Perhaps Co§9.2-2 could be interpreted as such that the capturing rule is also suspended when filling dame after stopping the game, so it would have no consequence to fill dame in proper seki situations. But I’m not sure if that’s actually the case. At least in my experience in finishing OTB games under Japanese rules, dame in seki are never filled, including eyes in seki. Those dame just remain unfilled throughout the scoring process.
Pondering about it some more, I know the capturing rule is suspended when filling territories with prisoners, so it might be the case that your interpretation of filling dame is what the writers of J89 actually intended, with the caveat that the capturing rule is already suspended so filling seki dame at this stage doesn’t collapse those sekis.
If that is the case then perhaps my finishing games with eye-seki up to now was overly cautious and all dame (including seki) could in fact be filled after stopping the clock and before removing dead stones in territory. It could make sense after all.
It could or it could not. In the end it is the responsibility of the rule makers to ensure the readability of their rulesets and to avoid confusion. It is not our task to get creative and twist words around until they finally roughly mean what the authors might have intended them to mean.
IMO rules for competitive games should regularily be revisited, challenged, improved and updated. But this does not always happen for every ruleset and therefore poorly written, contradictory or unclear rulesets exist.
None of what I wrote is supposed to be an attack on Japansese Go history or Japanese Go culture.
Just saw that Korea has actually updated their ruleset. Still bad but better than before.
No, that’s just not the case. I’ve taught dozens if not hundreds of beginners with Japanese rules, and they get proficient at scoring without the problems we see endlessly on OGS of trying to score unfinished games with open borders. A good teacher to explain things in person makes so many problems disappear.
A few years ago i suggested to organize some coaching system for beginners to pass the first difficulties like closing boundaries. Of course this sounded like fantasy, too complex but i don’t see any other way as to encourage and offer appropriate tools to help beginners start on OGS.
Indeed, when some software developer pops up every so often and asks if they should build a new go server as a hobby project, I tell them no, make an interactive scoring tutorial for beginners as that’s something actually useful we don’t have which doesn’t suffer from network effect.
True, we have a lot of threads about players being dissatisfied at the outcome of the scoring phase. Many are not aware that they can and should mark stones dead or alive before they agree. Others attempt to score unfinished boards. Is there an example in the compendium besides these categories?
I admit that my expression “real issues” is a little bait-y.
The majority in the poll (@Uberdude among them) disagrees with the AI scoring option, despite it solving the life&death and the unfinished boards issues. If that is off the table, is the existing implementation not as good as it gets? It is far from “bad”.
Many are not aware that they can and should mark stones dead or alive before they agree.
Yes, because it is poorly implemented, in large part due to being mixed with an AI score estimation.
Others attempt to score unfinished boards.
“Unfinished boards” is a misleading term. The players passed and agreed the game had ended, so the board is finished and can be scored. OGS’s system being confused by this situation is due to it behaving as a score estimator, trying to guess what the result should be, instead of a scoring tool which simply scores the game as is; another serious issue.
Is there an example in the compendium besides these categories?
Don’t know if it’s in the compendium, but a common issue is also that OGS’s system may reveals weaknesses that neither player saw, allowing a player to resume the game and take advantage of this AI help.
If that is off the table, is the existing implementation not as good as it gets?
Full manual or full auto seem both better than the current system (which is consistent with the poll’s result).
Let us revisit the older posts in this thread and the Chess analogy. Let’s bring it closer to Go: passing is now a thing. Both players pass in this position–how should it be scored?
Point to some ruleset that determines the score if you will; players will be unsatisfied. There is no proper solution and there is no real issue.
The majority do not want “full auto” and I respect that. Nobody likes the feeling of having control taken away.
Does the majority really want “full manual”? Not me at least. It feels clunky having to remove every loose stone. It is not in the spirit of OGS, a modern, convenient, feature-rich server.
That leaves us with auto-score. Since the final responsibility is with the players, it doesn’t have to be perfect. Even so, the current implementation handles finished boards and even unfinished boards pretty well in most cases.
To summarize, it is not bad.
ADDENDUM: overly helpful forum software blocks me from posting without this.
Misleading beginners into thinking they don’t need to finish games, by trying to score such games with an AI continuing to play well for them, instead of giving them the correct (silly) score of that unfinished games, prompting them to learn that they need to finish games to score them.
Sorry but this makes absolutely no sense. Chess does not end by players “passing”, and Chess is not “scored”; if you wanted to introduce those mechanics in Chess you’d need to rewrite the rules and then we could discuss on this basis.
On the contrary in Go, if both players passed to end the game and there is no disagreement as to the status of groups and no request to resume, then the game is finished and can always be scored without difficulty, whatever position the Goban happens to be in. If a beginner did not close properly a border the result may be surprising, but it will be consistent and without ambiguity.
That leaves us with auto-score.
Yes if you decide to disregard other options you are left with whatever is left, but this is a conclusion that’s only valid for yourself.
One that allows you to score unfinished boards in some way? OGS can already do that.
One that allows you to score unfinished boards according to Japanese rules? I believe OGS can even do that (not entirely sure).
A “full manual” scoring system? We don’t have the data to say for sure, but I believe that this is a fringe, purist preference. If the inclination for automatic markings is only valid for myself, then what is this?
One that does not mislead beginners? It sounds like a simple wish, but is actually a really high, subjective bar to meet. Whoever adopts this as a design goal has adopted the goal of trying to please everyone. They have my sympathies.
Sorry for skipping over this; it is a valid point.
How smart is too smart for auto-score? No amount of dumbing-down will entirely prevent giving away information, but it does break the feature. Therefore the auto-score should be as smart as possible.
Both full-manual and full-auto scoring prevent AI help to players. Too bad those are the fringe options.
Then there is acceptance. Is it such a big deal? It happens only at the end, the opponent knows what you know, and it can be a learning experience. Not bad for the basically unavoidable.
I’m not sure if its a good idea to mix the topic of “general Go rules” with the topic of the “OGS scoring features”. There of course is a connection between these topics but for large parts they’re separate.
Good question, luckily we already have a thread where we currently discuss this.
This depends on the ruleset you use and apparently also on who you ask. As established earlier in this thread I would argue that what you say is not true for the Japanese as well as the Korean rules.
I’m not sure why you believe those are fringe options, or insist that all is left is the current system.
The full auto or full manual proposals are not weird fringe proposals. Both are already in use on other servers, and the discussions and polls on this forum suggest the vast majority would prefer one or the other option rather than the current system.
If anything, I’d say it’s the current system that’s fringe, as it’s used nowhere else and seems to have few supporters.
For what it’s worth, I don’t think the current system is terrible. It didn’t push me away from OGS and I could continue to use it; but I do think it’s inferior to the other options, and so do many others apparently.
(read: not Japanese)
