I personally think the rules are pretty straightforward regarding that matter.
No. Passing does not end the game, passing stops the game (see §9.1). “Stopping the game” and “ending the game” are not the same thing. Passing is not prohibited at any point.
As explained, passing is not an issue. But I’ll think about this question as: “OGS did not implement the Japanese Go rules correctly what should we do?”.
That is actually an interesting question and I’m not sure how other servers or e.g. chess sites are handling this. I could think of three different approaches:
Remove the option to select Japanese rules
Implement the features required to play according to the Japanese rules
Maybe add an aterisk to the ruleset with a hyperlink detailling exactly which rules OGS chose to implement and which rules were not implemented / differ from the original ruleset
Ah, I misunderstood you. Thanks for clearing that up.
So they would pass and the game stops.
And what should happen then when the game cannot be scored properly in the strictest sense of J89 rules?
Would they be able to mark dead stones, or is resumption the only option available to them, or something else?
Well, I think that all of the above discussion does go to show that there is significant disagreement / confusion about the exact procedure of the Japanese rules.
While there are many other nuances and considerations, I think that a fundamental difference between the Japanese rules and various area scoring rules is the following:
Under various area scoring rules, life and death status is determined by the skill of the players, by playing out any disputes.
Under the Japanese rules, life and death status is determined by the rules. The concept of “hypothetical play” refers to determining whether or not stones can be captured under perfect play, and this process of adjudication should not (in principle) be limited by the skill of the players (i.e., if a skilled arbiter stepped in to resolve a dispute, then I believe that they should provide the correct judgements, even when both players have gotten the status wrong).
Resumption is an option. However, in strict application of the rules, if it is found that the outcome of the game would hinge on who gets to play next, the game should be declared as “both players lose”.
If that’s not the case, and neither player wishes to resume, they could still attempt to score the game. I think that unclosed borders simply nullify territory (and prevent the removal or otherwise dead stones that resided in “non-territory”).
It gets a bit weird with unsettled positions. It can be the case that somethings are judged as mutually dead (this is informally called as “anti-seki”), but these dead stones are actually not removed for scoring purposes.
First of all there is no “strictest sense” of the Japanese rules. The rules do not give some articles more weight than other articles, so I do not see any argument for having the reader of the rules decide himself which rules to follow and which rules not to follow. Either you’re playing according to the rules or you are not.
But let’s just go through the procedure from start to finish together. Btw this conversation is pretty funny since you are a way more experienced and better Go player than me
Scoring according to Japanese rules
Step 1: Read the rules before playing
This is very important for any board game. The Japanese rules are quite concise so it shouldn’t take too long. The first thing that really interests us for scoring is §8:
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.
This is acutely important for our endeavor because we need to understand what “Dame” is first.
Step 2: Play
Do your moves, e.g. until you end up having the board position that started this thread.
Step 3: Pass
Both players pass in succession therefore stopping (not ending) the game (§9.1)
Step 4: Agreement on Dame
Now both players need to agree on life/death and territory. We are at a junction now. During this phase several different options are possible.
Scenario 4.1: Agreement
Since both players apparently stopped the game without filling in the Dame they can now agree to fill in the Dame during this phase in accordance with Co§9.2-2
Scenario 4.2: Resumption
If a player does not want to agree to filling in Dame according to Co§9.2-2, he still has the option to request the resumption of the game according to §9.3, which the opponent has to oblige to.
We are now back to “Step 2” and continue on from there.
Scenario 4.3: Disagreement
If the players realize at this point that filling in the Dame does not make any sense because there are effective moves left and can therefore not agree to fill in the “Dame” and end the game, then both players lose in accordance with §13.1
Step 5: Filling in the Dame
The players need to fill in the Dame (see Co§9.2-1)
This step can only happen after 4.1 happened
If we end up at 4.2 then we go back to 2 and then (hopefully) at some point arrive at 4.1
The scenario 4.3 ends the game
Step 6: Confirmation
After doing this they can now agree on life/death and territory (see §9.2) and thereby end the game.
Step 7: Handle Prisoners
Now since the game has ended, we first remove dead stones from the opponent and add them as prisoners (see §10.1) and then fill them into our opponents territory (see §10.2)
Step 8: Counting
Now we can count the territory. The player with more territory wins.
Not sure where you got this from, could you give me a reference? There are clarifying examples for the confirmation of life and death in the ruleset. But situations that are not among the examples are unclear and need to be confirmed via playing them out in accordance with §7 as well as Co§7.2
This is not accurate. The rules do not state that the game “should” be handled that way. There is a clear procedure outlined on how to decide this.
They can only score the game after filling in the Dame (see Co§9.2-1)
Well, an experienced driver may be a better driver than a new traffic cop, but they may not be more knowledgeable about traffic laws.
My knowledge about the rules is enough to play and finish my own games and introduce other people to the game, but it is not enough to lecture rules enthousiasts.
What I teach about rules is what seems good enough to play and finish the vast majority of games, as I do myself. I know it doesn’t cover all edge cases, but IME this is not really needed anyway.
When you’re past beginner stage, scoring procedure caveats hardly affect game play at all, so I don’t think it’s neccesary to be a rules expert to progress as a player.
It is not entirely clear to me what happens if the players fail to fill all dame in this phase. I’d think article 8 (about territory and dame) still allows scoring the game, but stones touching those dame would be considered to be in seki and thus not score any points for any regions they surround.
I suppose that in OP position neither player would then have any territory under J89 rules.
(Still, when I’m scoring beginner games I’d rather use something like Lasker-Maas or Spight rules, because it seems a more beginner-friendly territory scoring method to me).
In the OP case there were effective moves left after they stopped the game, but neither player seemed to realize. Would this rule still apply and warrant a “both players lose” adjudication?
My point is that I don’t believe that this is actually “playing out” the position subject to the skill of the players in finding the correct lines, but rather a hypothetical analysis of whether stones could or could not be captured under perfect play (of course, performing this analysis accurately may be beyond the skill of the players). In this sense, I believe the Japanese rules are determining life and death status by rule, since it is (in principle) defined by perfect play.
I’m not certain what you mean by saying that it is “not accurate” or which “clear procedure” you are exactly referring to. My earlier statement is based on this:
Article 13. Both players lose
After the game stops according to Article 9, if the players find an effective move, which would affect the result of the game, and therefore cannot agree to end the game, both players lose.
So what were go players doing then before there were written rules?
Go has a long history of being played in Japan without formal rules, so it seems self-evidently possible to do and still call it go.
I consider J89 to be merely the latest official attempt to formalise the long tradition it has had in Japan.
There is nothing obvious here and that’s because of
The written japanese rules were not elaborated for online play in view.
There were different adaptation of rules from each servers including OGS.
It would be interesting to analyse these and especially pandanet as being a japanese server as a starting point.
From the discussion it seems that after 2 passes the game should stop and enter a kind of demo stage in which both players should agree on a countable final position. If not so, then this results in a mutual loss.
Confirmation of the life and death of stones and territory requires that the players fill the dame and add any necessary stones inside their territory, in accordance with Article 8.
I do not think this leaves much room for interpretation.
Nice analogy.
If the players fail to fill in all the dame they’re violating the rules. If this is recognized before the end of the game, I understand that it should be a loss for both players (see §14). If this is only recognized after both players have somehow (erroneously) concluded the game and agreed upon a result. Then this result (whatever it may be) becomes valid in accordance with §10.4 as well as Co§10.4
I mean, yes §8 by itself would allow for scoring the game but when we read through a ruleset we need to recognize all the rules, and as I stated before Co§9.2-2 does not allow this.
But if we want to do a thought experiment, act like Co§9.2-2 would not exist and just look at §8. Then Black would have two points (N13, M12) and White would have zero points, if we claim that all of Black’s stones are alive.
No.
The rules in that regard are indeed not perfectly clear since:
Clause 2 of Article 7 states how to settle unclear questions of life or death involving ko after the game stops (see Article 9, clause 1).
does not specify to whom it needs to be “unclear”.
At no point anywhere in the rules are outside stakeholders or outside view points referenced though. The whole document refers to the “two players” playing a game. I think given the context of the ruleset there is enough certainty to assume this section means “unclear to the two players”. I cannot understand where your “believe” comes from, that’s why I was curious if you have any reference for that claim. Since this is indeed not perfectly clear we should maybe check if the rule makers have issued a statement regarding that section.
Maybe this is a linguistic misunderstanding but the section you quoted does not state what “should” happen. It states how the game is played.
Never said anything else. This thread has only been about the (current) Japanese rules of Go.
If your conclusion is that seki can’t possibly remain on the board, surely almost everyone is in fact interpreting it differently, right? For example you could interpret it as “fill the dame as much as they want to”.
While skilled arbiters can determine the correct result in many cases, and in many more cases than less skilled players, they are still far from perfect, and cannot determine perfect play in all cases.
Is it better to limit the result to the arbiter skill than to player’s skill, considering that the former is a more limited resource, one that is not always available?
People cannot interpret something they’ve never read. The share of players that actually read the rules is basically negligible, and the share of players actually playing according to the rules is even smaller. It’s not my fault that the rules are badly written and that no one actually follows them.
The Korean rules (and I love Korea, lived there and speak the language, so I hope they forgive me for saying this) are even worse. I do not think it is even possible to play the game at all following those rules. Had some good laughs when I read through it.
Like I said earlier:
I mean language is language. Requires means requires. This seems to be a desperate attempt to somehow make sense of a poorly written ruleset but I do not see how this is a sensible interpretation of the words in that sentence. There is even an example further doubling down on this:
In the position in Diagram 15, the black and white groups are both alive, but in seki because of the dame at A, so neither side has any territory. A move at A is needed to make Black’s and White’s eyes into territory.
Well, I have mentioned a couple of times in this thread that to me the phrase “Japanese rules” does not neccesarily coincide with the J89 revision. So I wouldn’t say it’s only about the J89 rules.
When I learned to play go, the J49 rules were still in force and I think those were worded quite differently than the J89 revision (although I can’t find the J49 text right now), while supposedly still capturing most of the tradition that came before.
But this change of rules had no effect whatsoever on my games, nor on my understanding of the game.
So my notion of “Japanese rules” is not restricted to just the J89 revision. It’s more a rule set family (call it “Japanese Style rules” if you will). I would also include Korean rules into it, their ancestors like the J49 rules, more recent proposals to improve on the J89 rules and some alternative Japanese Style rule sets.
Especially with beginners, I’ll keep reserving some freedom for myself to transfer the (in my view) overall sensible and elegant ideas behind this “Japanese rules” family to clarify what the game is about, rather than focusing on a more legalistic and poorly written description for professionals, which might change again in the future.
I agree that the rules are poorly written and that you have to fill dame to get the associated territory like in the example, but you’re claiming all dame must be filled period, so there are never any seki remaining at the end, right?
I understand your general view point and respect it. I still think that I’m justified in thinking that “Japanese rules of Go” refers to the current official Go rules of Japan unless it is clearly specified otherwise. We also came a long way from the beginning. I do think that filling out Dame is a good thing to do for beginners before they figure out the score. But I could be wrong about this, I don’t really have much experience in teaching Go.
Yes, that is right. Seki requires the existance of Dame. The existance of Dame prevents the confirmation of territory. Without the confirmation of territory the game cannot end. Therefore Seki can only exist when either A. the game ended prematurely (resignation, forfeit, etc.) or B. someone violates the rules. If a game is played correctly there can be no Seki at the point of score determination.
Sounds weird but that’s what the rules state. I’m open to changing my view on this if someone can point me to anything in the rules that states otherwise. I mean I might have overlooked something or there might be a translation mistake. But I doubt that is the case. The Korean rules e.g. are written even more ambiguously in Korean than in the English translation.
I agree that filling dame is good for beginners (and I think for all players). When I help beginners in my club to finish their games, I always suggest or lead them to fill dame before counting. I tell them it’s to avoid confusion between territories and neutral areas while scoring. My intention is to make it 2nd nature for them to fill dame before scoring.
I think this statement lead to confusion for @Feijoa. If players were indeed required to fill all dame, then all sekis would collapse by self-atari during the dame filling stage.
I don’t think that was the intention of the writers of J89 rules. So arguably the players are not required to fill all dame.
But if the players leave a dame unfilled as in artice 8, it will neutralise areas that touch stones that touch this dame. So the result of this is that there is no territory in seki situations.
I think that was the intention of the writers of the J89 rules, trying to define “seki” as well as “no points in seki” in one go, without even referring to eyes as they might have done in the J49 rules.
OK, I didn’t think this was what you meant with that previous statement, but it seems I was confused and @Feijoa understood you correctly after all.
Anyway, don’t think that is the correct interpretation of J89 rules. I’ve never heard of players needing to collapse sekis before scoring.
In Co§9.2-1 and Co§9.2-2 it says the dame", not all dame. I suppose the original text just says “dame”, because Japanese doesn’t use articles.
Confirmation of the life and death of stones and territory requires that the players fill the dame and add any necessary stones inside their territory, in accordance with Article 8.
If the players agree, they may fill the dame and add other necessary stones after stopping the game, in which case these are not moves as defined by the rules, and need not be played according to the rules.
I interpret Co§9.2-1 as the players filling dame voluntarily, because of the consequence of losing points when they miss a “proper” dame (i.e. a neutral point not originally touching a seki, like diagram 15 you posted earlier).
It would make no sense if that statement meant an obligation to self-atari stones in seki before scoring. How would that even work? Which player should self-atari first?
I interpret Co§9.2-2 as the players are allowed to fill dame after stopping the clocks and game recording, and they can fill dame out-of-turn and simultaneously.
I suppose this just describes common practice in finishing pro games. At least I think I have seen this practice in videos of Japanese pros finishing and scoring their games.
