Odd Cases 🤔 in the Japanese Rules

I believe @hexahedron was referring to this discussion on the LifeIn19x19 forums:

https://lifein19x19.com/viewtopic.php?f=45&t=17089

1 Like

Nice!

The “non-Japanese-result” positions that I was able to extract from that discussion were only:

  1. Torazu sanmoku (KataGo gets the pre-1989 result though, so it is quite Japanese too :stuck_out_tongue: )
  2. Molasses Ko (I think that in Japanese rules an antiseki (0 points locally) should be the final result, while in KataGo a no-result should occur because both players want to be the first to pass, in order to win when the game resumes).

So far, the mismatch seems to be remarkably small!

1 Like

You may also be interested in positions A,B,C here:

https://lifein19x19.com/viewtopic.php?p=251494#p251494

2 Likes

Interesting!

However I have looked at position A and I think that Japanese rules would consider that white does not need to add an extra move there. This is because all white groups should be alive, the corner group in particular because in hypothetical play it enables placing new permanent stones (at F19, G19).

This is the very same justification of why the two black corner stones in https://senseis.xmp.net/?SekiInAreaAndTerritoryScoring are considered alive and not dead, and why this is seki and not “white lives with prisoners and territory” under japanese rules: White can capture those two stones, but capturing them enables the placing of new stones “in the locality”.

This also matches the Jasiek 2003 formalization, as far as I have been able to analyze, because the new permanent stones are within the group’s local-2.

So position A would be a difference because if my understanding is correct, KataGo rules imply that you must add an extra move to defend or you lose the corner group.

I have not looked at the other positions yet but they all seem very interesting.

2 Likes

I have seen all three positions. My understanding is that in all three cases, actual Japanese rules in use say that no extra defense is necessary (because new stones in the locality are enabled by the capture, which deems the corner group alive).

It seems that in KataGo rules, positions A and B require an extra defense in order not to lose the group during the extra phase. In position C, no extra defense is needed in KataGo rules.

So it would seem to me that only A and B generate a slight 1-point difference, and C is a “non-difference”.

1 Like

Fun discussion of anomalous positions here for KGS’s attempt to detect seki in order to implement Japanese scoring (taking advantage of human players marking life/death appropriately but still having lots of fun corners to consider):
https://senseis.xmp.net/?WmssDameSekiChallenge

4 Likes

Have you tried applying versions of the Japanese rules to 19x1 boards?

3 Likes

That’s an entirely different game. Except for sizes small enough that MCTS can brute-force the entire game, you’re not going to be able to say anything sane about that board, because it’s going to be entirely alien to the neural net. You’ll need to run the whole AlphaZero process to do specific training for boards that size.

2 Likes

I just meant the japanese ruleset(s) old, modern etc, (nothing to do with ai?). I saw people trying to debate that everything on a 19x1 board was seki and not territory

2 Likes

I’m not familiar with playing games on 19x1 boards, but I think that not everything would have to be seki. Let’s consider this 9x1 game.

:heavy_minus_sign::white_circle::heavy_minus_sign::white_circle::heavy_minus_sign::black_circle::heavy_minus_sign::black_circle::heavy_minus_sign:

With the center point being dame, I think the two stones next to it should technically be considered in seki. The other two stones are simply alive, and each player has one point of territory. At least, that’s how I would interpret the modern Japanese rules.

2 Likes

I don’t think it’s impossible to solve 19x1. It’s smaller in search space than 5x6, which has been solved. Up to 12x1 has been (weakly) solved here 12 years ago. It took less than 20 minutes, but the paper didn’t go beyond 12x1, to quote:

Results for 1×n boards are included for completeness. However, due to some design constraints MIGOS II is rather inefficient on 1×n boards. Consequently, comparison of search times with results on 2-dimensional boards is not straightforward. Therefore, we stopped computing 1×n boards after the 1×12 board. Similar inefficiencies also play a role on 2×n boards when n becomes greater than 10. Here we stopped after the 2×11 board.

3 Likes

It only looks like 6x1 to me?

2 Likes

Nothing - spot 1
White stone - spot 2
Nothing - spot 3
White stone - spot 4
Nothing - spot 5
Black stone - spot 6
Nothing- spot 7
Black Stone - spot 8
Nothing- spot 9

3 Likes

Doh, of course.

2 Likes

After you mentioned it I could see it as a 6x1, because normally we can see the lines between the stones.

maybe instead of

:heavy_minus_sign::white_circle::heavy_minus_sign::white_circle::heavy_minus_sign::black_circle::heavy_minus_sign::black_circle::heavy_minus_sign:

have

:heavy_plus_sign: :heavy_minus_sign::white_circle::heavy_minus_sign::heavy_plus_sign: :heavy_minus_sign::white_circle::heavy_minus_sign::heavy_plus_sign: :heavy_minus_sign::black_circle::heavy_minus_sign::heavy_plus_sign: :heavy_minus_sign::black_circle::heavy_minus_sign::heavy_plus_sign:

[I know the pluses make no sense since there’s no intersection on the 9x1 or nx1 board.]

That’s kind of interesting actually. I was wondering how to interpret it. So it’s because the dame can’t be filled without some kind of capturing occurring. It’s kind of like that send two return one position kind of.

1 Like

Yes, playing at the dame point would be like sending two, returning one. Hence, neither player should do that since it would only lose a point.

If plus signs are preferred, I would rather use just a single symbol, since having three emojis spaces out the stones too much.

:heavy_plus_sign::white_circle::heavy_plus_sign::white_circle::heavy_plus_sign::black_circle::heavy_plus_sign::black_circle::heavy_plus_sign:

:heavy_plus_sign::white_circle::heavy_plus_sign::white_circle::heavy_plus_sign::black_circle::heavy_plus_sign::black_circle::heavy_plus_sign:

Note that there are no spaces between the emoji codes, which leads to them being placed closer together.

:heavy_plus_sign: :white_circle: :heavy_plus_sign: :black_circle: :heavy_plus_sign: leads to
:heavy_plus_sign: :white_circle: :heavy_plus_sign: :black_circle: :heavy_plus_sign:

Another 2D diagram example

:heavy_plus_sign::white_circle::white_circle::black_circle::heavy_plus_sign:
:white_circle::heavy_plus_sign::white_circle::black_circle::heavy_plus_sign:
:white_circle::white_circle::black_circle::heavy_plus_sign::heavy_plus_sign:
:black_circle::black_circle::heavy_plus_sign::black_circle::heavy_plus_sign:
:heavy_plus_sign::heavy_plus_sign::heavy_plus_sign::heavy_plus_sign::heavy_plus_sign:

Code:
:heavy_plus_sign::white_circle::white_circle::black_circle::heavy_plus_sign: :white_circle::heavy_plus_sign::white_circle::black_circle::heavy_plus_sign: :white_circle::white_circle::black_circle::heavy_plus_sign::heavy_plus_sign: :black_circle::black_circle::heavy_plus_sign::black_circle::heavy_plus_sign: :heavy_plus_sign::heavy_plus_sign::heavy_plus_sign::heavy_plus_sign::heavy_plus_sign:

However, one has to clarify that this is the top-left corner

2 Likes

Could try and get the markdown that they use on Boardgame stack exchange where they $$ and it turns in to a go board image :slight_smile:

You can even do nx1 boards :slight_smile:

2 Likes

So is this the most modern version of the Japanese rules linked in the OP?

Is there a formulation of the Japanese rules that are a bit cleaner and don’t involve a case where both players lose?

Actually in that link I don’t like

Article 7. Life and death

1. Stones are said to be " alive " if they cannot be captured by the opponent, or if capturing them would enable a new stone to be played that the opponent could not capture. Stones which are not alive are said to be " dead ."

I know one of the comments says that you can apply this recursively to various snapback situations - (which just seems like a pain that should be ironed out of the wording of the rules).

image

I really hate that wording however, in principle it’s supposed to tell me that in situations like the above blacks two stones are alive because when white goes to capture them black can play at 4-2 (row-column) and that stone can’t be captured. But actually couldn’t you just read it where black just adds any stone to an already alive group? It just doesn’t make any sense.

That and when you have to add a commentary on your list of rules that’s actually longer than the list of rules, maybe the rules aren’t very clear.

1 Like

In my original post, I linked to an English translation of the official 1989 Japanese rules. I believe that is the most recent official rules, as used by the Go associations in Japan.

Several people have proposed alternative territory scoring rules that at least loosely adhere to the traditions of the Japanese rules, although they excise certain parts they don’t like, which necessarily departs from the official Japanese rules.

See the “Other Japanese Style Rule Sets” section on this page:
https://senseis.xmp.net/?JapaneseRules#toc3

I think the “Japanese-like rules” used by KataGo are a particularly interesting example. These are influenced by the Spight Japanese Style Rules, and although they introduce a significant departure from the Japanese rules by adding the ability to “play on” (under special rules) to decide life/death, they aim to produce almost the same strategic results as the official Japanese rules (except in some relatively exotic cases).

These rules were mentioned earlier in this discussion, involving the author of KataGo (and those rules), @hexahedron.

2 Likes

A Curious Position

In another thread, @S_Alexander offered an interesting position that occurred in one of their recent games.

Spoiler: strategically speaking, this local position most likely would ultimately be a seki, unless it turns out to be useful to sacrifice as a potential ko threat. However, even without an ongoing ko fight, understanding why this is a seki involves some interesting considerations in regards to the rules. Clearly, any further move by Black would be self-atari, although which self-atari crucially matters for finding an effective ko threat. White, on the other hand, could also self-atari to make a ko threat, but also have potential moves that would lead to some interesting considerations of the rules.

Possible Usage as a Ko Threat

Let’s first consider, hypothetically, if there was a large, ongoing ko fight elsewhere on the board.

By Black

If White has just taken the ko, and it’s Black’s turn to find a threat, then Black could potentially consider playing T3 as a ko threat. Of course, this would be self-atari, which would only strategically make sense if potentially winning the ko fight was worth offering the sacrifice of this large corner. If the ko fight elsewhere was indeed worth a very large amount of points, then White might even ignore this threat (to end the ko elsewhere), which would allow Black to then play T1, which would convert the status of this corner from a seki into another ko that would need to be settled.

Note that if Black instead attempted to play R2 as a ko threat, that would be a huge blunder that should be ignored by White, allowing White to win the ko fight elsewhere. Then, even if Black later followed up on the threat by playing at T1, White could just take back with S1, which would turn Black’s group into a dead “Bent Four in the Corner”. Hence, Black playing at R2 would only die in gote, since R2 by Black does not require an immediate answer by White.

By White

If Black has just taken a ko elsewhere, White might consider to play in this corner as a ko threat. White could do so by playing at T3, which would prompt a response by Black at T1, if responding to this threat is worth more than winning the ko elsewhere. White would now no longer have any immediate ko threats in this unsettled corner position. White could eventually take back at T2 to reset the local position into a seki, or Black might eventually follow-up with R2 to eliminate white from this corner (to gain some territory and captures), however, both of those moves are gote, which makes them ineffective as ko threats.

On the other hand, White could instead play a ko threat with R2, which would necessitate a response from Black at T1 (assuming that this threat is big enough). The seki position has now collapsed and White’s remaining stone in the corner is dead. However, White could eventually play at S1, perhaps as another ko threat, which would require Black to respond at R1 in order to prevent White from coming back alive (in seki) with a followup play at R1 to nullify Black’s territory in the corner.

Note that after the sequence W R2, B T1, W S1, B R1, White’s stones are dead, and Black is fully alive (not in seki) in that corner. This position is settled, and neither player should waste any more moves in that corner. However, it is of interest to note that this is an example of a “Dead Ko”, where playing on to actually capture the White stones could start a ko pattern, but White can never “win” this ko by resolving it in their favor. Ultimately, this position should be ruled dead during the confirmation phase, however, this situation becomes very peculiar when elsewhere on the board there is also a source of infinite ko threats (such as the double ko seki that we discussed earlier), which would cause this to become an example of Moonshine Life. I will discuss Moonshine Life in more detail in a future post.

A Cycle that Does Not End with “No Result”

Now let’s consider if there are no ko fights elsewhere and everything else on the board is fully settled, and imagine that White is behind in points (with this corner shape called as a seki).

In this case, although the position should locally be a seki, it is possible for White to still force the game to continue by playing the 3-move “sending two, returning one” cycle:

  1. White at T3
  2. Black at T1
  3. White at T2
  4. Black passes

Note that White yields two prisoners to Black in each cycle, while only getting one back, hence the name “sending two, returning one”.

Assuming that the game is close enough that Black cannot afford to lose those stones, this cycle is forced to continue (at least for a while), and thus White could force the whole-board position to repeat. White might mistakenly think that they could claim that this game is now a “no result” (which they might prefer over conceding the loss), however, recall the exact wording of article 12:

Article 12. No result

When the same whole-board position is repeated during a game, if the players agree, the game ends without result.

The crucial words are “if the players agree”. If White tried to pull such shenanigans of forcing a “sending two, returning one” cycle in order to attempt to claim a “no result” for cycling, Black should simply refuse. The Japanese rules do not have a superko rule, so this cycle can continue, if White persists, however, since Black gets a net gain of one prisoner in each cycle, eventually Black will gain enough points to simply abandon their stones in the corner and still win. Thus, there’s no strategic advantage for White in persisting with this cycle, as all they would accomplish is to lengthen a lost game, which would probably be viewed as rude (if White understood what they were doing).

5 Likes