@XanderHenderson @amWhy: And here?

@user21820 Who, me?

@user21820 Wow, that would have meant the deleted user cast roughly 265 upvotes for Jack. I trust that since the deleted user's votes were scrapped, they were either fairly new, or they, at one point, committed voter fraud. If newerish, that's an awful lot of upvotes from a newerish user!!.

@XanderHenderson Cannot help to think that such events, affecting a mod, are not a good sign for the way the site is run.

@Did In how far? Moderators don't have a switch "allow only 100% rule-abiding users to vote on my posts" any more than other users have. One can't prevent getting votes from people that also use socks. And if for whatever reason they decide to vote for a lot of your posts, you get a big drop when the user is removed. You can't infer much from such events.

@DanielFischer: Is the sock-puppet problem a significant one? If there are even a few that do things like what we just saw undone, it would be quite a big problem. And I have every reason to think there are many more than have been caught.

I understand much of what you're saying, @DanielFischer. But the particular mod affected answers not a few very low quality questions, and in doing so, leaves himself, perhaps, more vulnerable to receiving illegitimate votes? I don't know, I can't judge.

Maybe we should move this discussion to Math Meta?

@user21820 Usually, sockpuppets are more localised. Most socks vote primarily for the puppetmaster, and other users lose only little when the socks are removed. Occasionally, however, somebody uses socks not only - or not even primarily - to upvote their own posts, but to up- or downvote specific other users. I hope that's rare enough to not be a big problem, but who knows.

Typically, the users targeted by the sock votes in these cases have a very high reputation anyway, so it's not a big part of the rep that's going upon deletion.

@user21820 You mean Math Meta site or Mathe meta room? (Probably the latter - if the discussion is going to continue, that might seem a reasonable thing to do. Let's wait a bit to see what others say about it.)

@amWhy Primarily, he answers a metric fkton of questions, so there are a lot of posts by Jack one can vote on. I don't think the quality of the answered questions has much of an influence on the legitimacy of the votes.

@MartinSleziak Yeap the Math Meta room.

@user21820 Oh, my.

@Did No doubt. It is a worrisome phenomenon.

@DanielFischer I see.

@DanielFischer perhaps not, unless a user and their sock-puppet asked a sizable number of the questions that this mod answered, with user and sockpuppet upvoting each of them, plus some. Why are you surprised that people might question the appropriateness, when history has shown the mods as a team go on witch hunts to assume the worst of a situation which a much less sinister explanation?

@amWhy I'm not surprised people question the appropriateness. But I find the jumping to conclusions premature. As for the witch hunts you mention, can you give an example or two?

@DanielFischer I haven't jumped to conclusions. I know all too well the danger and the damage when folks jump to conclusions.

@amWhy I didn't mean you with that. I meant Did's "Cannot help to think that such events, affecting a mod, are not a good sign for the way the site is run.", which I find premature. Anybody can be affected by such events.

@DanielFischer Indeed I cannot "infer much from such events", since making serious inferences would require having access to information I do not have access to. Nevertheless, the size of the rep cancelled is unusual, you would surely admit, and I do not find healthy, cancellations of this magnitude. Anyway, maybe you could confirm two things: 1. All 2655 points are due to a single user, or possibly to the votes of several? 2. Is this an all-time record on the site?

@DanielFischer Where do you see some "conclusion" in my comments here?

In fact, 2.5k points is definitely not a record - somebody pointed out a bit higher loss on meta: User deletion causes 13,000 point reputation loss?

But the timing suggests that it is more-or-less the same (or related) incident.

@Did Yes, losing that amount of rep in a single day is unusual. Re 1, without database access I can't confirm. It looks like being due to the deletion of a single user, but I don't know whether the software groups such events consistently. Re 2, not even close. As Martin just linked, Brian M. Scott recently lost 13k due to somebody else's sockpuppets being deleted. And that isn't the all-time record either.

However, you're most likely correct that regular user cannot say much about the question whether the reversed votes came from the sam user.

BTW maybe moving this to another room could be reasonable, as user21820 suggested: chat.stackexchange.com/transcript/message/44081072#44081072 ...? After all, it seems quite likely that this topic might continue for a bit. But probably the decision should be made by users most active in this discussion.

@DanielFischer Thanks for these factual informations (and thanks to @MartinSleziak as well).