12

I find the user experience of having to click the same arrow (up or down) to undo a vote counter-intuitive. I would expect to click the opposite arrow to undo a vote i.e. if I have up-voted I would expect to click the down arrow to undo my vote.

Maybe it's just me, but this gets me regularly.

Would it be possible (in addition to the existing undo behaviour) to make clicking down and up arrows decrement and increment by 1 respectively when a vote has already been cast. In this way I suspect undoing votes would work more intuitively for a bunch of users.

2
  • Hurrah! Please fix this, I didn't know it was eve possible to neutralise a vote, and I have been on SX for quite some time!
    – danixd
    Commented Feb 4, 2011 at 15:34
  • @danixd I am embarrassed to admit that I had been using SO for a long time before I worked out that you had to click the same arrow again. I found what I thought was the correct behaviour just plain confusing. I am guessing that there is a silent minority of users who just don't get it and think WTF!
    – chibacity
    Commented Feb 4, 2011 at 15:47

1 Answer 1

5

I propose as a solution that clicking either arrow will "reset" the vote. This to me makes more sense anyways.

Note: "reset" is seen (by me) as different than "increment or decrement the vote counter"

8
  • 7
    Personally I think it makes sense exactly the way it is; I don't know what website has the "downvote cancels upvote" convention, but that's kind of weird. If I want to reverse a vote I'd rather not click the new arrow twice Commented Jan 19, 2011 at 4:41
  • @Michael Just the way my mind works ....
    – jcolebrand
    Commented Jan 19, 2011 at 4:42
  • If I were increasing or decreasing a number, then clicking a +1 arrow to cancel out a -1 arrow would make sense. However I am not increasing or decreasing a number. I am toggling my reaction to a post, and thus toggling the down arrow by clicking it again makes total sense (it is an option I can light up or snuff out, not a button that does a subtraction). Likewise if I click the up arrow that means I want the "I like this answer" beacon to be lit, and to allow this the downvote beacon must be quenched. Commented Jan 19, 2011 at 5:52
  • 1
    @Jonathan @Michael I am not suggesting that the existing behaviour is illogical. I can see that it works for you: great. I am suggesting that for some people it is counter-intuitive. Some things are just experienced differently by different people. I am describing a change that would work for multiple sets of users. You will still experience the feature in the existing way. This potentially broadens the church in terms of a good UX without affecting you. In my mind, that is a good result.
    – chibacity
    Commented Jan 19, 2011 at 9:22
  • @drachenstern Good point re. reset. Including detail on increment and decrement was a bit too literal on my part. But yes, I agree that clicking any of the arrows once a vote has been cast should undo the vote.
    – chibacity
    Commented Jan 19, 2011 at 9:23
  • @chibacity It does affect me; now clicking one arrow and then clicking the other arrow doesn't do what I expect; it cancels my vote instead of reversing it. Once you've voted there are two buttons, one for "cancel vote" and one for "switch vote direction" -- you're proposing we change them both to "cancel vote" Commented Feb 4, 2011 at 15:54
  • 2
    @Michael You are quite right, I had only been thinking of the "cancel" use-case - good point. I suspect though that there are many users who are not undoing their down-votes to neutral (i.e. no vote) though, because those users press the up arrow and it goes from "-1" to "+1" which is not what they want. I am one of those users who took absolutely ages to work out you had to press the same arrow twice. I know this might look dumb, but if there are a significant number of users that have the same issue it is a UX problem.
    – chibacity
    Commented Feb 4, 2011 at 15:59
  • well, you can plea your case here as well: meta.ux.stackexchange.com/questions/382/undo-vote-expectancy but the data is weak Commented Jun 14, 2011 at 11:22

You must log in to answer this question.