I tend to fill the dame all the time when I play live, but forcing the players to prolong a correspondence game for weeks because of dame to fill sounds a bit sadistic, especially if those dame are worth exactly zero points.
Note that not filling dame is somewhat common in Chinese rules too. If the remaining number of dame is odd, then the player on turn should play a dame. But if the remaining number of dame is even, then the player on turn can pass without losing points.
In fact, I suspect that the main reason dame are filled when playing face-to-face is because it reduces the probability of making a mistake during the scoring phase. Scoring on a physical board requires moving a lot of stones around, and being careful not to accidentally “create” territory. Unfilled dame are just confusing and look like they belong to someone’s territory. This issue is not present in computer games, since no stones are moved when the computer computes the score.
Also note that in face-to-face games, especially in non-tournament games, there is a widespread custom of filling the dame “after the game”. That is, players will fill all teire points during the game, then take a second to make sure there are no more teire points remaining, then say something like “do you agree that the game is over except for zero-point dame?” (or don’t say anything, but silently look at their opponent and nod in understanding); and then finally, both players will fill the dame “informally”, without looking at where their opponent has played (and maybe not even alternating play correctly if using territory-scoring). This dame-filling part is thus somewhat “after the game” and not part of the game itself.
This isn’t possible in internet games, and I understand the desire to cut a correspondence game short by not wasting time on the dame.
On some occasions I use the “conditional play” feature to gain time during dame-filling; but since you can’t predict in which order your opponent will fill the dame, this requires building an extensive conditional tree, whose size is exponential in the number of remaining dame.
On some occasions I use the “conditional play” feature to gain time during dame-filling; but
since you can’t predict in which order your opponent will fill the dame, this requires building
an extensive conditional tree, whose size is exponential in the number of remaining dame
I imagine this could fairly-often be replaced with
One player uses “conditional play” for a single line which fills all the dame, and in chat
tells the other player what the line is, so that hopefully the other player just goes along it
.
So filling dame by playing actual moves (ie. before passing) is not needed (unless there is disagreement). And in online games (without a dispute) the counting/confirmation procedures after passing are skipped/automated of course.
After further thought, I don’t believe that people using apps are the cause of the wrong score in the game I posted above, nor in this additional game:
I reported this and was told that the whole scoring transaction action took 2 seconds. The wrong winner accepted the score immediately, while the proper winner clicked on White’s group marking them dead, but the correction did not work presumably due to lag. This BTW points to a side issue, that the scoring clock sometimes times out very fast, a problem that has existed off and on for years.
If use of an app fails to trigger the autoscore then I would expect that the board would be either completely unscored or one color (the app user) would be unscored. Instead the wrongly scored color is partly unscored and partly scored. That doesn’t make sense.
Hence, I still believe these two games I posted are examples of the continuation of the 2020 autoscore bug.
Another, more personal example, convinces me that the 2020 autoscore bug is ramping up. Before both sides corrected the board (they still need to accept after corrections), the autoscore had failed to mark all the dead stones and consequently also left the adjacent territory unmarked per the rules of go. It’s as if the autoscore does not recognize any dead stones. That’s a pretty big failure of the system, and it is devastating for beginners.
Again, I’m not sure I’d describe this as an autoscore bug exactly. It’s not that autoscore is getting it wrong. It’s just that autoscore does not run if the relevant player is not showing as online. This happens in Rengo where the scoring phase I don’t think really understands which player(s) can/should do scoring or are online and also with app users, who do not appear online while using the app.
The autoscore DID run. It just did its job incompletely and with very specific and peculiar omissions on both sides that exactly match the 2020 autoscore bug.
just make it VERY CLEAR that opponents should click something on the board before accepting the score, then it wouldn’t matter how auto-score paints something by default
Very large numbers of beginners and even some experienced players do not know they can mark the board, let alone how to do it. It appears that the majority of new players do not read the documentation. Since the autoscore is available, they assume that it takes care of scoring correctly.
If one wants a hard-nosed shift to making the players utterly responsible for the scoring, then it would be better to abolish the flawed autoscore, forcing the beginners to learn how to score. If that is not done, then bugs and design flaws documented in this and the earlier thread on this subject should be fixed.
That is another example of the autoscore substituting its own knowledge, in the form of a multi-move tesuji, for the players’ knowledge. As such, it violates the site rules because White, in this case, is being assisted by AI. Similar cases were the subject of the previous thread about the autoscore flaws, and I think some appear in this thread too. In addition, I know of several other cases that I have never bothered to post.
As I have said before: until this is fixed, players must defeat not only their opponent, but the autoscore as well.
A good one to highlight the difference between score estimation and final scoring. This seems reasonable for a score estimation but if both players have passed then the left side/centre and top right should all be dame.
