KataGo planned rules - drafted

[Bill Spight]

Effectively, on the board outside the molasses ko, the game becomes one of no pass go.

I think you are assuming a kind of superko rule is in effect. Using SSK, the game will likely end leaving the local shape as a seki. Using PSK, ugly(?) no pass go will result.

Correct about my assumption. Sorry, I was relying upon memory from the 1990s. People say that situational superko produces a different result than positional superko. I don't think so. See SGF file.



The first player to pass can be made to violate both a positional and situational superko.

However, Japanese, Korean and Chinese rules do not adopt any superko rules.

Right about Japanese and Korean rules. Chinese rules have a positional superko "in theory", but allow the organizers or authorities to have different rules, which I gather they usually do.

[moha]

any (balanced) repetition before first stop is a real draw, the rest is much easier handled in dispute/resumption phases.

It is not much easier in my experience. Yes, if repetition occurs after the first phase, there is a high chance that someone is just trolling. But who is to blame? In a rule where a connected moonshine life is allowed to live, the repetition after the first phase may still result in a draw, and no one is to blame.

For me the very reason to separate the main phase from resumption phases is to restrict special ko rules (necessary for moonshine) to those later phases. So a player who wanted to force repetition but forgot (?) to do so and passed instead may not be able later, whether a simple hack like superko or some more sophisticated ko pass rules are used there. This is the price of having the first phase (the main game) free from extra rules, and to ensure nothing unexpected will rob the players from their legal moves in it. Basically, to allow the main game to progress in its natural way.

I don't understand the exact situation you are talking about. For me, whichever decision (draw or dead) is acceptable for the connected moonshine life. The problem is that the sequence of repetition for connected moonshine life and separated moonshine life are virtually the same, which means a logical rule allowing the life of a connected moonshine life is likely to allow a separated one to live

I meant a connected moonshine is not automatically draw, even if allowed, the player who can repeat may choose to play for a win (letting it live), and only "cash out his draw option" if that doesn't work out. This issue does not arise if it can be killed in dispute phase. BTW there are logical rules that can differentiate the two cases, for example, explicit localization and control (Pauli) - even though I prefer identical handling (cycle/ko behaviour over locality).

For players who are playing normally, they wouldn't even notice that a sequence analyzer is included in the rule. Only when some weird situation occurs or someone is really trolling, it's the time a program or referee to come up and try sequence analysis.

This is still quite different from other board games, where the players themselves have no problem applying the rules (which may even be necessary for gaining wide adoption).

Chinese rules have a positional superko "in theory", but allow the organizers or authorities to have different rules, which I gather they usually do.

I wonder if there were ever any serious games (under Chinese rules) where triple ko was NOT treated as draw? I always thought this is unversally accepted in all three main go playing countries (and their histories), with maybe even casual players being aware of this rare case.

[jaeup]

I meant a connected moonshine is not automatically draw, even if allowed, the player who can repeat may choose to play for a win (letting it live), and only "cash out his draw option" if that doesn't work out. This issue does not arise if it can be killed in dispute phase.

I keep saying "anomaly" is a subjective term, so if you think it is an anomaly, I can't say it is wrong. However, I just want to point out that most people does not think having a choice of draw is not an anomaly. In a game under Chinese rule where a bent-four-in-the-corner and a double ko seki shape is left, one player may have an irresistible choice of making a draw. It is up to the player's calculation, and he can simply try scoring and win the game.

I wonder if there were ever any serious games (under Chinese rules) where triple ko was NOT treated as draw? I always thought this is unversally accepted in all three main go playing countries (and their histories), with maybe even casual players being aware of this rare case.

As far as I know, the answer is no. A triple ko is a de fecto draw in China, whatever the rule text says. (These kind of things make the life of rule theoreticians much harder. You may have years of rule study, but a newbie pro may still know some practical application of the rule you never learned from texts.)

[jaeup]

Correct about my assumption. Sorry, I was relying upon memory from the 1990s. People say that situational superko produces a different result than positional superko. I don't think so. See SGF file.



The first player to pass can be made to violate both a positional and situational superko.

I am a little bit unsure if we are talking of the same subject. Just to make life easier, let us assume that White move 0 was at (1,C), and all the upper sides are played so that no more meaningful(in a traditional sense) plays are left other than the molasses ko.

Under every rule, Black 1 is necessary unless Black gives up one's group.

In PSK, if Black chooses to pass as 5, Black's move 9 will be prohibited. Thus Black must jump into the pass fight as 5, where the first passer loses the game.

In SSK, Black can pass as 5. Black's move 9 is allowed, and White's 10 is a pass. Now Black passes as 11. (Maybe this is something you missed. White cannot play at (1,E) as 12.) In AGA rule, both players will pass one more time and the game ends leaving a molasses ko seki. (Any other choices are either prohibited or results in one's loss.)

[Maharani]

In SSK [...] In AGA rule [...]

Not sure if this applies to that situation, but Sensei's further distinguishes between situational superko and natural situational superko. I find this interesting:

Positional Superko – Forbids a board play to repeat a board position – Chinese rules (in theory), Tromp-Taylor Rules
Chinese Superko – A play may not repeat a board position by means of basic ko or sending two, returning one – Chinese rules (in practice) – This, according to the sixth meeting of the International Go Rules Forum, is what the Chinese rules were intended to specify.
Situational Superko – Forbids a board play to repeat a board position with the same player to move (regardless of how it previously arose) – AGA rules, New Zealand rules
Natural Situational Superko – Forbids a player’s board play to repeat a board position that they created with an earlier board play – BGA Rules 2009 – Note that it could allow an immediate ko recapture if the first capture was preceded by a pass. BGA rules combine this with the basic ko rule to forbid immediate ko recapture. According to Terry Benson, AGA rules were meant to use Natural Situational Superko, but everybody else interprets them with Situational Superko.

NSSK sounds the most elegant to me.

[Bill Spight]

In SSK, Black can pass as 5. Black's move 9 is allowed, and White's 10 is a pass. Now Black passes as 11. (Maybe this is something you missed. White cannot play at (1,E) as 12.) In AGA rule, both players will pass one more time and the game ends leaving a molasses ko seki. (Any other choices are either prohibited or results in one's loss.)

Thanks.

SSK plus two passes to stop play plus preservation of the superko ban in the resumption or encore ends play in position A. Whether to call that a seki or not is another question.

[moha]

I meant a connected moonshine is not automatically draw, even if allowed, the player who can repeat may choose to play for a win (letting it live), and only "cash out his draw option" if that doesn't work out. This issue does not arise if it can be killed in dispute phase.

I keep saying "anomaly" is a subjective term, so if you think it is an anomaly, I can't say it is wrong. However, I just want to point out that most people does not think having a choice of draw is not an anomaly. In a game under Chinese rule where a bent-four-in-the-corner and a double ko seki shape is left, one player may have an irresistible choice of making a draw. It is up to the player's calculation, and he can simply try scoring and win the game.

No, I wouldn't go as far as calling this an anomaly (which I'd only say on clearly wrong rulings). It is just a suspicious situation that would not otherwise arise. The difference to bent4+dks is that bent4 is a known and accepted not-always-dead and play-out shape (area rules, unremovable threats). With moonshine life, the only reason one would think of it as alive (in the connected case only) is a side effect of explicit localization - which in itself is an unnecessary and ad-hoc rule invention in my opinion. OC there is also the case where actual capture is necessary during the main game already, for nearby fights - but in that case it is not moonshine life anymore!

I also agree on subjectivity - which I think is another advantage of dividing main phase <> resumptions: all subjectivity is pushed to dispute phases, leaving the main game clean and unquestionable.

BTW, about molasses ko: the fact that in almost all such superko analyses there are some oversights can form a rather strong argument against superko (at least its use in the main game).

[jaeup]

SSK plus two passes to stop play plus preservation of the superko ban in the resumption or encore ends play in position A. Whether to call that a seki or not is another question.

Yes, it is somewhat different from a normal seki. In some sense, the whole board is frozen and players just can't touch anywhere on the board. A Whole board repetition with period 5 is another example that SSK and PSK are working differently. https://senseis.xmp.net/?SendingThreeReturningTwo

Another important lesson is this. Lifting ko ban or superko ban after a pass (or other reasons like resumption of game, change of phase and etc.) can be dangerous. In my exeprience, efforts to remove one anomaly usually results in another unexpected anomaly.

[Bill Spight]

SSK plus two passes to stop play plus preservation of the superko ban in the resumption or encore ends play in position A. Whether to call that a seki or not is another question.

Yes, it is somewhat different from a normal seki. In some sense, the whole board is frozen and players just can't touch anywhere on the board. A Whole board repetition with period 5 is another example that SSK and PSK are working differently. https://senseis.xmp.net/?SendingThreeReturningTwo

Well, the molasses ko has an odd repetition only if you count a pass as a play. The game tree of the molasses ko has no passes. AFAIK, Ing was the one who introduced the idea of a pass as a play. Otherwise, relinquishing the right to play was a typical locution.

Another important lesson is this. Lifting ko ban or superko ban after a pass (or other reasons like resumption of game, change of phase and etc.) can be dangerous.

Well, the main difference between the framers of the Japanese 1949 rules and Honinbo Shusai and Go Seigen was the question of whether there should be a final ko ban in effect. Shusai, Go Seigen, John Tromp and others say yes. The Nihon Kiin, Ing, and others say no. Yasunaga's "Draft Constitution" said no.

第７条 交互着手の権利を放棄せざる場合には同型反復を禁止す。
第８条 終局、交互着手の権利を連続３回放棄せる場合。

( http://park6.wakwak.com/~igo/igorule/yasunaga.html )

That explains his 3 "pass" rule. The first pass may have been forced by the superko rule, but the second and third passes are unencumbered, and may be taken as indicating a willingness to stop play.

In my exeprience, efforts to remove one anomaly usually results in another unexpected anomaly.

I don't think that the question here is one of anomalies, but of what it means for a position to be final. If the rules embody the preferred definition, whatever it is, then let the chips fall where they may.

[moha]

Another important lesson is this. Lifting ko ban or superko ban after a pass (or other reasons like resumption of game, change of phase and etc.) can be dangerous. In my exeprience, efforts to remove one anomaly usually results in another unexpected anomaly.

I think the latter statement has little to do with the former. IMO in all cases where passes lifting bans - supposedly - leads to problems, it is actually some other, poorly formed rule producing the anomaly, with passes only uncovering it. With normal ko, bans are temporal not positional, so a pass works naturally (unlikely to produce unexpected things by itself). Not being able to recognise this is the very source of some superko anomalies.