 2021 update of our Igo Hatsuyôron 120 website have reached the "final 5 percent" (hopefully for the last time this fourth or fifth time), which are always extremely boring. A little relief is good.Gérard TAILLE wrote: 1) you can prove
2) you can prove
3) but you cannot
##########################
##########################
The procedure is as follows.
##########################
##########################
Step
: Determine the status for each individual group There are no dependencies on other groups during the individual status determination!
##########################
##########################
Step
: Combine the resultsMutual dependencies are only now to be considered, especially for determining seki (if non "two-eyed alive" groups were determined that could not be captured, or could be re-established only partially after having been captured -- but which is not the case here).
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)
-
Gérard TAILLE
- Gosei
- Posts: 1346
- Joined: Sun Aug 23, 2020 2:47 am
- Rank: 1d
- GD Posts: 0
- Has thanked: 21 times
- Been thanked: 57 times
Re: Japonese counting
Thank you Cassandra for having explained what you mean by "two-eyed alive".
It is clear to me and eventually quite easy (I mean easier that the local-2 concept of J2003 ).
Now how it works with territory? In the position above the normal play for black is to play on the dame "d" and black wins the game by one point (6 against 5).
But both players pass and confirmation phase takes place.
In J2003 the result is jigo because, due to the dame, the two big groups are in seki => no territory and no prisonners.
What about your proposal? Do you take care of the dame (I mean an empty point adjacent to a white and to a black two-eyed alive group) or do you simply ignore it?
In the case this dame prevent you to recognise a territory, does that mean that white will win the game by one prisonner to zero?
It is clear to me and eventually quite easy (I mean easier that the local-2 concept of J2003
).Now how it works with territory? In the position above the normal play for black is to play on the dame "d" and black wins the game by one point (6 against 5).
But both players pass and confirmation phase takes place.
In J2003 the result is jigo because, due to the dame, the two big groups are in seki => no territory and no prisonners.
What about your proposal? Do you take care of the dame (I mean an empty point adjacent to a white and to a black two-eyed alive group) or do you simply ignore it?
In the case this dame prevent you to recognise a territory, does that mean that white will win the game by one prisonner to zero?
-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: Japonese counting
The "駄" (= "da") in "駄目" (= "dame"; "目" is "point") means (in my understanding; likely amoung others)Gérard TAILLE wrote:Now how it works with territory? In the position above the normal play for black is to play on the dame "d" and black wins the game by one point (6 against 5).
But both players pass and confirmation phase takes place.
In J2003 the result is jigo because, due to the dame, the two big groups are in seki => no territory and no prisonners.
What about your proposal? Do you take care of the dame (I mean an empty point adjacent to a white and to a black two-eyed alive group) or do you simply ignore it?
In the case this dame prevent you to recognise a territory, does that mean that white will win the game by one prisonner to zero?
-- of poor quality
-- not especially valuable
-- useless
It should be evident that the main application is with empty board points that are located between groups with two eyes (or between a group with two eyes and the fencing group of a seki, just to be complete) and where it does not matter from which side they are occupied (aka "neutral points").
Empty board points in seki are also referred to with this term. But in seki, these points should NOT be occupied by any side. Therefore, these points have another characteristic than the DAME mentioned above.
DAME in seki are equally "useless", in that they cannot / will not become territory.
But nevertheless they have a special value, in that they prevent strings from becoming captured (leaving e.g. "three points without capturing" aside here, in order to keep the text as simple as possible).
############
Now let's return to the ruleset discussion.
Under the concept of "uncapturable" (used e.g. by J89 and J2003), DAME (in seki) are absolute necessary to distinguish
-- uncapturable groups that have two eyes (and therefore surround territory) from
-- uncapturable groups that are a compound of a seki.
The need to occupy DAME outside groups with two eyes (to confirm territory inside) is a side effect of using the same term for two different kinds of board points.
If there was a specific term for "DAME in seki" in Japanese, it would be completely unnecessary to occupy "neutral points" in connection with "territory".
The remaining benefit of forcing both players to occupy ALL "neutral points" on the board would be to guarantee that NO protective move inside a group with two eyes is missed, which occupation will become mandatory due to a potential shortage of liberties (of parts of the group with two eyes that are not yet "securely" connected within that group's compound of strings).
--------------
Under the concept of "two-eyed alive", empty board points are NOT needed to distinguish
-- groups that have two eyes from
-- groups that are a compound of a seki (and as such do NOT have two eyes).
In your example, Black will win the game by one point on the board, independent of the point of "d" being unuccupied or not.
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)
-
Mike Novack
- Lives in sente
- Posts: 1045
- Joined: Mon Aug 09, 2010 9:36 am
- GD Posts: 0
- Been thanked: 182 times
Re: Japonese counting
WHY do you say that? WHY do you think both those groups are in seki? They are both alive (in the two eye sense -- achieving two eyes cannot be prevented). The dame at "d" is irrelevant, not a point for either and filling ti does not require a response (say to maintain connectivity)Gérard TAILLE wrote:Thank you Cassandra for having explained what you mean by "two-eyed alive".
It is clear to me and eventually quite easy (I mean easier that the local-2 concept of J2003
Now how it works with territory?
In J2003 the result is jigo because, due to the dame, the two big groups are in seki => no territory and no prisonners.
zero?
-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: Japonese counting
Mike Novack wrote:WHY do you say that? WHY do you think both those groups are in seki? They are both alive (in the two eye sense -- achieving two eyes cannot be prevented). The dame at "d" is irrelevant, not a point for either and filling ti does not require a response (say to maintain connectivity)Gérard TAILLE wrote:Thank you Cassandra for having explained what you mean by "two-eyed alive".
It is clear to me and eventually quite easy (I mean easier that the local-2 concept of J2003
Now how it works with territory?
In J2003 the result is jigo because, due to the dame, the two big groups are in seki => no territory and no prisonners.
zero?
J89 wrote:Article 8. Territory
Empty points surrounded by the live stones of just one player are called "eye points." Other empty points are called "dame." Stones which are alive but possess dame are said to be in "seki." Eye points surrounded by stones that are alive but not in seki are called "territory," each eye point counting as one point of territory.
J2003 wrote:Definitions 3
…
In the final-position, a black-region / white-region is in-seki if at least one of its intersections is adjacent to a dame.
…
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)
-
Gérard TAILLE
- Gosei
- Posts: 1346
- Joined: Sun Aug 23, 2020 2:47 am
- Rank: 1d
- GD Posts: 0
- Has thanked: 21 times
- Been thanked: 57 times
Re: Japonese counting
That is clear Cassandra and quite simple which is a great advantage. The counterpart is that the results are not always the results expected by the japonese professionnals but I see another subtle drawback.Cassandra wrote:The "駄" (= "da") in "駄目" (= "dame"; "目" is "point") means (in my understanding; likely amoung others)Gérard TAILLE wrote:Now how it works with territory? In the position above the normal play for black is to play on the dame "d" and black wins the game by one point (6 against 5).
But both players pass and confirmation phase takes place.
In J2003 the result is jigo because, due to the dame, the two big groups are in seki => no territory and no prisonners.
What about your proposal? Do you take care of the dame (I mean an empty point adjacent to a white and to a black two-eyed alive group) or do you simply ignore it?
In the case this dame prevent you to recognise a territory, does that mean that white will win the game by one prisonner to zero?
-- of poor quality
-- not especially valuable
-- useless
It should be evident that the main application is with empty board points that are located between groups with two eyes (or between a group with two eyes and the fencing group of a seki, just to be complete) and where it does not matter from which side they are occupied (aka "neutral points").
Empty board points in seki are also referred to with this term. But in seki, these points should NOT be occupied by any side. Therefore, these points have another characteristic than the DAME mentioned above.
DAME in seki are equally "useless", in that they cannot / will not become territory.
But nevertheless they have a special value, in that they prevent strings from becoming captured (leaving e.g. "three points without capturing" aside here, in order to keep the text as simple as possible).
############
Now let's return to the ruleset discussion.
Under the concept of "uncapturable" (used e.g. by J89 and J2003), DAME (in seki) are absolute necessary to distinguish
-- uncapturable groups that have two eyes (and therefore surround territory) from
-- uncapturable groups that are a compound of a seki.
The need to occupy DAME outside groups with two eyes (to confirm territory inside) is a side effect of using the same term for two different kinds of board points.
If there was a specific term for "DAME in seki" in Japanese, it would be completely unnecessary to occupy "neutral points" in connection with "territory".
The remaining benefit of forcing both players to occupy ALL "neutral points" on the board would be to guarantee that NO protective move inside a group with two eyes is missed, which occupation will become mandatory due to a potential shortage of liberties (of parts of the group with two eyes that are not yet "securely" connected within that group's compound of strings).
--------------
Under the concept of "two-eyed alive", empty board points are NOT needed to distinguish
-- groups that have two eyes from
-- groups that are a compound of a seki (and as such do NOT have two eyes).
In your example, Black will win the game by one point on the board, independent of the point of "d" being unuccupied or not.
In common japonese rules the territories have a remarkable property that is never mentionned : territories are really independant to each other. As soon as a territory has been identified for say white then, even if you allow white to rearrange her territory as she want (provided it remains white territory) then the analyse of all other groups on the board does not change. OC, this gives a great consistency for all territories.
An example to understand what I mean: The upper right corner is white territory according to confirmation phase
Even if white rearranges her territory like this then the analyse all all group in the marked area will be the same.
This is quite interesting because as soon as you have identified a territory you can safely analyse the other groups of the board by excluding any play in this territory and by assuming the corresponding stones being definitly uncapturable (even if the owner of the territory pass always).
To assure this property common japonese rules conclude often to a seki where you conclude to "two-eyed alive" and territory. That means that you lost this consistency between territories which is a pity for me.
-
Cassandra
- Lives in sente
- Posts: 1326
- Joined: Wed Apr 28, 2010 11:33 am
- Rank: German 1 Kyu
- GD Posts: 0
- Has thanked: 14 times
- Been thanked: 153 times
Re: Japonese counting
I am sorry, Gérard, but I really do not have any idea of what you intended to say.Gérard TAILLE wrote:To assure this property common japonese rules conclude often to a seki where you conclude to "two-eyed alive" and territory. That means that you lost this consistency between territories which is a pity for me.
If you like to re-arrange the position of the two eye-points of a group (if technically possible), please do so. Nothing outside a "two-eyed alive" group (a Black one in your example) will be affected by anything inside. And vice versa.
The only single example that I know of, where J89 has a seki, but the application of the "two-eyed alive" idea has not, is life-and-death example 4.
However, I see this as absolutely irrelevant. I hope to have made clear, why.
Contrary to a potential hidden rule for achieving the intended results of examples 16 to 18 (which lack the derivation of the intended results), I am unable to imagine which hidden rule (please remember the interdependency with example 1) should lead us to the intended result of example 4, while having absolutely no side-effects on other positions.
#############################
By the way, as far as I can remember, we have not discussed "seki" in general yet.
#############################
By the way #2, it will be impossible to avoid any differences between the intended results of J89 / J2003 and those of applying the "two-eyed alive" idea, just because the concepts of "uncapturable" and "two-eyed alive" have one very decisive diffence in their foundations. Which we haven't discussed yet, too, as far as I can remember.
The really most difficult Go problem ever: https://igohatsuyoron120.de/index.htm
Igo Hatsuyōron #120 (really solved by KataGo)
Igo Hatsuyōron #120 (really solved by KataGo)