Ah I’m not that good at psychology lol. Besides i don’t like to win like this.
There is still a small problem with the poll as we cannot chose 2 solutions
2 Likes
I’m surprised that the 15k is playing with white. I think there is a case to be made that the 15k gets 2 handicap.
I can understand that the 15k is unaware that black’s lower side groups are hanging, but I would expect that a 10k knows the lower side groups are unsettled. Why did the 10k let it come to this? Or is the 10k also unaware of the status of those groups?
On the poll questions, it depends:
Is this a competitive ranked game or an unranked teaching game? If this is a teaching game and the 10k knows the status of those groups, the 10k could consider explaining the situation, save one group and let white kill the other.
If it is a casual IRL game and they both pass and I’m spectating, I would probably let them score the game without any comment from me, but afterwards I might point out to both players that black’s lower side groups were unsettled.
If it is a competitive ranked game, I guess the 10k passes and neither player marks the black groups during the stone removal phase, those black groups are considered alive and black is awarded 12 territory points on the lower side.
If the 10k passes and scoring is left to an autoscore algorithm without players marking dead stones, I suppose that several different results could come out of it:
-
Annullment (game is not finished)
-
Both black groups alive
-
One black group alive and one dead
With options 2 and 3 there is the option of substracting teire points for living/killing.
I think a case can be made for either of those options. But I think option 3 with teire substraction requires a very sophisticated autoscore algorithm. And perhaps a mod/spectator would need to explain the result in option 1 and 3.
3 Likes
Hmmm… 
Very interesting case. I voted “pass” but now I’m not so sure. Let’s think through the options…
Main thing is that black’s lower groups are unsettled, i.e. life and death depends on who plays first. So, assuming no other prisoners (I think komi value doesn’t matter, if any) the variations are:
-
If black passes and asserts they are alive:
a) white may agree that they are alive and black wins; or
b) white asserts they are dead, in which case they will have to resolve the dispute by resuming.
(i) If white wants to resume, then black gets to play first, so black saves one group, white kills the other and black loses; or (ii) if black wants to resume, then white gets to play first, so white kills one group, black saves the other and black loses -
If black protects one group:
a) white successfully kills the other group and black loses; or
b) white fails at killing the other group and black wins
Question: If black can read that the groups are unsettled but tries to win by variation 1. a), i.e. asserts they are alive when they know they are actually unsettled, relying on white to agree due to their inferior strength, is this cheating?
So, maybe, black is better off going for variation 2. and hoping for white to make a mistake so that black wins by variation 2. b) rather than losing by variation 2. a). On reflection, this seems like maybe the fair option as it gives white their deserved opportunity to win but they have to prove their ability to kill black’s rectangular six. Conversely, one could view it that variation 1. relies on tricking white into believing the black groups are alive when they are not.
4 Likes
BTW, I think Japanese vs Chinese rules does’t make any difference here, except for the question of teire points substraction in the autoscore cases.
2 Likes
These are both very nice ways to deal with it 
3 Likes
I wouldn’t call it cheating, maybe I would call it an intentional overplay. If black plays a move that doesn’t work and black knows that, but hopes that white doesn’t find the correct answer, is that cheating?
If white knew how to kill the corners, white should just go ahead and kill one. What is white trying to achieve by passing? Do they hope, that black gives away how to kill the corner so white can do it? Would that be cheating?
3 Likes
I think @dragon-devourer was asking whether it would be cheating to call them alive during scoring.
I’d say for a 10k player who knows the actual status, it’s dishonest to call either group alive. Either group “can” be hypothetically killed by White, so shouldn’t they be called dead under an honest application of Japanese rules?
2 Likes
I find this example the best case so far that the players should do the scoring completely themselves.
It shows that it does not only depend on perfect play, but also strongly on the assumptions made about the skill level of the players.
8 Likes
Oh well, I guess I can’t really judge that without looking into the rules. According to the CMU link there are some things to consider:
i: These rules must be applied in a spirit of good sense and mutual trust between the players. (Preamble, was mentioned here before)
ii: Go is a game in which two players compete in skill on a board, from the beginning of the game until the game stops according to Article 9. (Article 1)
iii: When a player passes his move and his opponent passes in succession, the game stops. (Article 9)
iv: Stones are said to be “alive” if they cannot be captured by the opponent, […] ( Article 7)
So under Japanese rules the competition in skill ends when the players pass in succesion. Does that mean the outcome should not depend on skill once both players passed? Then black would have to be honest about the status of the corners.
But that raises a question: Are the players allowed to use help (from stronger players, AIs, Sensei’s Library etc.) to determine which stones are alive?
If yes, I actually think a strong bot should do the scoring.
And another question: If help is allowed and white thinks the corner might be unsettled, would white be allowed to pass in order to get the help of someone to actually find the move that kills? Or does that violate the application of the rules in a spirit of good sense?
Huh, I wouldn’t have thought scoring could get that complicated, but I find it very interesting. I’ll have to read some commentaries on the Japanese rules.
2 Likes
In japanese rules in case of a disagreement, players resume playing to fix status, but this stage is only for this for what i understand. After agreement, the games comes back to the double pass time for the counting. I don’t know how external help is welcome during this extra time, i guess not but because it’s clearly stipulated that the competition in skills ended already i dunno now. Then AIs, other players may interfere and the final decision taken by the referee (the mod here)
So adapting it to online it would be consistent to propose a semi-manual scoring, letting players mark the stones and then to use an AI in case of disagrement (option for other players and ultimate decision by mod) To cover all cases an ultimate confirm of AI analysis may be asked to the players and if still disagrement, case given to the mod.
This system will be in the spirit of japanese rule but will remove the “please continue the game” advice from the mod (not completly if a player call a mod before passing.)
For the cases of unfinished games (thinking of beginners) mod has full liberty (but a knowledge how to proceed too) to determine the result.(annulation included)
For the weakest players who still have problems with boundaries, group status… we could plan another more friendy and adaptative procedure of scoring in unrated games.
2 Likes
In theory yes. Since the opponent gets to play first neither player will request the game to continue, if the outcome depends on this move. The game just stays in scoring phase until an outside party solves it.
As if. I need the rules when there’s disagreement. If good sense and mutual trust were enough, there wouldn’t have been the need to write down the rules in the first place.
5 Likes
(Post withdrawn by author)
Re this
And this
I wouldn’t worry too much about interpreting the Japanese rules literally. As noted in the commentaries by Robert Jasiek, the official Japanese rules are full of inconsistencies to the point that they are unplayable as written. He proposes “applicable Japanese rules” that are as true to the original as possible with the inconsistencies ironed out so they are then playable. 
As for the recent debate about scoring phase, common sense says that the game must be scored by the players without external help. Only in the case of an unresolvable dispute should assistance be call for. In such a case, the “referee” (moderator) can make the decision. Preferably, the referee should be somewhat stronger than the players in order to judge the position with authority. In cases where this is not possible, it seems acceptable for the referee to seek further advice, e.g. from another stronger player and / or AI. 
4 Likes
Just because he says so doesn’t make this true. Japanese '89 rules are not that bad actually, but they are not aimed at beginners.
About resumption and disputes/scoring: the rules have both as separate options. A stopped game can be resumed by the players and normal play can continue. Or, alternatively, the rules define how the board should be scored if nobody requests resumption but the players disagree. This latter is where competition of skill is not a factor, what is alive depends on what can live theoretically. Strong bots seem to fit in here.
The tricky thing is players requesting resumption AFTER the stopped position was analyzed.
4 Likes
True 
But in this case, what he says also happens to be correct. In the commentaries, he gives examples that highlight the inconsistencies and show how that makes the rules unplayable if taken literally. From reading through the rules and commentaries, it seems to be more a problem of sloppy transcription leading to inaccurate wording and self contradiction, as opposed to an inherent flaw in the logic of the rules themselves. While that may be the case, there are some traditional aspects of Japanese rules that makes them more complicated than other rulesets. Some concise rulesets advocates really hate that but I don’t mind (nor am I much of a fan of overly concise rulesets like Tromp-Taylor anyway 
)
2 Likes
Robert looks for things that can be misunderstood (with effort). Players need rules that can be understood (with effort). The distinction you wrote above is important: J89 rules are remarkably free from gross logical flaws and incorrect rulings - unlike rules created by Robert himself.
Logical territory scoring is not easy. But many players still like it because of its beauty.
1 Like
Over the past few months, there are several lengthy debates going on L19 about the interpretation of J’89 rules: L19 Go Rules subforum
3 Likes
It’s a bit of a shame that I spend all of my time here and have largely ignored L19, as it seems that is where the bulk of the rules pedantry discussion occurs.
However, I simply enjoy these forums so much more, and the software is much better.
6 Likes
If you like philosophising and debating about rules and beasts, there is plenty of material on L19.
This beast for example (No result game without loop (in japonese rule) ? • Life In 19x19)
AFAICT, it “exploits” a literal interpretation of this:
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 suppose the 2nd clause of the 1st sentence is meant to declare that the single stone in a snap-back is alive, because capturing the stone will result in restoring that stone by a recapture (snap-back).
But that description is maybe not restrictive enough. Take the position above. If black captures those stones, it is unknown if white can live inside the newly created space. So what is the status of white’s stones (assuming that 2nd clause of the rule is to be taken 100% literally)? How would autoscoring work for this final position?
4 Likes
I think if you take a quick glance at some of the discussions there, you won’t feel like you’re missing out. It’s mostly endless fighting where the parties involved don’t agree on some fundamental points, and thus never manage to convince each other of anything.
7 Likes