I’d just like an opinion on whether my reading of a position is accurate, please?
In the final position of this game:
At first I thought the black group at A9 etc. was dead, but the automatic scoring disagreed. As I look at it more though, I now think the key play here is G15, which either side could have played since clearly we both passed without playing it.
I think if white plays here then it is seki, but if black plays here then the whole white group connected to G19 can be killed.
Interesting position. I believe you are correct, and white had a couple opportunities to turn this position into a seki. Black G19 removes white’s potential eye and kills. For the AI analysis, this doesn’t make much difference as black is far ahead anyway.
Note that the game is played with area scoring, so each occupied intersection is worth one point.
Note that automatic scoring has a hard time if territories and group status are not very clear.
In playing bots, this means you really need to resolve a situation like that one through play before passing.
Otherwise you are left to the whim of the automatic scorer (because bot games are decided by the automatic scorer, because programming bots to do scoring phase is not a thing, for whatever reason).