What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 134 Stack Exchange communities.

Almost every time a colleague of mine asks a question, we have discussed this problem beforehand. Also, we sometimes meta-discuss the question for stackoverflow (why did I get the "this might be subjective" warning, is this understandable enough, etc). This is less so with answers, but I could believe for some groups this works the same. For now I'll stick to questions, but you could s/question/answer if you like.

The resulting question is almost always

  • Interesting for me, and a problem I don't know the answer to.
    • I probably would favorite it
  • According to me a valid and good question (as we talked about it beforehand)
    • I would probably want to upvote it.

Now I believe to read that "ring voting" is a thing, and I tried to find some questions about this, but didn't find many. I did find and read this blogpost from 2008: http://blog.stackoverflow.com/2008/12/vote-fraud-and-you/ .

It reads to me that upvoting 'fraud' is more commonplace then downvoting and it might seem anomylous that I would upvote almost all questions from one user. If all colleagues would do this to their team, you'd get some mediocre questions getting a lot of votes, and some users getting a lot of attention, points etc from that. If you would -as seems to be done -, sort the amount of upvotes for my colleague, you'd get me being 'anomalous' in front of the crowd.

On the other hand, as explained above, I do want to upvote almost all stuff some of my colleagues put on here for (I believe) the right reasons: they are questions I consider good and interesting. It's just that I know of all their questions (as they tell me about them), which obviously isn't the case for a random user.

So what is the deal?

  • When is upvoting content from one user frowned upon?
  • Would my usecase be in danger of being a (what I consider 'false') positive for the fraud-detection? (and would you agree with that or not)
  • How to prevent this false positive (if you agree with me it is one, otherwise this is a moot question) -> As noted in the comments: I'm not looking for tricks, so an answer could be "don't vote for your colleagues". I agree that there should be no to 'tricks' around that could be used for real fraud of course
share|improve this question
7  
Knowing how to prevent this false positive would allow people with actual voting rings to bypass the fraud scripts... –  Lix Jul 3 '12 at 12:58
3  
How's that for being meta: my colleague just upvoted this question. –  Nanne Jul 3 '12 at 12:59
    
@lix I agree, this is not the goal of the question. Clarified a bit. –  Nanne Jul 3 '12 at 13:02

5 Answers 5

up vote 28 down vote accepted

When is upvoting content from one user frowned upon?

When the purpose of your account's existence is to upvote the other person.

Would my usecase be in danger of being a (what I consider 'false') positive for the fraud-detection?

Yes, it would be in some danger. No, I don't consider it a false positive, but you wouldn't be treated as harshly as someone who created a fake account just to upvote their main account.

How to prevent this false positive?

There are lots of things you can do, including contributing content other than votes, but the votes you cast weigh most heavily in moderator's decisions in these cases. Without getting into specifics, please put yourself in the shoes of a moderator for a moment. If 90% of your votes went to your colleague, that would certainly look bad, right? What about 50% or even 25%? We're starting to reach a gray area, but we honestly don't have a line drawn anywhere specific.

The best advice is to avoid voting for your colleague's posts. Since we can't reasonably expect you to refrain completely from voting on topics where you probably have some expertise, I would caution you to vote for their posts sparingly, and make sure you spend time on the site contributing content and voting for questions and answers posted by people other than your co-workers.

Sorry I can't be more specific than that, but I'm sure you understand that if we reveal exactly how we detect vote rings and exactly what our tolerance for them is people would just use that information to conceal cheating.

share|improve this answer
1  
Why I consider this false positive is that if I would 'organically' stumble on these questions I would have upvoted them. But this is more personal taste then fact, so I think discussion on this point would be pointless. The purpose of my account obviously isn't upvoting this person. I could link you to the last question that triggered this but I didn't in the first place to not come across as "sneakily trying to whore a question around" :D. I'm sure that a human looking at my voting pattern would decide it to be nothing to worry about. Anyway, thanks for the clear answer. –  Nanne Jul 3 '12 at 13:16
1  
@Nanne I was referring mostly to the general case of co-workers upvoting each other in high proportions. How you discover questions and answers ("organically" vs. going to your co-worker's profile) isn't a huge factor. A lot of users have developed "fans" who follow them, and if you looked at my own voting record there would be slight anomalies just because I happen to share common interests with other users. We expect to see some anomalies like that for most active users. –  Bill the Lizard Jul 3 '12 at 13:29

It's unlikely that the team will reveal their cross-vote detection algorithms here. They wouldn't be a good deterrent if there were clear rules on how to stay below their radar.

I would recommend abstaining from upvoting colleagues' or friends' contributions, just like you don't participate in competitions run by your own company, or add a disclaimer when writing about something that you're affiliated with. It's just good style, even if there is no foul play involved.

If your friends' contributions are any good, they will gather upvotes on their own.

share|improve this answer
    
I still hope for another sollution, but I'm afraid this might be it. It would be too bad though, because they are generally questions I really would like to see an answer to... –  Nanne Jul 3 '12 at 13:03
5  
I do strongly disagree with the competition comparison. That would also cast a shadow on answering my colleagues' or my own question, but that is explicitly allowed. They are somewhat the same as they are both to prevent real as well as imagined fraud, but it compares 'reputation' too much to something like a game, and not a representation of the current feeling users have about someone (hmm, that didn't come out too good, but I hope you get my meaning) –  Nanne Jul 3 '12 at 13:07
2  
@Nanne not taking part in your company's competition (even if it would be perfectly squeaky clean to do so, with exactly the same chances as everyone else) is just to avoid even a shadow of doubt. The same should apply to friend votes IMO. And it's just one vote - who cares? If the contribution needs that vote, it's probably not great anyway. Answering is a different matter - except for the check mark, it's still the community that gets to judge whether your contribution is any good. Also, friends systematically answering friends' answers is a pattern that mods also look for AFAIK –  Pëkka Jul 3 '12 at 13:11
    
Yes, I do agree on the shadow-of-doubt thing, it's just the "game" part of a competition that kinda tickles me. Anyway, I see what you mean and agree that this is one side of the story (a heavy side). But in my defence, the other side also has some validity, it being that they are questions I would've voted up without a shadow of doubt if I'd just come agress them naturally. That's where it "stings". But maybe refraining from voting would be good as you say. –  Nanne Jul 3 '12 at 13:18

It's likely that your colleagues are not posting several questions per day. If you up vote their questions in moderation, then you shouldn't have any problems. A few up votes in a day to the same user will not trigger the fraudulent up voting trigger. However, if you wait until the end of the week and up vote 30 of their questions, then that would likely set off the trigger.

share|improve this answer
7  
You're only talking about the serial voting script. Independent from that the moderators have additional tools for detecting sock puppets that are not public. –  Mad Scientist Jul 3 '12 at 13:13
    
True, but if the moderators detected something wouldn't they inform the user. Then the user would be able to explain why the votes are legit and not fraudulent. –  Linger Jul 3 '12 at 13:15
    
Also, there are many users that consistently up vote Skeet, do they get trigger by moderators –  Linger Jul 3 '12 at 13:16
    
@linger i think if a user suddenly upvotes 5 JS posts, yes. Otherwise, no, I think. –  Manishearth Jul 3 '12 at 13:17

Normally, I say that your votes shouldn't be influenced by the user behind it. But this case is special.

Unfortunately, telling you exactly how to vote in a way to avoid being false-positive'd would also tell vote ringers how to avoid detection. Not that I know of a way myself :P

What I would say is this:

  • Keep significantly higher standards for stuff posted by colleagues. Of course, I'm not saying you should downvote their mediocre questions--I'm saying you shouldn't upvote unless it is absolutely-unicorn-waffletastically-amazing.

  • Vote rarely on these posts. I don't know the algorithm, but once every two weeks seems fine. Don't do this systematically--don't keep a "hitlist" and upvote them one by one with an interval of two weeks. Just that, if the next post by your colleague is fantastic but is within a two week period after your last upvote (for a question by any of your colleagues), then don't do it.

  • Keep such votes in a small percentage of your total votes.

But really, if the posts are good they can get enough upvotes on their own :)

share|improve this answer

Personally I vote for colleagues posts like any other, if I see it and it's a problem that I like or solution that helps me I upvote it. The only difference is if we are pairing or they have either asked me offline about the problem or told me they have put it on stack overflow I'm more likely to see it.
I think if you just treat it as a normal post you are going to be ok, it's not as if you are deliberately setting out to get them more points, your just happen to have a shared interest in problems.
Quite often with colleagues you prevent them even posting to stack overflow because you have shared the problem offline and solved. Even in that case, I sometimes get my team to put up the question & solution so other people will benefit (or we could find an even better solution). I would assume the abuse detection is smart enough to take account of this.

share|improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .