Too bad you didnโt say that earlier, you could have saved yourself a post. I was only addressing the question of whether it was an autoscore mistake.
I never said it was an autoscore mistake.Too bad you didnโt read my post, you could have saved us both a post.
Yeah, I read your post. I donโt believe you really think I didnโtโbut I commend your knowledge of cliches. Your post said nothing about the marking tool. What it said wasโฆ
Strictly speaking, the server didnโt score those dame, a human did. Since your post drew attention to the serverโs scoring, that is what I addressed, and it is naturally the first question most people are going to think of. The marking tool is another issue.
The irony, as you should know if you have been paying attention for the past year or more, is that I agree with you on the issues in this thread.
1 Like
What if my opponent and I both agree that a group is seki, and weโre playing with Japanese rules? How do we mark the eyes of the seki as โnot territoryโ?
3 Likes
Chinese rules FTW 
2 Likes
Hell Yeah, Chinese rules all the way, so much simpler and so much more confusing at the same time
1 Like
Here is a new example of the autoscore bug that started in August 2020. The bug is characterized by unmarked territory and occasionally by unmarked dead stones as well as unmarked territory. This is one of the worst examples I have seen. The bug became rampant up to the time of the last autoscore update, which very greatly reduced incidents of the bug, but did not eliminate it.
I reported this game, and the mod told me that, indeed, neither player marked the board in scoring. Therefore, this is definitely a case of the old autoscore bug.
1 Like
Just shift-click on each point of territory until it turns to dame.
Iโm not sure this is an autoscore bug exactly. I would guess that this is a case where the players are using an app (at least one, Iโm not sure if it makes a difference of itโs one or both players using an app).
It seems that autoscire doesnโt automatically run it a player is โofflineโ. So itโs not so much that itโs a bug with autoscore but maybe a bug in triggering (or not) the autoscore system.
I actually prefer this as it leaves the players to mark the dead stones manually. But itโs probably not the desired behaviour, especially if the players donโt realise they need to mark the dead stones or that autoscore has not run.
3 Likes
I am not familiar with that. How long has such a use of an app been going on. Can that explain the cases dating back to August 2020? No one then or subsequently could explain the phenomenon. More importantly, what is the purpose of โusing an appโ? What does that mean exactly (in a go context) and what does it do?
Getting rather OT but the Android app has been around for a few years and there is also an iOS one too. They just allow you to pay go on OGS. I use the Android one as itโs just easier to use on mobile then the main site directly. Itโs obviously designed for mobile whereas the OGS site is a desktop thing that works for mobile but itโs clearly not designed for mobile primarily.
2 Likes
Thank you. It sounds like this could be a growing thing. If so and it causes such problems as above, it is something that urgently need to be addressed in some way. I suspect most players expect the autoscore to work under all circumstances. And we already know that huge numbers of players do not know how to correct the score.
I have no idea about numbers of app users as a fraction of the player base but Iโm not sure that auto score not working is necessarily that much of s problem. Itโs clear in the app what is dead/alive/territory and tapping live updates the view. The specific case posted on Reddit is one where the position on the site didnโt match what was shown in the app. I donโt know how that arose but I know the app developer is looking into it. However I think it would be a rare event.
The problem of people not knowing what is alive or dead or how to mark things one way or another is probably nothing to do with whether they are using an app or the site.
3 Likes
People not knowing how to mark the board, or even knowing that it can be marked, is already a big problem, leading to the large number of cases in which score cheaters succeed in their cheating.
If wrongly scored games like the one I just posted are due to the use of an app, and if that is due to the app not triggering the autoscore, then I think many more of these wrongly scored games will occur if use of an app becomes more popular.
2 Likes
I play ogs via my phone, I use Chrome, not the app. I frequently encountered delays in scoring and often had to refresh the page, or close it and open it up again. Is that related? It for sure is annoying. 
1 Like
Rather than marking the status of both chains of stones and empty regions during the scoring phase, I believe it should be possible to score by ONLY marking the status of chains of stones and not empty regions, because the correct status of regions follows as a logical consequence of the stonesโ status. In Japanese rules stones are either alive in seki or alive with territory (or dead). Alive in seki is deduced from being adjacent to dame (dame being empty regions adjacent to live stones of both colours). So eyes in seki would be not territory by virtue of the surrounding group being in seki status, the user has no need or chance to mark (or mis-mark). And then if stones are marked alive but not adjacent to any dame then their surrounded regions are territory. Again no need or chance to mark or mis-mark. And in Chinese rules the eyes in seki are territory because โthems the rulesโ, again no need or chance to mark.
1 Like
While this definition of seki might be in the Japanese rules in theory, in practice players usually do not fill dame at the end of a game played in Japanese rules, especially on the internet. Even Japanese professional players in Japanese tournaments appear to not fill dame.
So, this definition of seki would require the computer to differentiate between โdame that the players were too lazy to fillโ and โdame in a sekiโ; or to be super-strict and force the players to fill all dame to avoid losing their territory.
You can argue that filling dame is a good idea for beginners anyway, because filling dame might force my opponent to fill the โteireโ points that they might otherwise not fill.
What we certainly donโt want is a truly alive group losing its territory because the system considers it seki because of an unfilled dame somewhere on the board.
Anyway, I donโt think that this is how OGS currently handles seki?
4 Likes
As a rules pedant, I would delight in making such a change, but pragmatically speaking, and wishing to avoid a widescale revolt, I think weโre almost forced to allow people to play with this informal application of Japanese rules.
There is a very widespread mistaken belief that one does not need to fill in dame when playing with Japanese rules. When there have been surveys and discussions about ruleset preferences on these forums, that particular misconception has often been brought up as a reason for some to prefer Japanese rules. Being lazy about dame filling seems to be a widespread practice, and the server essentially supports it by even allowing players to mark teire points (points of apparent territory that must eventually be filled in response to dame that are never filled).
To clarify what I mean in the last point, often players might both pass with a situation like this left on the board:
The correct thing to do under Japanese rules is for White to first fill the dame, and Black to defend the cutting point, however, people are often take a lazy shortcut by leaving those two moves unplayed.
However, OGS allows for scoring such positions (and pretending that those moves were played), by allowing the players to mark both squares as โdameโ and the rest as territory.
The server currently allows such informal application of the Japanese rules, for the sake convenience that many might prefer, and taking that away might make some users upset.
2 Likes
A shortcut that saves two clicks at the minimal expense of two additional clicks. Totally important!
2 Likes
I agree that itโs an ugly shortcut, and I prefer not to do it myself.
However, Iโve seen many others perform this practice, particularly in the case of correspondence games, where they are too lazy to play out two moves, but not too lazy to just simply resign in order to avoid scoring altogether.
If I had it my way, I would just require strict application of the Japanese rules, by requiring all dame (besides those required for preserving seki) to be filled in order to score without losing territory. However, I think many others might object to this.
1 Like