So it looks like it might be the autoscore adding random markings on the board in the first place, and clicking the autoscore button “fixes” the problem by adding or removing markings randomly.
It’s more like some group will be wrongly be alive or dead or some territory will be marked incorrectly as dame, and I don’t know if one of us misclicked or they are trying to cheat or what. The autoscore always just fixes it and we can proceed. I haven’t noticed any problems with randomly incorrect scores in a properly finished game.
Fair enough. But I’d be pretty upset if I disagreed with my opponent on the status of a group, and then they used the autoscore to force the score despite our disagreement.
Take the bent-4 shape, for instance. If you’re experienced enough, you probably know how to kill it when it arises. But if you’re playing against a player who has never seen that shape before, and the shape arises in the game, they’ll probably think their group is either seki or depends on a ko.
Now, if you resume the game, and play a correct sequence to kill their group, that’s a satisfying ending for both players. You will have proven that the group is dead, and your opponent will have learned something - at the very least, if they still think that it should have been a ko or seki, they’ll have to wonder where they went wrong in the sequence, and they can choose to study it or not later.
But if you use the autoscore to force the bent-4 group to die without actually killing it, and I was your opponent, I’d felt really cheated. Here is my group that looks like a ko or seki, and you used an AI to force it to be dead instead. That would be a really disgusting experience.
Now, you can argue that you’re not really cheating, because you know that a bent-4 group can be killed, but if that’s true, why do you need an AI to do it for you?
1 Like
It’s just the realities of Japanese rules. You can’t force your opponent to start the ko, and you can’t force them to fix all the potential ko threats to be able to ensure they can win the ko if they do start it. Bent four in the corner is just dead as it stands as well a few other other shapes involving double or triple kos.
You don’t even need the autoscore to do it. You can do it manually and it’ll look the same to your opponent more than likely, unless it takes multiple clicks that they can see.
The autoscore doesn’t settle anything, it just set statuses and both players still need to agree to it.
Well, if White is allowed to unilaterally pass, that doesn’t really sound like they’re “playing it out”. That’s why I said “if they pass, you pass”. In my mind there would always be consecutive moves by both players until they agree. I guess in my mental model, the burden of proof always falls on the player who first passes.
By the way, for the same reason I stated at the end of my previous reply, I believe “passing” is another approximation that the rulesets make in the attempt to find a rigorous formalization of “when the two players agree, the game is over” – in other words, it feels to me that passing is only “in the spirit of the game” if it’s done with the intention of ending the game.
Your opponent does not need to continue playing moves.
No, this is not correct under Japanese rules. You could simply have a “dead ko” situation, which would require additional move(s) for you to resolve by capturing in a hypothetical play out.
However, even more crucially, is that under Japanese rules, a different sort of ko rule is applied during the hypothetical play outs for life and death confirmation. There are positions that cannot be successfully captured by simply resuming play (e.g., bent-four-in-the-corner with unremovable ko threats or moonshine life). There are also plenty of other weird situations involving sekis, false sekis, and kos.
The only way to resolve some tricky life and death disputes is to actually convince your opponent (or arbiter/moderator) that some esoteric part of the Japanese rules applies in your favor during the consideration of hypothetical play outs for life and death determination.
1 Like
I think you are confused about how to handle life and death disputes under Japanese rules. There is no “playing it out” to settle life and death disputes under Japanese rules, nor is there any compulsion for players to continue playing moves during life and death determination with hypothetical play outs.
What you describe with the compulsion for players to continue playing stones is kind of like the concept of pass stones (under AGA rules), but that is not part of the Japanese rules, and anyways it would not quite work properly, if the players have not also filled in all dame, or if some pathological situations arise where a player would be forced to reduce a living two-eyed group to one-eye, given a lack of another legal move elsewhere.
Your mental model about how Japanese rules works is incorrect.
It’s not my mental model for Japanese rules, it’s my mental model for the game of Go.
There’s “the game of Go”, which is, reasonably speaking, the set of rules you can teach in five minutes, and there are different attempts at formalizing a set of rules to handle complicated situations. Most if not all fail miserably at actually being rigorous, as far as I can see, which makes them a pointless exercise in bureaucracy unless you really need them for tournaments or on a multilingual online playing platform. For amateur games, I believe disputes should be solved subjectively and relying on reciprocal honesty.
Looking back, I realize in a few places I did word it that way (“players being forced to play it out”), but I never meant “there should be a rule forcing players to play disagreements out” as much as “imho it would be in the spirit of the game if players played disagreements out (especially if the alternative is relying on AI)”.
Also, isn’t this picture
something called “alive in double (triple, here) ko”? If there are no ko threats, White is uncapturable. It feels to me that that would make it unquestionably alive.
EDIT: oh, whoops, never mind, White can’t survive 
but in my mental model, they should still tenuki. If White “passes”, Black “passes” and says “I think those stones are dead”. If White disagrees, White has to play a move because it’s White’s turn. In my mental model, passing is not really a thing.
I might be wrong, but this also means that the score doesn’t change. For every move Black is “forced” to play to capture, reducing their own territory, there’s a move White is “forced” (for burden of proof reasons) to play in tenuki (or they can fill in the fake eyes if they want, to speed up the process).
I don’t know whether anyone noted this as I have read only about half the posts I missed, so I am sorry if I am repeating what has already been said.
The autoscore button is essential for one reason if for no other: to quickly counteract score cheaters by resetting the score.
2 Likes
I think all rules agree that the burden of proof falls on the player who says that some stones are dead.
If I think that all of my stones are alive, and you think that some of my stones are “dead”, then I think that the game is over, and you think that the game is not over. So it’s okay for me to pass, and you’re the only one who has to keep playing stones to kill mine.
This is an example of a 9x9 game with AGA rules. Black played the last move. Now White passes first, and Black passes too.
Then it turns out, White says the group at H8 is alive, whereas Black says it’s dead. Certainly the burden of proof falls on Black now; the game resumes so that Black gets an opportunity to try and capture the White stones, but White is allowed to pass and is not forced to add moves to the board.
So, Black will play a move, White will pass (still thinking that their stones are alive and that there are no more useful moves to play). Now if Black passes too, the game is over and White’s stones are considered alive. So Black doesn’t pass, and keeps playing moves to try and capture the White stones. White can either pass or answer by playing moves too. But you can’t force White to play if White thinks that their stones are alive.
It’s really important not to force White to play if White thinks that the game is over.
Imagine instead we had the following situation:
White still thinks all stones are alive, and Black still thinks the northeast White group is dead. If you force White to play, then any move by White will result in White stones dying.
2 Likes
I think it’s in the nature of Go that many things depend on tempo and not just on the configuration on the board.
Whether or not you can win a ko, whether or not you can make life for all your groups, whether or not you’ll be able to take advantage of the apparent potential you have in the early game, they are all things that often depend on the exact order you play your moves, on how well you can take advantage of sente, or in other words, on how much tempo advantage you gain or lose.
As the game gets closer to the end, there’s an “optimal order of moves” that becomes more and more clear, and better players will (or would) of course be able to start seeing it much earlier. Though the path was probably always there.
(It changes during the game because the players, including AIs, can’t see it and just make guesses, but there probably is an optimal game of Go for any position.)
Anyway, I think it’s in the spirit of the game to respect the importance of tempo management, and I think it’s very much in the spirit of the game that “passing” not be considered an actual legal move, but only an artificial convention.
Consequently, I think it’s also in the spirit of the game to let unsettled groups, and perhaps even sekis, be decided by the natural flow of tempo of the game. Whichever player happens to have the move that will kill their own stones has the burden to play it or accept that those stones are dead.
Plus, unless I hear a good argument against this, I don’t think it makes sense to assign the burden of proof and/or the obligation to play first to either “the player who thinks their stones are alive” or “the player who thinks the opponent’s stones are dead”.
In the end, both players are simply claiming an area on the board as theirs, there’s no qualitative difference and no reason to give two consecutive moves to a player whether they want it or not.
As I said before, I recognize the use for rulesets in situations where games “need” to have an official result, such as tournaments and online play, and “playing a game through” is understandably considered a waste of time – but I will never take that to mean that rulesets have any authority whatsoever on what “playing in the spirit of the game” means.
Anyway, I want to make it clear that I appreciate that you’re taking the time to discuss this with an inexperienced player, even though you haven’t convinced me yet
As a sort of compromise to resolve group status (alive or dead) disagreements while scoring a game under territory scoring rules, you could use pass stones to preserve the score during the “playing it out” stage for establishing group status. Both players should play the same number of moves in the process (including passes).
This is not strictly following official Japanese rules, but it does have the (pragmatic) benefit that players don’t need some arbiter to settle their disagreement.
That is exactly what AGA/British/French go rules do.
The French rules say: “The game is scored using area-scoring; except if both players agree to score using territory-scoring.”
I think players are supposed to give pass stones no matter whether the game is going to be scores with territory or area scoring, although of course if using area-scoring, then the pass stones don’t affect the score.
Thanks to the pass stones, both methods arrive at the same score, and if players disagree on the status of a group, they just play it out. Not only is a referee not needed to decide the status of groups but even if you do call a referee during a tournament, the referee should tell you to just play it out.
Also, the pass stone is really cool because it’s a clear way to signal passing, without the need for the players to say “I pass”. This is especially cool in international tournaments where the players don’t speak the same language.
3 Likes
I’m actually not a fan of pass stones. I don’t care much for this feature of AGA/British/French rules to make the territory scoring method arrive at the same result as area scoring.
I’m merely suggesting to use pass stones only in situations where a group status disagreement arises while scoring the game under territory scoring rules, and there is noone around to help the (inexperienced?) players settle this status disagreement without affecting the score.
I’m not aware of “pass” being potentially misunderstood in international tournaments (especially when accompanied by pressing the clock). Which languages use a different word than “pass” when finishing a game of go?
All languages except English, I guess. And in games using the pass-stone rule, you’re not expected to say anything at all.
It’s just quite elegant that passing is an actual physical action, rather than something entirely intangible / virtual where you have to tell your opponent what you did.
As far as I was aware, “pass” (perhaps spelled with 1 s) is used by all countries in the West and it is also understood (if not used) in Japan. I don’t consider “pass” an English word. It is also used (with 1 s) in other games in Dutch.
I don’t know which word Korean and Chinese players use, if not “pass”. Perhaps a nod or a bow instead of a word?
What did French players do or say before they adopted the AGA rules?
Edit: looking up the origin of the Dutch word “pas”, it is a loanword from English (probably from the game of Bridge and usage in other similar card games).
In this informal Japanese description of the Japanese rules they use the term パス (“pasu”) in katakana, suggesting it is a loanword, probably with a similar origin. But the official Japanese rules rather seem to describe passing as “renouncing a move”(?) according to this page, without a description of how to signal a pass.
2 Likes
Not exactly, since AGA also have white pass last, which switches to area scoring completely (both counting methods give the area score, no way to get the territory score).
What gennan mentioned is using pass stones in encores while retaining full territory scoring resolution. The AGA route brings all usual area scoring problems (wrt. ko, score accuracy, random bonus to B etc.).
2 Likes
As a mid-DDK, may I ask: is this a seki?
If Black plays G9, White captures J9 and when Black recaptures H9, the position repeats (not sure if ko rule prevents this).
If Black plays J6, White captures J9 and when Black recaptures J8, White can make an eye with J7 and his group is safe (in semeai, one eye beats no eye).
If Black plays J9, White passes and if Black plays G9 or J6, White captures the Black group and his top-right group can easily make two eyes.
4 Likes
Yes it is.
2 Likes
Under various area scoring rules (such as Chinese, AGA, etc.), the superko rule prevents this repetition.
Under Japanese rules, this cycle can happen repeatedly, but the captures are unbalanced, with two Black stones captured during each cycle, while only one White stone is captured. Thus, it is not in Black’s interest to persist. If Black did do so, then White would also cycle until enough points had been accumulated in White’s favor that White would still win the game, even if White abandoned this cycle. This pattern is called “sending two, returning one” and probably would not actually be played out in practice, among players that understood it, since it is essentially time wasting.
Note that this is essentially a three move cycle, where the fourth play by White must be a pass, to avoid White from eventually losing points elsewhere.
This is an example of a situation where @espoojaram’s model of “no passing allowed” breaks down for life and death dispute resolution.
2 Likes