0
\$\begingroup\$

Recently I have noticed that the order in which answers are displayed does not necessarily go from highest voted to lowest. It used to, at least for me... maybe a function of privilege change.

Is this now a standard until acceptance?

The reason I ask is that it appears that most people looking at a question only read the first one or two posted answers, those late to the party with actually better answers or at least things no one else has considered don't seem to get votes by virtue of the person looking at them doesn't bother to scroll and read through all posted answers.

I realize this is human nature and the nature of these types of venues but would it be possible to withhold the ability to vote until you actually have at least scanned through all answers. This seems to be fairer to me instead of fastest to post wins as it seems now. I might be over simplifying it or might be interpreting it incorrectly but it could be that some don't post because they realize they won't get read.

Most people want that dopamine hit for a good answer and even going into old posts where rules have changed and posting to them doesn't seem to glean any benefit.

\$\endgroup\$
3
  • 3
    \$\begingroup\$ Related on Order of Answers \$\endgroup\$
    – NotArch
    Commented Apr 25, 2018 at 20:15
  • \$\begingroup\$ @NautArch Thank you, I must have changed my order on accident. That at least explains that part. \$\endgroup\$
    – Slagmoth
    Commented Apr 25, 2018 at 20:17
  • 3
    \$\begingroup\$ related: SE's Fastest Gun in the West meta post. \$\endgroup\$
    – nitsua60
    Commented Apr 26, 2018 at 2:45

1 Answer 1

Reset to default
3
\$\begingroup\$

Welcome to democracy

The candidate(s) with the highest profile will win irrespective of their quality - this is true in all forms of democracy ever tried.

For stack exchange "highest profile" means at or near the top so all we have to do is use the best algorithm for ordering the answers. Fortunately, we know what that is - the lower bound of Wilson score confidence interval for a Bernoulli parameter.

Unfortunatly, stack exchange doesn't use this method - the default method is the really, really bad Up votes - Down votes method. They do this even though they have known it was wrong for 7 years. It isn't going to change.

\$\endgroup\$
4
  • \$\begingroup\$ The issue with the "Lower bound of Wilson score confidence interval for a Bernoulli parameter" is that upvotes are far more common here than downvotes, so you would have to adjust the algorithm accordingly. That would involve some kind of reasoning (and likely a series of studies) for why downvotes are less common (probably due to the deletion/improvement of poor answers and the active presence of quality users leaving answers in an efficient manner if I had to guess) \$\endgroup\$
    – David Coffron
    Commented Apr 26, 2018 at 11:12
  • \$\begingroup\$ I was just thinking of a site mechanic that prevented votes until you got to the end of the answers. Sort of like the EULAs that prevent you just hitting "I Agree". I suppose that may not work with the coding and format but it was a thought. \$\endgroup\$
    – Slagmoth
    Commented Apr 26, 2018 at 13:12
  • 1
    \$\begingroup\$ @DavidCoffron why down votes are less common. They cost rep. \$\endgroup\$
    – KorvinStarmast
    Commented Apr 26, 2018 at 16:18
  • \$\begingroup\$ I think the means of sorting the answers should be transparent to a person that does not have a math degree. \$\endgroup\$
    – Bloodcinder
    Commented Apr 26, 2018 at 17:45

You must log in to answer this question.

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