That’s not what I asked, though. I asked how many. It could be hundreds. If only tens are discovered, there is a huge hole of games that were very probably wrongly annulled.
My impression is game results are largely “fixed” so that the algorithm isn’t upset. Algorithm expects certain player to win. Games are annulled when the results don’t agree with what is expected.
This is my impression, nothing that has been said about the subject across the forum in all relevant discussions counters that impression, other than vague “mods know, mods are volunteers, mods do their best, shut up”.
Every time I read a thread related to moderation and to what moves are okay or not in games played on ogs, I keep telling myself that the reality is probably not as bad as the way it looks when reading messages posted by moderators on these forums.
Well, this thread has finally convinced me that it is. The OP was wrongly accused of sandbagging, but even after the mistake has been recognised, the OP is still asked to change the way they play go?
The reason I’m on a go server is to play go, not to play the guessing game of moderators’ idea of Go.
I will finish the correspondence games that I have started, and then I will stop playing on OGS.
You have stated such suspicions many times over the years, but I haven’t seen it happening, also not when I was a moderator (or if it happened, it was a rare occasion). This probably won’t convince you, because you just want to believe otherwise it seems.
Well, it’s difficult to prove that something doesn’t exist (Russell’s Teapot).
Naming and shaming may be uncomfortable, but I do not think it is better than the alternative of silently cancelling games. If you really want to avoid pointing directly to the game you could always (automatically, see below) batch the notifications and say “One or more of your games in the last X hours were cancelled”.
Surely it could be automated? I do not know if you specify a reason for annulment, but if one player was held to be at fault, I think that players should be notified of that.
on a sidenote here:
I find it strange to try to go on hard against sandbagging in this direction, but don’t do anything about the system that allows sandbagging really well in the first place.
Simply creating a new account within a mouseclick, not even a mail adress needed - you even start at 12k?. One lost game, and you are ready to go.
Or is it the algorithm only that matters?
It has been explicitly, very clearly stated that games are annulled if AI or mod believes resigning player could have won. A player’s reality of misreading, misjudging or plain and simple being wrong about a goban situation is thrown to the trash.
You are supposedly against naming and shaming, yet you had an innocent player very publicly have to prove that his sandbagging didn’t exist.
If you don’t see how that is wrong, we disagree on a very fundamental level and it has nothing to do with what I want to believe or not.
And I see you are still evading the question of the magnitude of altered results. I don’t expect this to change and you can weave it how you want, but it’s probably because the numbers are considerable.
I don’t expect this to go anywhere, it never does.
Where did I do that? OP asked a public question in the forums after being warned for suspected sandbagging. I commented on the positions that @KAOSkonfused posted thay they previously considered suspect after OP was reported for sandbagging.
I think there is nothing that can be said or done to prove that moderators are not sifting through unreported games/game histories to annull games that have an unexpected result (which would be a lot of work). I mean, even if OGS sends you all its logs on moderator annulments for all of OGS’ existence, you could still claim that OGS purposefully deleted the records beforehand that would support your suspicion.
But surely you can fairly easily dig out and post the numbers of games played in (e.g.) the past 24 hours, 7 days, 30 days and 365 days, and how many of each those were annulled? That would help clarify the situation.
Come on, this wasn’t a generic you. This was directed at me, because how dare I be suspicious of the fact that mods openly admit altering game results at will.
Nobody even batted an eye when all the “jokes” about me “sandbagging” (a very serious accusation in these lands) was going on, although I stated I was uncomfortable with it.
But as I said, I see you’re closing ranks again.
Also as I said, fine, you win. I’m not the target user base anyway.
However, you don’t dare to let the whole community know. Because you know this practice is wrong.
Mods could maybe be a little less strict about this stuff but I don’t think there’s anything nefarious going on @Gia.
Sandbagging is a real issue. It makes sense that the mods acted on the reports in the way that they did, imo. It’s a lot of work to go through a player’s games in detail and to judge whether or not they are intentionally resigning or not and it’s not easy to always get this right, especially with someone like me, who resigns in really stupid looking situations sometimes.
The only thing I would have done differently is not use the “system” user to notify me, because it meant I had no way of entering into a dialogue and explaining myself.
I don’t even have access as a non-moderator, and perhaps only anoek could provide this data in that format.
But when I was still a moderator, I estimate that many dozens of games were annulled by moderators daily. The vast majority being for sandbagging, score cheating and botting, usually following directly from a report investigation.
And some are reported misscores, of which many are caused by a weak bot (human opponent cannot mark dead stones when playing a bot, so they are at the mercy of the bot’s potentially mistaken life/death evaluation). Besides those situations, I can’t think of some other reason right now that occured somewhat frequently.
Sorry, your post made me think you were speaking for the moderators, and thus as a moderator, and I forgot to check. I meant of course that moderators could do this, and it would be helpful.
