Renaissance rules

Some thoughts about rules, and a new proposal:

Japanese and Chinese rules are often thought of as illogical mess of traditions. But AGA/NZ rules also have some dubious decisions behind them. In particular, the removal of long cycle draws, disallowing play in triple kos always seemed like an arbitrary and unnecessary change.

Why this change? The game becomes less interesting and incompatible to how pros (and most amateurs) actually play. You may think this is to avoid draws - but at the same time NZ sets komi 7, for even more frequent draws. You may think this is the price of simple and logical rules - but returning to the complete game is also easy.

For AGA/NZ rules, one can play the main game with basic ko + a threefold repetition rule, and only apply superko in disputes, after resumption.

For an even better solution (without phase-specific rules) superko can be replaced by a minimal variant, which only forbids recreation of the last position both players passed in successively. Which not only behaves much better, it is also easier to use than full superko.
(anybody with interest in rules, please review and check for problems)


This statement seems a bit vague and confusing. Couldn’t one force the application of superko by raising a dispute? I guess the key part is that one can only do that for certain situations where one can safely pass to reach the scoring phase in order to resume. I think I roughly understood, but I’m not too sure.

Overall, I don’t like the practice of making rules sets more complex in order to engineer certain outcomes for rare edge cases. I subjectively prefer a simpler rules set.

From the page that you linked (Renaissance Rules at Sensei's Library), this phrase seems a bit condescending and smacks of colonialism:

the Western rediscovery of the complete game as played in China, Japan and Korea


Yes, I didn’t mean anything more with “dispute” than a new phase, requested resumption after the first phase stopped by successive passes. As you wrote, the ability itself to reach this phase could serve to differentiate between true repetition like triple ko (which lead to draw and don’t reach dispute phase) vs. things like moonshine life (which can be postponed until then).

This seems like a reverse direction: I think the rules need to be just as complex as to describe the actual game most people play. (Chess would be simpler without en passant, wouldn’t it? :slight_smile: ) AGA/NZ doesn’t do this at the moment, but could do with only minimal changes.

Thanks, I will try to rephrase.


I would put a full stop here :laughing: There are already too many rulesets! The ultimate simplification is to have just one ruleset to rule them all (to find them, to bring them, and in the darkness bind them :wink:)



Upload of the comic behind @jannn 's link for those too lazy to click:





Realistically, there will always be at least two rulesets (Japanese and Chinese), since both area and territory scoring has some significant advantages which are missing from the other. Maybe we can make that three with the territory+area hybrid.

The Korean<>Japanese split is a curious situation, which may not remain forever since Korean rules - while based on an attractive idea - do not always work well (moonshine life sometimes alive etc.). In any case the differences are minor.

But the AGA/NZ schism is different. Removing long cycle draws IS a major change that results in a new go variant. Maybe not of an everyday significance but definitely an everyyear one (triple kos happen regularly in pro play).

And again, where are the benefits? This actually loses the most significant advantage of area scoring (beginner-friendlyness).

What are the differences between the Korean and Japanese rules? Are the official Korean rules available online somewhere?

Two versions are linked from sensei’s (and a third was uploaded to L19 by jaeup). The difference is how L/D works. Japanese use restricted ko in hypothetical play, Korean is about explicit local view (that’s why a local moonshine behaves differently). There was also a recent topic about this on L19.

1 Like

There are multiple versions of the Korean rules? Are they materially different from each other? Do you know which one is the official version that would be used by the Korean professional Go association for its tournaments? Could you help clarify exactly which links are pertinent?

IIRC they are minor revisions only (with maybe some more detailed than others). In any case should be a good start. The text itself is oc not too clear, the L19 topic I linked above has a backlink to jaeup’s old post that also includes his explanations.

1 Like

Yes, I’ve seen the Sensei’s library page, but I’m trying to understand what is the definitive source text of the rules.

What is this text that you are referring to? Is that online somewhere? Or just in a book?

Is the most definitive English translation simply this ??

The other one (“asian games 2013”) is from the bottom of the sensei’s page, the third one uploaded in that old jaeup post I mentioned above. The sensei’s page also has some info on revisions. I don’t know more myself. :slight_smile: (Edit: that third one also seems to be a 2016 translation, with a few words on revisions.)

Is this the correct file that you are referring to?

So, the other two versions are these?

However, all of those versions seem incredibly vague about what removable/unremovable (or capturable/uncapturable) stones are. They just provide some example cases, but neither seems to provide enough clarity to understand why or how:

Vague indeed. The 2013 version is the most detailed I think, at least that includes the key sentence “each position considered locally without regard to the whole board position”. As for local moonshine, this is confirmed by jaeup and matches the above quote.

Ok, dumb question: who exactly is jaeup? And why do they seem to have a better understanding of the situation?

I guess there is some Korean source text that explains the rules in more detail, or maybe just a better shared understanding among Korean professionals (even if that is not clearly written down)?

It just seems odd to have any sort of careful analysis of any rules beasts when some clear understanding of the technical details of the rules is not widely available.

1 Like

He is the author of a korean book about rules, who consulted korean pros trying to clarify things for his book. Local moonshine is also shown as an example in both 2013 and 2016 versions I think (including the “only one side can repeat” explanation, and the draw result).

1 Like

What is the origin of the name “Renaissance” rules? The Sensei Page does not say why they are named like that.

You say that AGA rules could be changed by moving superko only to the second phase, allowing draws/noresult by infinite repetition. Is there any actual proposal / scheduled revision of AGA rules in place? Or is it just a theoretical possibility, but without anybody at AGA planning to do or discuss anything? I ask because it is not clear to me which of these is the case based on the written article.


It just seemed an expressive name for what the rules try to do: return to the original Chinese/Japanese game with draw on repetition. As the “no territory in seki” topic on L19 shows, most traditional rule ideas have strong logic behind them, forged in centuries of go history. Draw on repetition seems the same.

I honestly don’t see why AGA didn’t try to do a logical Chinese ruleset, why did they invent a new go variant instead. The minuses (losing CJK and pro compatibility, superko application difficulties, beginner unfriendlyness) seem far outweight the pluses (one line shorter ruletext). But no, I don’t think AGA will change this even as an option.

  1. This objective seems a bit vague since there are important distinctions between the Chinese and Japanese rules, as well as some tweaks over recent history. Depending on what one means by “original”, perhaps omitting komi seems in line?
  2. Games are not “drawn”, but rather they are declared a “no result”, which requires replaying the game from the beginning to produce a final result (which is unlikely to be a tie, even with integer komi).
  3. Further, not all repetitions end with “no result”. For example, sending two, returning one is forbidden under Chinese rules, while under Japanese rules, it is strategically discouraged by the counting of captures (allowing one player to eventually break the cycle). Moonshine life is a arguably another example of a potential repetition that is essentially avoided by both rule sets treating that situation as dead.

What exactly do you mean by “a logical Chinese ruleset”?

It’s not really possible to be simultaneously compatible with the Chinese, Japanese, and Korean rules at the same time. There are discrepancies (usually in rare circumstances) between each pairing among those three.

So, are you saying that your proposal is literally just amending AGA rules by adding a single line? Could you clarify what that line would be?