Japanese rules, komi 7.5
What is the result of this position ? No result?

If I counted correctly and had the right sequences in mind then it appears to be similar to triple ko and neither player would wish to resolve it. Do you have a different idea?

Yes kvasir for me the RESULT should the same as the result of a triple ko.
The analysis looks for me quite different : it looks only a two stages ko with each side having an infinity of two stages ko threats. Quite uncommon isn't it?

Here is the loop in ten moves (the ko threats are too big to be ignored):

If I allow myself to call a direct ko "a three move ko", because the number of moves that it takes each player to win the ko is three when added together, then this position has three "four move kos". That is the send two, return one can, in this way, be said to be a four move ko. Maybe the position presented is unusual but many things are possible, for example what about a triple ko that mixes three and four move kos? I am not sure if that is unusual, maybe there are real game examples. Probably you could also make triple ko that centers around a ko with a larger and larger move counts, maybe there are interesting patterns? or there is no real pattern? Humans are often quick to see patterns that then only work for a few cases and then they will see a pattern of when their pattern works

In my example I understand you call the bottom left corner position a four move ko due to the possibility of the sending two return one possibility. Isn'it a quite unusual genrelization of the ko concept?

Let my try to understand what you mean by a n move ko.
Do you call the following position a nine move ko?



Anyway your idea is interesting:



if black puts a stone at "a" black is unconditionnaly alive.
If black puts a stone at "b" black creates a nine move ko which is not as good!

In the sense that this is a "send eight, return one" that could be recycled it is would be a ten move ko. White can capture in nine moves, black can prevent this sequence in one move, that is ten move in total. If it can't be recycled then there is no point in equating it to a ko, but "send two, return one" can sometimes be recycled and also "send three, return one". Maybe you can make this "send eight, return one" work in a triple ko?



Yes. People tend to count approach moves or steps in the ko fight. In that case "send two, return one" doesn't appear to be similar to ko but it is sometimes similar. I have been thinking in this way instead because many kos cycle through steps that are hard to understand, but counting the moves that white and black would need to make appears to me to be the key thing to understand even if one fails to understand other subtlety of a ko fight.

The idea of extending the ko concept is fine for me but I see a major difference between a ko and a ko threat (or an extended ko and an extended ko threat).
What happen in the example above?
Black plays in "A" zone which appears to be a ko (an extended ko). Then white answers by playing in zone "B" in order to be able to retake the ko in zone "A". In that sense zone "B" looks like a ko threat and not a ko. The global sequence for a loop is:
Black "A", White "B", Black answers in "B", White "A", Black "C", White answers in "C", Black "A" ... which is a logic of a ko fight with an infinity of ko threats.

The logic of a triple ko is different for me:
Black "A", White "B", Black "C", white "A", Black "B", White "C", Black "A" ...

In any case I agree the results are obviously a loop in any cases, but I prefer to distinguish the two different logics :
1) a ko fight with an infinite number of ko threats
2) a triple ko

With this distinction "A" zone is a ko (extended ko) while "B" and "C" zones are ko threats (extended ko threats).

Could this approach make sense for you?

Sure. It is still interesting to note that the ko threats need to be recyclable for this position to work as a cycle. Sometimes there are different but useful ways to look at the same thing.

Click Here To Show Diagram Code

[go]$$c One ko or two?
$$ | . . . . . . . . . . .
$$ | . . . . . . . . . . .
$$ | . . . O . . . . . . .
$$ | . . O . X . . . . . .
$$ | . . O O O X X X X X .
$$ | . X O X X O O O O X .
$$ | X O . @ X X O O . O .
$$ | . O O X . . X . O . .
$$ ------------------------[/go]

Interesting position.
When looking at the expected sequence for me the logic is clearly a direct ko, each side having an infinity of direct ko threats. Here again it is a loop but not a triple ko logic:
White has just taken the ko, black plays a local ko threat, white answers the ko threat, black retakes the ko, white plays a local ko threat, black answers the ko threat, white retakes the ko ...
In your example the ko and the local ko threats are just "simple" direct ko and "simple" direct ko threats instead of n move ko and n move ko threats. In that sense our examples are quite similar.

Just for fun I tried to put your example in a 7x7 board. Here is the result: