A few remarks about mass-downvoting

17 points

expand_less expand_more

Feb 13, 2014

To whoever has for the last several days been downvoting ~10 of my old comments per day:

It is possible that your intention is to discourage me from commenting on Less Wrong.

The actual effect is the reverse. My comments still end up positive on average, and I am therefore motivated to post more of them in order to compensate for the steady karma drain you are causing.

If you are mass-downvoting other people, the effect on some of them is probably the same.

To the LW admins, if any are reading:

Look, can we really not do anything about this behaviour? It's childish and stupid, and it makes the karma system less useful (e.g., for comment-sorting), and it gives bad actors a disproportionate influence on Less Wrong. It seems like there are lots of obvious things that would go some way towards helping, many of which have been discussed in past threads about this.

Failing that, can we at least agree that it's bad behaviour and that it would be good in principle to stop it or make it more visible and/or inconvenient?

Failing that, can we at least have an official statement from an LW administrator that mass-downvoting is not considered an undesirable behaviour here? I really hope this isn't the opinion of the LW admins, but as the topic has been discussed from time to time with never any admin response I've been thinking it increasingly likely that it is. If so, let's at least be honest about it.

To anyone else reading this:

If you should happen to notice that a sizeable fraction of my comments are at -1, this is probably why. (Though of course I may just have posted a bunch of silly things. I expect it happens from time to time.)

My apologies for cluttering up Discussion with this. (But not very many apologies; this sort of mass-downvoting seems to me to be one of the more toxic phenomena on Less Wrong, and I retain some small hope that eventually something may be done about it.)

17 points

expand_less expand_more

Comments

sorted by

Highlighting new comments since Today at 10:57 AM

All comments loaded

Eliezer_Yudkowsky

17 points

expand_less expand_more

I've asked someone trusted to try to write a program to detect mass-downvoting and even check particular individuals, but we haven't been able to find anything! It's possible that the database export we're getting from the server admins is incomplete? I don't know.

17 points

expand_less expand_more

3 points

expand_less expand_more

Huh. Now that someone has been caught very much doing this, did you find out why you couldn't detect it before?

3 points

expand_less expand_more

3 points

expand_less expand_more

Many thanks for looking into this!

What have you/they been trying to do? Unsupervised detection of mass-downvoting, or exploration of particular specific cases alleged to have occurred? If the latter, do the records not show the downvotes at all, or do they show downvotes from many different individuals in each case?

(Does the database contain information about when any given downvote happened? I guess probably not, which makes diagnosis more difficult.)

For example: I've been having ~5-10 recent and older comments downvoted per day, I think all during UK night-time or early morning (i.e., roughly 4pm to 2am Pacific), most days (I think all) for about the last 4 or 5. Approximately all of the recent-ish comments I've checked appear to have been downvoted exactly once. (A few very recent ones haven't been. A couple have been downvoted more than once; I guess that they were genuinely disliked on their (de)merits.)

If someone has the time to look, it would be interesting to know: Do the records show that those comments have been downvoted, and by whom? One downvoter, or a few, or many? Any signs of sockpuppetry, if many?

One possibility (though an unappealing one, not a very likely one a priori, and one that I feel a bit paranoid even mentioning) is a sort of downvoting ring of people willing to cooperate on downvoting at a rate just slow enough to avoid suspicion. That would be bad. And also sad.

[EDITED to add: I'd offer to help with the investigating, but as an interested party I probably shouldn't.]

[EDITED again to add: at 2014-02-15 08:41 GMT, it looks as if I haven't had a pile of downvotes in the last 8-10 hours. Maybe whoever it is has got bored, or maybe they've noticed evidence of someone looking. Or, of course, maybe I'm making the whole thing up, but anyone who finds that likely and cares to check can look at my comment history and see the evidence.]

3 points

expand_less expand_more

2 points

expand_less expand_more

This is exactly the pattern for my downvoting too.

2 points

expand_less expand_more

1 points

expand_less expand_more

Can you give us some details on how the votes are stored in the server? This may be difficult/impossible to do in an offline fashion if the right sort of data isn't available.

1 points

expand_less expand_more

KnaveOfAllTrades

6 points

expand_less expand_more

I suspect that as with site modifications, those of us suggesting ways to find downvote stalkers would do best to figure out how LW works and do as much of the work as possible ourselves. So in this case, that'd probably mean downloading LW source code, figuring out the database structure, thinking of approaches to finding downvote stalkers, formalising them as database queries, then trying to get someone with database access to security check then run those queries. I suspect this because from what I gather Eliezer and those with database access (e.g. presumably Trike) tend to be busy enough or doing important enough other things that they are not willing to or it is not worth their time to do all this themselves, so we should do as much of it as possible to make things quicker for them.

Small amount of money to mouth: I did read through some of the webpages surrounding LW's source code, downloaded it, and spent a little time trying to figure out how the site and database work. But by the time I got to the point of looking at the code, I had little enough temporary motivation left and the relation of the scripts to each other and the difficulty of figuring out where to start was enough that I didn't get very far before I burned out for that night and haven't looked again since. :z

A guide to (learning) LW's code and database (even if just a few paragraphs along the lines of 'Start by looking at the main article display script, then move on to...' or commenting the scripts or something) might be higher leverage at this point with respect to improving the site than submitting small code improvements, since it might encourage several others to submit improvements. On the other hand, part of me suspects that the set of people held back just by that might actually be quite small (polarisation of would-be contributors into hardcore and indifferent with few in the middle--'if they were going to do it, they would have done it by now').

Given the distribution of coding ability here, it certainly seems ridiculous how slow stuff like this gets done, and I think it's due to trivial inconveniences, ugh fields, etc., of which figuring out the site and how to submit code etc. is possibly a large part.

Since Eliezer's response, I have slightly decreased my distribution over the level of downvote stalking, but there is still way too much evidence for me to honestly believe that there aren't any downvote stalkers; it would take at least an explanation of exactly what had been tried and possibly significant knowledge of database structure to convince me it's not happening at this point. So at present I defy the data.

6 points

expand_less expand_more

3 points

expand_less expand_more

The server needs to explicitly remember every vote from every users for the interface where anybody can change or retract any of their past votes to be possible.

3 points

expand_less expand_more

5 points

expand_less expand_more

Right- but if it doesn't have a timestamp, then it's difficult to determine whether or not one user downvoted another user many times in a few minutes, which is a more reliable sign of the karmassassination problem than just how many times one user has downvoted another user.

You could go off comment timestamp- "has user X downvoted a contiguous block of comments from user Y, or are there holes (i.e. comments user X did not downvote)?"- but that's less useful, and more likely to catch the false positives of norm-enforcing users downvoting a repeated norm-breaker.

5 points

expand_less expand_more

1 points

expand_less expand_more

The past 80+ comments from me have all had at least one downvote. There is no reasonable way to interpret this other than as having a stalker.

And the solution to how not to catch false positives is to use some common sense. You're never going to have an automated algorithm that can detect every instance of abuse, but even an instance that is not detectable by automatic means can be detectable if someone with sufficient database access takes a look when it is pointed out to them.

1 points

expand_less expand_more

1 points

expand_less expand_more

There is no reasonable way to interpret this other than as having a stalker.

Suppose we find the list of users who downvoted your recent comments, and there are fifteen users on that list, each of whom is an active poster in their own right. What conclusion would you draw from that?

(It may be that, when we actually find that list, there is one account, or a handful of mostly inactive accounts, that represent almost all of the downvotes, in which case 'stalker' is a reasonable conclusion. But it's not the only way the data could turn out.)

And the solution to how not to catch false positives is to use some common sense.

Common sense is costly. The point to doing this algorithmically is that you get a query result that says "these are the twenty cases that might be karmassassination" instead of "these are the twenty thousand cases that might be karmassassination" or "these are the zero cases that might be karmassassination."

It's also not particularly wise to run this check just on people who complain- part of the point of this is to prevent karmassassins from driving users away, which hasn't happened to the people who stuck around to complain (somewhat)- and at least a few users have a habit of downvoting any comments complaining about karma loss because they don't like comments that complain about karma loss, and so they'll be extra likely to show up on that list.

1 points

expand_less expand_more

9 points

expand_less expand_more

Time for a poll!

Systematic downvoting of users (that aren't spambots or obvious trolls) is wrong. [pollid:600]

Systematic downvoting of users (that aren't spambots or obvious trolls) is against community norms (i.e. people should already know it's wrong). [pollid:601]

It would be nice if admins had a way of automatically detecting such behavior (e.g. running an SQL query to pick up patterns of mass-downvote) [pollid:602]

By the way, dear LessWrong reader, have you ever been a victim of mass downvotes? [pollid:603]

And have you engaged in it? [pollid:604]

(not that I think such polls should have power of law, it's just nice to have an idea of the opinion of the community)

9 points

expand_less expand_more

13 points

expand_less expand_more

I've done a minor version of mass downvoting-- I normally let mildly annoying comments go, but I'll downvote them if they're from one particular poster.

13 points

expand_less expand_more

FeepingCreature

12 points

expand_less expand_more

I think if I'd notice that I always tend to downvote some person's comments when I see them, I might end up looking at their history for comments I'm likely to also disfavor. But I wouldn't downvote their comments just because they're the one who made them; I'd judge each comment individually. I don't know if this is against community norms.

12 points

expand_less expand_more

8 points

expand_less expand_more

I have two questions:

(a) Has there been a common feature of victims of mass downvoting (perhaps political views?), and

(b) Why do people care about karma so much? I don't think there is a lot of signal there.

8 points

expand_less expand_more

9 points

expand_less expand_more

(a) There is some evidence that engaging in discussions on gender and taking a view nearer the "feminist" than the "reactionary" end of the spectrum sometimes results in getting hit by mass-downvoting. I'm not sure whether other (leftish?) political opinions provoke it.

(b) I don't know how much is "so much", and hence in particular whether it's true that people care about karma "so much". But the karma system is part of how LW is built; it's meant to be somewhat motivating for writers and somewhat informative for readers. Indiscriminate downvoting (or for that matter indiscriminate upvoting, though oddly one rarely hears of people complaining that this has happened to them) of a particular user's comments is liable to distort both.

9 points

expand_less expand_more

5 points

expand_less expand_more

It's more informative for writers than readers, IMO.

Some of my most upvoted posts were ones where I was venting and expected to be dinged on. I often get downvoted for tone, so when a rant gets a large and consistently positive vote, I find that an interesting tell. I don't think I've hit on an interpretation yet, but those are my posts that I should be interpreting. I notice that I am confused, etc.

5 points

expand_less expand_more

5 points

expand_less expand_more

I don't know how much is "so much"

Enough to write a top level post about it.

5 points

expand_less expand_more

10 points

expand_less expand_more

I care about mass-downvoting because I think it is damaging to the LW community. If that counts as "caring about karma so much" then I suppose that indeed I do, but I don't really see why it should be surprising.

10 points

expand_less expand_more

2 points

expand_less expand_more

Why do people care about karma so much?

Because people are not rational, and, in particular do not like negative feedback.

2 points

expand_less expand_more

9 points

expand_less expand_more

In this community it might be more accurate to say that people do not like useless negative feedback. I'm perfectly happy to get useful negative feedback, and I think most who participate here would agree.

[ETA: A heavily downvoted post is useful information; it says people don't like what you posted, and this is something you can analyze, ignore, or change appropriately. Mass-downvoting is discouraging without providing anything useful; it says someone doesn't like what you posted because you're you, and that's kind of hard to change.

My personal opinion is that mass-downvoting is basically trolling and that complaining about it is counterproductive for the same reason that responding to trolls is counterproductive.]

9 points

expand_less expand_more

7 points

expand_less expand_more

I agree. Getting downvoted feels bad man, no matter the reason.

7 points

expand_less expand_more

5 points

expand_less expand_more

This along with the semi-regular accounts of downvote abuse makes me question what advantages a +/- system has over a strictly + system. The ego threat of being downvoted seems more like a contribution deterrent than a learning signal. Is there anyone who could explain to me why the current system is better?

5 points

expand_less expand_more

14 points

expand_less expand_more

Downvotes (plus a mechanism for making heavily-downvoted things less prominent, which LW has) provide protection against spammers and trolls.

Downvotes provide a non-cluttering way of indicating when a particular kind of comment (e.g., low-value template-y attempts at humour, which if not discouraged are apt to start taking over everything) are not appreciated by the community.

Downvotes provide a signal to people who simply aren't able or willing to make a positive contribution that they might do better to go elsewhere. Perhaps this last one is some of the motivation behind some or all mass-downvoting, but what evidence I have suggests that it's often done to people whose contribution is obviously positive to anyone who isn't so politicized that they see the mere presence of someone who thinks differently from them as a threat.

14 points

expand_less expand_more

1 points

expand_less expand_more

A less extreme modification of the karma system would be to keep the downvotes but change how karma is calculated for the users.

Karma could be defined as the sum of all votes of posts with positive total score. An alternative change would be to count only the upvotes and ignore downvotes completely for the karma calculation.

In both cases the general correlation between users that post great content and high karma would stay intact but mass downvoting would no longer feel as threatening. All the signaling benefits you mentioned would still work in this modified system.

Do you think these are acceptable changes to the karma system?

1 points

expand_less expand_more

7 points

expand_less expand_more

Check out this comment by Eliezer, within the last month.

I have always been skeptical of the value of posts like this. I have since become annoyed at the quantity of them. The fact that Eliezer responded in a particular place that was not of this form is evidence that they are not a good way of attracting his attention. Maybe you should contact him directly? (though I'd wait a full month from the above comment) If you want a statement from moderators, maybe you should contact them directly?

7 points

expand_less expand_more

3 points

expand_less expand_more

Maybe you should contact him directly?

In my experience contacting him directly has been useless and he claims that there is no way whatsoever to check a suspect's voting history.

3 points

expand_less expand_more

3 points

expand_less expand_more

His assertion that there is no way to check seems to me a better outcome than these posts shouting into the wind that don't get any response.
But now that he made the particular comment, one could contact him to ask for an update, rather than a whole request.

3 points

expand_less expand_more

5 points

expand_less expand_more

Engaging any controversial topic might earn you a fan it seems. You contributed to White Lies. There was a fair bit of disagreement in that thread and it wasn't pretty. Prismattic and Wedrifid also remarked in the previous open thread after participation in White Lies that they were mass downvoted. I got unusually many downvotes in the thread too, which I don't find surprising, but haven't experienced systematic downvoting.

Previous top level post

5 points

expand_less expand_more

4 points

expand_less expand_more

Yeah, I've been getting this too (see here for when I posted in the open thread about it).

So far, I've lost about 300 karma.

4 points

expand_less expand_more

3 points

expand_less expand_more

It is interesting to me that the karma system continues to be a discussion on LW.

Is there a correct answer to what an optimized karma system entails? How optimized is the current system? Do people really care that much?

At this point, a high comment score signals to me (A) good, useful comment and (b) someone who is super invloved in LW and LW-approved stuff like CFAR or MIRI or whatever. I'd say A is the dominant signal, but B is getting stronger all the time.

As far as mass-downvoting, I think it is problem for some users and I believe there has got to be some way to mitigate its effects. I sense a apathy from whoever makes the decisions to do anything about mass-downvoting, or the karma system writ large. Why? Who knows. I'm apathetic about their apathy.

3 points

expand_less expand_more

4 points

expand_less expand_more

I don't feel this is very much an issue with the karma system, just a problem with "mean people suck". No matter what technical system is in place, there will probably be some ways for people to be meanies---sending lots of nasty personal messages from behind a TOR proxy, for instance. The fact that in this case they decided to (ab)use the voting system is incidental, and a solution need not involve making any dramatic changes to how the voting mechanisms work.

4 points

expand_less expand_more

3 points

expand_less expand_more

What does "make my votes public" do?

I've never noticed any indication of anyone's identity associated with a vote. Does everyone have it set to private?

3 points

expand_less expand_more

5 points

expand_less expand_more

"Make my votes public" only applies to votes on top level posts, not votes on comments. Mr Hen turned it on, so his about page has, in addition to the normal "Comments" and "Submitted" tabs, tabs called "Liked" and "Disliked" (also, "Hidden," whatever that means*; some people have a "Drafts" section, but it's usually empty). Liked and Disliked show posts that he voted on. The point is for people who share his taste to get recommendations. It is not convenient for posters to find out how he voted on their posts, presumably by design.

* I'd expect Hidden to be related to the preferences choices "don't show me posts after I've [dis]liked them," but it doesn't seem to match in his case.

5 points

expand_less expand_more

3 points

expand_less expand_more

Relevant post from the stack overflow blog: Vote Fraud and You (not so much for the post itself as the comments and links).

Given that the karma system has value, does not function well when people mass-downvote, and that people are actually mass-downvoting, it's important to deal with the problem.

3 points

expand_less expand_more

2 points

expand_less expand_more

How easy would it be to have downvotes be on some kind of timer, where you could only downvote once every N minutes? (A time is arbitrary and flexible based on experimentation)

This seems as if it would prevent someone from trivially systematically going through and downvoting every post by a poster, but still allows for someone to read something and downvote it on a general basis.

2 points

expand_less expand_more

5 points

expand_less expand_more

If someone's goal is to mass-downvote all comments made by someone else, this wouldn't stop them at all. Just keep the page open in another tab, and every N minutes downvote the next comment.

For people who read quickly and vote in real time, it would be inconvenient. (For example, Stack Exchange has a similar thing: I can read a short comment in 1 second, but I am allowed to vote only once per 5 seconds. So I read 20 comments in a row and then close the page without voting, because I don't see a point in waiting.)

5 points

expand_less expand_more

3 points

expand_less expand_more

I don't know anything about LW's internals, but in most sane architectures you'd want to have a last-downvote timestamp linked to each user's account, and to compare the timing of incoming downvotes against it. If the difference is too small, you reject the input and display a message similar to the one you get when you're about to post in a heavily downvoted thread. That message has a few issues -- it doesn't come up if the vote total upthread dropped below the threshold between when the page was loaded and when the post was submitted -- but since we only care about session-local behavior here, that shouldn't be a problem. Synchronization between the session and server time might be harder but that depends on details I don't know.

I can see a behavioral pitfall, though. I typically decide on votes after reading through a screen's worth or so of a thread, and this sometimes implies two or more downvotes (for example in the cases of short low-quality posts or discussions inappropriate for the site). I'd find it gratingly aversive to be forced to wait N seconds or minutes between posts in these cases; I don't know if this happens often enough to outweigh the mass-downvoting problem, though.

3 points

expand_less expand_more

Stefan_Schubert

2 points

expand_less expand_more

The Karma system has its advantages but the present version strikes me as a bit too crude. For one thing, a down-vote from someone who down-votes everything is not as good evidence that the post is actually bad than a down-vote from someone who barely ever down-votes anything. Also, a down-vote from a poster with lots of Karma is better evidence that the post is bad than a down-vote from a poster with lots of Karma.

My suggestion is therefore that for any given level of Karma that you have, there should be definite limit to the number of up votes and down votes you can make. If you went over that limit, your votes would get diluted, so that, e.g. one up-vote from you only would give the poster .8 Karma.

If such a system would be introduced, it would also be natural to let people give less-than-full votes (say .5 up-vote, .3 down-vote, etc).

The drawback of such a system is of course that it is more complicated, but I think it would give a better picture of what posts are good and what aren't.

2 points

expand_less expand_more