In this game for move 60 I wanted to play A12 but was prevented from doing so. It didn’t tell me why in any way, and it was not a repeated board state, so I have no clue what rule could be causing this. Is this a bug, or some rule that I am missing?
3 Likes
It would actually be a superko repetition from a few moves prior. Looks like a rare PSK vs SSK difference
7 Likes
Ah, that is an interesting one; have not encountered superko before, thank you
2 Likes
That’s a new one. What’s a super ko?
For what it’s worth, even if you could retake the white group is dead anyways, black could simply play F9 to take your liberty. It’s not a seki. This situation is called “Eye kills no-eye.” Black has a real eye but white doesn’t, so White’s group will die once outside liberties are gone.
1 Like
There is the basic ko rule, which I’d paraphrase as “you can’t repeat the previous position”. This is the traditional ko rule.
Then there were some people in modern times who thought it made sense to change that to “you can’t repeat any previous position” and they managed to convert some Western countries to use this modified rule.
That modified ko rule is called “superko”, and there are even a few variations: Superko at Sensei's Library.
2 Likes
Yeah, Superko is nice because it is extremely simple, and yet handles all the edge cases which arise under Simple Ko (undoing the opponent’s immediately prior move is forbidden) with no need for any special cases to be defined
1 Like
The superko rule is indeed simple in theory, but in practice it can be hard to verify a violation (i.e. detecting a long cycle in your IRL games, such as Cycle at Sensei's Library, especially when both players make an effort What do you think Shin Jinseo and Katago's ranks would be on OGS? - #8 by gennan).
Edit: OP’s question is a similar case as the example under 3-move cycles on that senseis page.
No more so than the repetition rule in chess (which also has no limit on how long ago the repeated position can have come from). I like chess’s approach of just saying it only applies if one of the players claims the draw. They also say that if it repeats twice more, it’s automatically a draw regardless of player desires, but that’s not essential to using such a rule irl. In both chess and Go, the vast majority of applications are incredibly trivial to judge and do not rely on remembering positions from many moves prior
In chess the repetition rule ends the game by declaring the game a draw. It doesn’t ban long cycles.
To me that seems more similar to ending the game by No Result (Japanese/Korean rules and Chinese rules in practice), than by banning long cycles by the superko rule (AGA/BGA/FFG/NZ rules and Chinese rules on OGS).
The only thing that different is what happens when it triggers. The relevant identicality is in what information it requires (strictly speaking) to enforce (all previous game states). My point was that since that requirement does not prevent three-fold from working well over the board in chess (and I mentioned the implementation detail they use to make it more forgiving for fallible humans, one which can be easily adapted to Superko), it also shouldn’t be seen as preventing Superko from working well over the board
What’s that?
Positional Superko (just the position matters; who’s turn it is to play is ignored) and Situational Superko (the position and who’s turn it is to play must match to trigger the Superko Rule). In both cases captures are ignored, of course
And you suppose that the OP knows this (even I doubt that many understood it)
I assumed that I had provided standard terms they could look up or ask for clarification on if needed. The only reason I never defined Superko until now in this thread, is that genann already did
1 Like
I assume PSK and SSK are far from being anything like obvious abbreviations for a vast majority of players but ok if you feel so. I won’t go further.
I just looked up superko on sensei’s and it had PSK vs SSK; Samraku was descriptive enough.
5 Likes