I am sure that we all know the purpose of vote lock-in. It's to prevent "tactical downvoting." We have also run into the following scenario before:

(See Can we remove vote lock-in?)

You open some old question, see that you were there a few hours (or days, weeks, months, years) ago, and that you upvoted something, and you look at it now and think "what was I thinking? That's obviously wrong!" But you can't remove the upvote because "tactical downvotes"?

Well, my situation is similar, but entirely the same. I upvoted an answer, but the question itself was soon after changed such that the answer did not solve the OP's issue. I wished then to retract my upvote, but of course I ran into everyone's "favorite" feature of Stack Exchange - vote lock-in.

I understand that the current solution to this problem is making a "bogus edit," but that does not always work because:

  1. You can have to wait for the edit to be "peer reviewed."
  2. It might not get approved.

Furthermore, no one wants to be cheap and submit a bogus edit.

I really hope that someone finds a good solution to this problem as a whole. As for me, I have a proposal to resolve at least this one aspect of the problem that I am experiencing.

If you voted on an answer, the question is changed, then your vote should no longer be locked-in. This way you can retract/change your votes on answers that lost their value without needing to submit a bogus edit.

On the flip-side, what if you downvote an answer, but then the question has information added that gives the answer value? Yes, this will only work if the question itself is changed, but at least addresses a portion of the much larger vote lock-in issue. I am right?

  • 3
    If a question is changed in such a way that it invalidates existing answers, that's a problem. The question needs to be edited to put it back the way it was—or to add a note that explains how it used to be and why existing answers were legitimate at the time. – Jason Bassford Jan 6 at 18:41

I don't think the vote-lock should be removed for answers when the question is edited.

My argument is that no edit to any question should ever change your view on an answer.

There are two situations where this proposal could be applied:

  1. Users answering unclear questions, providing nothing more but guesses. Asker adds more info --> Answer seems better now
  2. Askers invalidating good answers with their edit. --> Answer seems worse now

The first problem could be solved by this feature-request but the real solution would be to make the answerers clear that they should not answer close-worthy questions but vote to close them instead.

For the second problem we need to apply that question askers should never change the question so that it invalidates prior acceptable answers.

In such a situation you should rollback the edit and comment on the question inviting the asker to ask a new question when they have a new problem. If the OP reverts the rollback, do not rollback again. This would probably start a rollback war and this is something that needs to be avoided. Instead flag for moderator attention with a custom reason and the keywords "answers invalidated by edit". A moderator will handle the problem with special tools only available to them (lock, private message).

Note: If you don't have enough reputation to edit any post, just ask in chat. There will be enough users with sufficient privileges. A suggested edit will be rejected as too drastical. You can also try to solicit consensus on the per-site meta before performing any of these actions if you are not 100% sure or if you can't chat yet.

  • You just made a very good point. However, when I posted this question, I was specifically thinking of a Stackoverflow post where the user did not change the essence of question, but he did provide some code samples that did work with the answer. However, I did not want to restrict this discussion to situations such as this. I understand that all StackExchange sites have vote lock-in, and that it is a slightly common issue for the OP to change the direction of his question. This answer does not solve the situation I had in mind, but it does solve many others. +1 – Cardinal System Jan 6 at 18:01
  • 2
    @CardinalSystem I think this answer answers your post. The problem/fault is most probably not from the answerer but from the asker. Either the original post didn't had enough information so that an incomplete answer was submitted or the question was changed in a way that invalidates the existing answer(s). Both things are not okay. When the answer was ok for the revision of the post it was added to, the post may not be invalidated. You can ask them to create a new question. [1/2] – MEE the setup wizard Jan 6 at 18:05
  • 2
    The most important thing is to roll the edit back because otherwise a valid answer is downvoted by persons unaware of the change and that would be bad. [2/2] – MEE the setup wizard Jan 6 at 18:06
  • 2
    The other way this can happen is answerers posting "answers" before they're sure what the solution really is. They should vote to close instead. But that one's a lot harder to prevent. – Josh Caswell Jan 6 at 18:09
  • @CardinalSystem can you provide a link to the post – MEE the setup wizard Jan 6 at 18:09
  • @JoshCaswell can you edit this in please, I was just about to turn off my computer. ;) – MEE the setup wizard Jan 6 at 18:11
  • Okay, I like what your saying. This leaves one more scenario which I forgot to mention (just updated my question). What about the flip-side of this whole thing? What if you downvote an answer, but then the question has more information added which promptly gives the answer value? – Cardinal System Jan 6 at 18:14
  • @CardinalSystem I can’t think of any legitimate case where this would happen. Only when someone attempts a guess at an unclear question but here would Josh’s link apply. – MEE the setup wizard Jan 6 at 18:17
  • Ah, of course. Why didn't I think of that... – Cardinal System Jan 6 at 18:19
  • 1
    I think this answer is only valid when you want to retract an upvote. If the question is being edited to be more clear and you wish to change your downvote on an answer that now looks like a good answer to the question, you shouldn't be rolling back the question. You should be retracting your downvote. – Laurel Jan 6 at 18:25
  • @Laurel please look at my edit – MEE the setup wizard Jan 6 at 18:44

You must log in to answer this question.

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