Yes I know that this might sound like a previous question but bear with me I think its a bit different.

  1. I come across an answer that I think is wrong.
  2. I downvote.
  3. I start writing a comment on why I downvoted. Of course at this point the brain has engaged and starts to question some possible assumptions that lead to the downvote which makes me less sure the downvote is deserved.
  4. I remove the downvote.
  5. I go away and do a bit of test/research to check my assumptions and find them correct.
  6. I return to the answer and attempt to re-apply my downvote.
  7. Fail: "Vote to old to be changed, ..."

I agree with the reasons for this limitations for upvotes but I don't agree that a removal of an existing downvote is tantantmount to an upvote.

  • Not just one. Not just two. But oh my the dupes you will be able to line your pantry with! – random Nov 27 '09 at 13:44
  • 1
    +1 - this gets my vote – ChrisF Nov 27 '09 at 13:44
  • 1
    Bad, bad, bad! Before you downvote an answer, you should first take it in consideration. The fact that you undid a downvote just shows that you weren't sure at first. Basically, you downvoted because you thought something was wrong but weren't sure. You should downvote yourself for this! :-) – Wim ten Brink Nov 27 '09 at 14:50
  • 1
    @Random: Perhaps if you actually used the non-meta sites more than this meta one your opinion would carry more weight! – AnthonyWJones Nov 27 '09 at 14:53
  • 1
    @Workshop: Your absolutely right but that isn't human nature. Tell me you've never responded to something a little too quickly, then thought about it, then realised you we're right all along. The current system is counter to human nature. – AnthonyWJones Nov 27 '09 at 14:55
  • You do realise that the voting mech works the same across all the sites right? – random Nov 27 '09 at 14:59
  • 1
    @random: I don't use other sites I use Stackoverflow. Othersites can make their own minds up about how things should work for their community. I'm only interested in how Stackoverflow works and am only really interested in the opinion of fellow activity SO users. – AnthonyWJones Nov 27 '09 at 15:06
  • You're right. More people on SF and SU need to learn how to vote on answers before they know what they're voting on. – random Nov 27 '09 at 15:17
  • Yeah, I'm not perfect too. I did downvote an answer once, then had doubts too. But I didn't undo the downvote but took my time to do some research to check if I was right. As it turned out, I was wrong, but the downvote was too old. So I edited the answer with just a comment: "Edited to undo the downvote". That worked like a charm, because you can change your vote if an answer has been edited. :-) – Wim ten Brink Nov 28 '09 at 11:16
  • @Workshop: I hadn't thought of that. Would that allow you (and others) to remove an upvote? What happens if someone rolls back your edit? I'm not sure I'd be comfortable editing someone elses answer for my own convenience. – AnthonyWJones Nov 28 '09 at 15:10

This is exactly the behavior that the restriction is designed to prevent.

  • question is posted
  • within 5 minutes: flurry of new answers come in
  • post your answer
  • tactically downvote ALL other answers on the question (except your super-cool answer of course)
  • wait 5 more minutes
  • undo your downvotes, removing all evidence that you ever cast those naughty downvotes!

If you think something is wrong, go test your assumptions, and when you're sure, come back and vote for it.

It's quite simple, really: if you are unsure about your vote then


  • 1
    The logic here is just a tad too clear. Needs more obfuscation. – random Nov 27 '09 at 15:05
  • To my mind, the only reason you can remove a vote at all is in the case of "oops, I meant to hit up but I hit down instead" – Kevin Laity Nov 27 '09 at 15:15
  • never thought anybody would do that..this will give ideas to the people. – user131638 Nov 27 '09 at 15:31
  • 1
    @Jeff: If I'm unsure I don't vote. Its amazing how many times you can be sure and go ahead and vote then in a matter of seconds question yourself, "unvote" then after some consideration realise your initial view was correct but its too late you can't re-apply the downvote. I guess I could just let the answer go un-downvoted, I've goofed up my chance and I can still comment. Still think there should be some few seconds of grace where a quick blip of "vote/unvote" is treated as "nothing at all happened". – AnthonyWJones Nov 27 '09 at 15:32

I have the opposite problem sometimes.

I'll see an answer, that looks like the right answer, and I'll upvote. Then maybe I'll try the code or investigate further and see that it's actually a wrong answer. I'd like to change to a downvote or, at the very least, just remove the upvote. Votes are locked in entirely too quickly. The idea is to prevent "strategic" voting (downvoting someone so that they are a lower vote than you, regardless of whether their answer is correct), but I don't think it's working so great.

  • I agree with you in an ideal world where there weren't 'orrible individuals trying to game this whole thing would be unnecessary. I'm actually in favor of this limited time to retract vote feature I just don't think it should be applied to the "un-vote". – AnthonyWJones Nov 27 '09 at 14:58

I've also experienced this problem multiple times.

I think there's a solution:

  1. When voting on a post the vote goes into a temporary table.
    The vote is only visible to the voter while it's in there.

  2. If the vote is changed after less than 2 minutes the vote in the temporary table is changed and the timer reset to 2 minutes.

  3. After two minutes there are two possible outcomes:

    • The vote in the temporary table is equal to the permanent table.
      -> The temporary vote gets removed.
      -> Voting is still possible. (As if no vote was cast.)

    • The vote in the temporary table is not equal to the permanent table.
      -> The temporary vote gets moved into the permanent table.
      -> Voting is now no longer possible unless the post gets edited.

Any flaws you can think of?

  • The flaw is that within those two minutes a flurry of inertia based upvotes would have taken place. You can trick the order by downvoting, unvoting, downvoting, unvoting and constantly making sure you're on the other when the 2 minutes ticks over. – random Nov 27 '09 at 14:09
  • I thought of 2 minutes before the votes are shown to the public. (Except to the person casting the vote.) That would have the same effect as people thinking 2 minutes before voting. – Georg Schölly Nov 27 '09 at 14:43
  • @random: That seems like a lot of effort that someone has to put in for a very limited gaming exercise. – AnthonyWJones Nov 27 '09 at 14:48

You're not reading the message properly.

You voted. The first time you vote, that's when the counter starts. You have a short amount of time before the next edit to lock in a vote. After that, your vote is now too old to be changed.

Doesn't matter if the final was down or up or un.
Or if the first was up or down.
You started the voting clock.
You ran out the voting clock.

Also, look over your scenario over the downvote:

  • You downvote an answer
  • Answer is now ranked below others on sort by votes
  • Other answers may potentially float up (maybe you posted one) by collecting upvotes via inertia
  • You remove the downvote

Right there, that's the tactical voting part.

  • I can read the message fine and I'm not really interested in what it says. I can understand that there is a tactical scenario here that would prevent the removal of a downvote (or upvote) after a short time. However this should not apply when the last action was an "un-vote", I can see no "tatical" advantage given by relaxing this. – AnthonyWJones Nov 27 '09 at 14:46
  • Clearly you can't read the message fine. No-one said an unvote was an upvote. – random Nov 27 '09 at 15:14

You must log in to answer this question.

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