Is it possible to collect statistics on how quickly questions are answered? Without collecting piles of answers back and processing them locally? In an Elastic Search instance you'd do this with the aggregations functionality.

I've read over the API, though I haven't studied all of it in detail and could be missing something obvious.


The Stack Exchange Data Explorer is an excellent resource for this type of analysis.

First, let's have a look at all questions answered within an hour:

enter image description here

The value for 0 minutes is a bit skewed because of self-answers. Otherwise, it looks like a Poisson distribution; most questions are answered within about 10 minutes.

Here is an overview for two days:

enter image description here

You see that the vast majority of the (answered) questions is answered within the first two hours after posting. (To compare: about 6.5k questions are unanswered – they don't appear in these graphs.)

  • 'merged questions can have older answers' your query comment cleared up some confusion for me. Thanks this is excellent.
    – Jontia
    Dec 4 '20 at 18:33
  • 1
    You're welcome. This is what happens if you leave that out. I'm actually surprised that I thought of this while writing the first version of the query, not afterwards. Especially considering that it's Friday evening here 🍺
    – Glorfindel
    Dec 4 '20 at 18:38

You must log in to answer this question.

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