5
$\begingroup$

I saw this question on the HNQ list, and then this one a little while later.

enter image description here

enter image description here

enter image description here

Now I don't have anything against these question per se, but I find it very odd (from my spending a little over a year on Chem.SE) that the first one managed to make it to the HNQ list with only 18 views (I was the 19th viewer, I suppose), and that the second one managed to do the same with a zero net score.

Going by my experience (which isn't very much to begin with), a question must usually gather at least 40-50 views in an hour or two (in addition to having a net positive score) in order to make it to the HNQ list.

Now I'm curious, is this some sort of bug? Or is this some recent change brought about by the SO overlords Devs (I don't frequent meta.SE so I wouldn't know)?

I suspect I'm probably being a little too pedantic here :(

$\endgroup$
4
  • 4
    $\begingroup$ The formula for the HNQ list is probably more complex than the theory of everything. $\endgroup$ Nov 21, 2017 at 12:19
  • $\begingroup$ @Mart Judging from the humor, I suppose this isn't really a bug, huh? (and that this post was unnecessary)? Still trying to find my feet on Chem meta... so not yet completely sure what posts are O.K here O:) $\endgroup$ Nov 21, 2017 at 12:21
  • 1
    $\begingroup$ We are very liberal on our meta page, so go ahead and ask whatever you are concerned about relating to our main page. You can (and probably should) do a courtesy search on mother-meta. In this case you can dig through hundreds of questions about hotness and the purpose of hotness and anything related: meta.stackexchange.com/questions/tagged/hot-questions $\endgroup$ Nov 21, 2017 at 12:35
  • $\begingroup$ @Mart Duly noted! I skimmed through the first page on that list (and then I sorted it by votes and checked again), but I don't think there's a very similar question around. I saw the post on factors that (should) affect a question's "hotness"... insightful, but sheds no light on the 18 views issue I'm seeing. Thanks! o/ $\endgroup$ Nov 21, 2017 at 12:42

1 Answer 1

8
$\begingroup$

The reason is very simple and documented on mother meta: question views do not count towards the arbitrary hotness points formula that is used to determine whether a question is ‘worthy’ of making the HNQ list.

The only metrics that influence HNQ are:

  • question score
  • answer count
  • cumulative answer scores
  • question age in hours

The formula is documented on mother meta, but with MathJax enabled here, we can make it look much nicer:

$$\begin{align}\text{hotness} &= \frac{\lfloor n(\mathrm A), 10\rfloor \times s(\mathrm Q) \times 0.2 + s(\mathrm A)}{\lceil a(\mathrm Q) + 1, 6 \rceil^{1.4}}\\[1em] n&:\text{number of}\\ s&:\text{score}\\ a&:\text{age given in multiples of 1 hour}\\ \mathrm Q&:\text{question}\\ \mathrm A&:\text{answer}\end{align}$$

The numerator increases linearly with the score of the answers and a higher question score as well as a higher number of answers will also make it increase. When ten answers have been added, no additional answers affect the hotness points any more.

The denominator is a weighting fraction that increases once the question age hits $\pu{6h}$.

Further weighting factors are added. One of these is a traffic consideration factor to help smaller sites appear in the list. The other is a multiple question decourager that will penalise each additional question from the same site by a certain factor. Aside from these two mentioned in the post of mother meta, it is unknown whether additional weighting factors are added.

To conclude: The question already had three answers with a question score of 1, meaning the first expression of the numerator was $0.6$. The cumulative answer score would be added to that. Since the question had not degraded past six hours, the denominator was still at its lowest possible value. This probably gave the question just enough edge to make it onto the list.

$\endgroup$
3
  • $\begingroup$ Senpai I've edited my post to include another example. Your answer still stands, but perhaps you might want to take a look? >:) $\endgroup$ Nov 21, 2017 at 16:23
  • 1
    $\begingroup$ @paracetamol I suspect the answer gained upvotes quickly. I know I upvoted it. As the formula shows, upvotes to answers are the most effective way to bump a question into HNQ. $\endgroup$
    – Jan
    Nov 22, 2017 at 9:15
  • $\begingroup$ Wokay! Just thought I'd let you know about the edit O:) $\endgroup$ Nov 22, 2017 at 9:17

You must log in to answer this question.

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