23
$\begingroup$

Something like the following scenario has happened to me two or three times since I've been active on this site (and happened again today).

I read through an answer, like it, and upvote it. Then I take a closer look and realize I'm not so sure the answer is correct. I remove my upvote, which the system allows through (I believe) five minutes. Then, after thinking through what bothered me about the answer, I realize the answer actually is correct and click to restore my upvote. However, instead of recording my restored upvote, the system gives me this message:

You last voted on this answer
[x time units] ago
Your vote is now locked in
unless this answer is edited

The net effect is that I now cannot upvote an answer that I think deserves it.

I don't know if the same problem occurs with downvotes; I've only experienced it with upvoting.

This sure looks like a bug to me.

$\endgroup$
  • 4
    $\begingroup$ I've been wishy-washy with upvotes too sometimes, and it bothers me as well why I can't give back an upvote I took away... $\endgroup$ – J. M. is a poor mathematician Dec 19 '10 at 4:55
  • 4
    $\begingroup$ This is a duplicate of meta.stackexchange.com/questions/18360. Generally questions about the general behavior of the system are supposed to be asked over there, while site specific questions belong here. $\endgroup$ – Noah Snyder Dec 19 '10 at 4:59
  • 6
    $\begingroup$ Anyway I agree with your point, and I don't actually understand the reply given at meta.SO, but procedurally there's nothing really to be done once the question has been asked at meta.SO and the SE people have decided not to change it. $\endgroup$ – Noah Snyder Dec 19 '10 at 5:04
  • 1
    $\begingroup$ @Noah: Well, I guess that answers my question. And thanks for the heads up about this kind of question needing to be asked at meta.SO. $\endgroup$ – Mike Spivey Dec 19 '10 at 5:37
3
$\begingroup$

There are a lot of vote-based exploits based on vote undo.

Therefore,

  • voting can only be changed in the first 5 minutes after casting the vote
  • you may only toggle your vote a limited number of times within that 5 minutes, this is really only intended for "oops" type misclicks.

It's best to reserve voting until you are quite sure what your vote should be.

That said, you can always re-vote if the post has been edited since the last time you voted on it.

Edit

This has been changed - if there is no visual indicator of an up/down vote, you're allowed to cast a vote.

$\endgroup$
  • 5
    $\begingroup$ Thanks, Jeff. Would you mind giving a few examples of vote-based exploits based on vote undo? Apparently I haven't thought hard enough about how to exploit the system to figure out what those are. :) In case it isn't clear, I am asking out of curiosity, not because I want to game the system. $\endgroup$ – Mike Spivey Dec 19 '10 at 6:41
  • $\begingroup$ @Mike, @Jeff: The following is just an idea, don't know if it has a drawback. (I'd highly appreciate if I was told the drawback if it exists.) So here's the idea: If in a short time frame (say one minute) I cast a vote and withdraw it, then the system doesn't let me vote on that post for some time (say one hour), saying something like "please think through what you want before voting again" (bad wording so far). Then after this time frame, everything is forgotten, and I can decide if I want to vote up or vote down (or not vote at all). $\endgroup$ – Hendrik Vogt Dec 19 '10 at 14:08
  • 2
    $\begingroup$ @mike given sufficient undo windows, there's a way to prevent any user from gaining any reputation for an entire day, for example. $\endgroup$ – Jeff Atwood Dec 19 '10 at 14:42
  • 4
    $\begingroup$ Right, certainly unlimited undo windows are a bad idea. What I'm confused about is why having a short undo window where you can actually undo is a problem. $\endgroup$ – Noah Snyder Dec 19 '10 at 18:00
  • 2
    $\begingroup$ @Noah put on your game theory hat. What would your reaction be if you saw three downvoted answers and one at zero? Then undo, before anyone has time to complain. Answers tend to arrive very rapidly and early mover advantage tends to be a "rich get richer" $\endgroup$ – Jeff Atwood Dec 20 '10 at 0:36
  • 1
    $\begingroup$ @Jeff: Thanks for the interesting example. I'm just not sure; does the same apply to upvotes? $\endgroup$ – Hendrik Vogt Dec 20 '10 at 8:41
  • 2
    $\begingroup$ I think what's going on here is that at SO you get answers much faster than at the SE-style sites I've spent time on (here and MO). I now understand where the concern is coming from and it makes sense from the pov of SO. I think for lower volume sites it'd be better to worry less about this and just suspend people who game the system, but your probably hoping/planning that other sites will end up high volume like SO. $\endgroup$ – Noah Snyder Dec 20 '10 at 18:33
  • $\begingroup$ @Jarrod: Thank you for changing this feature. $\endgroup$ – Mike Spivey Apr 14 '11 at 19:33
  • 1
    $\begingroup$ If it's possible to game the system with "undo", then it's the system that's broken, not the undo. $\endgroup$ – endolith Sep 18 '12 at 14:45
20
$\begingroup$

Let me go on record as being against this "feature". If I only voted on answers that I was 100% sure were mathematically correct, I would be a referee instead of a participant in a math Q&A site. Even as a referee, if I realize that I have made a mistake, I have a reasonable amount of time to correct it.

This feature seems destined to reduce the amount of voting on the site, which in my view is directly antithetical to its main purpose. That some users may game the system for reputation seems like a lower order consideration to me: the reputation system is there to channel the natural desires for competition and acquisition of capital into the the socially useful outcome of creating prompt, useful and correct questions and answers.

If there is real concern about unauthorized reputation, this could be handled by the system tracking users who rapidly change many votes and moderators monitoring this data and acting accordingly.

Let me also say that the reaction on meta.stackoverflow.com to this change in site mechanics was overwhelmingly negative. It seems to be yet another discouraging instance of unresponsiveness by the founders to the needs and wishes of the site users.

$\endgroup$
  • 6
    $\begingroup$ +1. I don't vote a lot, but this encourages me very much to vote less. It goes against the very nature of the scientific enterprise to not allow me to change my mind... $\endgroup$ – Mariano Suárez-Álvarez Dec 20 '10 at 0:29
  • 1
    $\begingroup$ Moreover, reading that meta.so thread it is quite apparent that (the few, in proportion,) people who voiced an opinion were not exactly thrilled with the change. On the other hand, it appears that the impossibly ugly UI for spoilers was kept because of votes :) What do I know! $\endgroup$ – Mariano Suárez-Álvarez Dec 20 '10 at 0:32
  • 1
    $\begingroup$ Moreover, even if there is a mathematical error, it might still deserve an upvote. $\endgroup$ – Douglas S. Stones Dec 20 '10 at 6:27
  • $\begingroup$ Here's the latest request for this "feature" to be removed: Please get rid of “vote locking” $\endgroup$ – endolith Sep 18 '12 at 14:47
4
$\begingroup$

I'm going to answer this question with a question: Is it possible to edit the length of the "question window"?

My point is, if it's so unpopular, why not just change it?

$\endgroup$
  • $\begingroup$ Because the developers think they know better than all of their users. $\endgroup$ – endolith Sep 18 '12 at 14:46

You must log in to answer this question.

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