What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 134 Stack Exchange communities.

I have been trying to contribute now for about 2 months and find that on some tags I am playing the helpful poster with relevant experience (i.e. KML, Arduino) and on others it is cut-throat competition to get the quickest, best and most upvote-able answer down (Regex and MySQL/SQL being the best examples I have seen).

I approach answering questions in these categories completely differently, and note others do as well. It is much more game-ified and everyone is willing to take a stab at answering poorly framed questions with low initial effort by the OP in the name of grabbing some reputation (not a complaint - these categories are fun in a different way than the others).

I am curious if there is any current means of determining the relative competitiveness of various tags on the site? I can think of several interesting metrics that might describe this:

  1. Averages answers per question?
  2. Average time to answer?
  3. Number of answers provided before poor questions are closed?
  4. Average number of quick answers subsequently deleted? (I note a lot of posters post half baked placeholders while trying to improve their answer in an edit - these seem to frequently be deleted later)

I see at least some of these answers would be calculable through data.stackoverflow.com.

share|improve this question
    
A good metric would be the number of posts a certain tag gets per day. The more active a tag is and the more popular it is, the more competitive it gets... Take a look at the jQuery or PHP tags... m҉̸a̛d̢͏n͜͏͡e̡s̡s̕̕... –  Lix Feb 7 '13 at 16:59
    
answers per question is a pretty decent metric for competitiveness, and it's something easily accessible. –  Servy Feb 7 '13 at 17:02
    
@Servy Agreed - but I am wondering if popularity always reflects competitiveness? It certainly enables it in the implied larger group of users and opportunity to compete. –  Matthew Feb 7 '13 at 17:07
    
@Matthew In a very small data set it might not be a decent metric, but over the pool of questions that the language or other large tags get it's fairly applicable. Generally when there are lots of people all trying to answer a question right away you get lots of answers, when there are few answerers answering more slowly they tend to be more likely to skip already answered questions, so you get a lot more single answer per question pages. –  Servy Feb 7 '13 at 17:10
    
data.stackexchange.com –  Kevin Feb 7 '13 at 17:14
    
@иɪvэЖєvɪɴ Agreed this is the place to do the work (indicated in my post) I am more interested in discussion around anything that might already be there, and what metrics some might suggest / be interested in. –  Matthew Feb 7 '13 at 17:25
    
Answers per question is not very good metric. It could imply an imprecise question, a wide formulated one or a flame war ignitor. –  pelaillo Feb 7 '13 at 19:32
    
What about computing average time between answers and comments. Those who have larger average time between answers and however are still being commented or answered mean that the answers are being thoughtly prepared. –  pelaillo Feb 7 '13 at 19:38

You must log in to answer this question.

Browse other questions tagged .