0

I'm sure this has been proposed already, but sorting the answers by votes count has the side effect of bringing the first answers to the top (FGITW problem), with them not always being the best (this happened to some of my answers).

The default sorting has been changed already from sorting by the newest answer (in order to solve the SCITE problem) to the current sorting, but this introduced this new issue.

At first I tought that since answers can be manually sorted by votes count already, my proposal could have been to customize the sorting behavior on a per user basis and to randomize the default one, which seemed like the best option to give anyone a chance, perhaps only for logged in users (we want not logged in users who don't have the capability to cast votes to see the best answers first) and giving them the chance to change it, so that conscious users could actively choose whether or not to sort answers by votes count as it is right now, but as it has been pointed out by Seth obviously this is not good from an anwer's user perspective. I hope that someone can come up with a good solution to this.

  • 1
    Umm, the whole point of voting is to show the best answers at the top. – Seth May 31 '15 at 14:57
  • @Seth Yes, I agree with that, to be honest I posted this question mostly to bring the topic up. I agree with the fact that for the users of the answers this is not good. I hope someone can come up with a better solution – kos May 31 '15 at 15:05
  • Actually votes count won't always show the First Answer. I have several answers that were not the first answers and have far more votes than the first answer, and in some cases got me rep bounties. Where do you base your 'first post gets most votes' on? – Thomas Ward Jun 4 '15 at 12:04
  • @ThomasW. Precisely what I mean is that sometimes there's disproportion between upvotes given to first answers and upvotes given to equally good later answers; briefly searching my top answers this shows up and it's upvoted 10 times, but if that answer deserves 10 upvotes this answer in my opinion does as well – kos Jun 6 '15 at 13:16
  • I fail to see how that factors into your logic though. You say sorting by votes count brings first posts up higher, but if the later posts are better you have more votes, and an Accept beats all. I fail to see your logic on this as a result. – Thomas Ward Jun 6 '15 at 13:48
  • @ThomasW. Being more explicit (sorry for the confusion), that answer (the second one) should be the before the above awk answer which is far more complicated, but it's after because it has been posted later, and sometimes (maybe not in this case, but think of say a third answer addedd to an old question on which the first and the second answers have tons of upvotes) this is likely to stay as it is. Of course being a good first answer having been present from more it will usually always have more upvotes of a newer one because of that, [cont...] – kos Jun 6 '15 at 14:16
  • @kos add that clarification to your question ;) – Thomas Ward Jun 6 '15 at 14:17
  • @ThomasW. [cont...] , but I think that this is also due to the fact that not everyone goes through all the answers when visiting a question, and I think that sorting in a different way, for example, might help to make all of the answers being visited (roughly) the same number of time over time, mitigating what I suspect being the average user's behavior, on which of course I might be wrong. – kos Jun 6 '15 at 14:21
  • @ThomasW. I will, probably I should rephrase everything, anyway to sum up what I think overall is that not always best answers end up being at the top somehow, and that I think that sorting this way doesn't help – kos Jun 6 '15 at 14:25
  • @kos You don't propose a solution you simply state there's a problem. Hate to say it but I don't think anyone's going to put time or effort into an alternative, so you may have to propose a solution yourself. Or move it to meta.se? – Thomas Ward Jun 6 '15 at 14:39

You must log in to answer this question.

Browse other questions tagged .