StackOverflow reputation has turned into a handy currency for people to use in determing expertise. Some employers are using it; some people are using it as a useful proxy measure for inquisitiveness, knowledge, communication, programming ability (hmm... doesn't correlate with being able to answer questions), intelligence (double hmm...), sex appeal (now we are getting silly) and so on. But the raw reputation score isn't that useful if you want to find out how good the person is in a particular language or technology niche. The fact that someone has 5,000 reputation is a nice thing to know if you are trying to decide whether to give them a job as a C# developer - but if that's 5,000 points all from subjective and not-programming-related, it's not quite so much of a ringing endorsement.

Would it be possible then to expose reputation grouped by tags for users? Would it be desirable? Would it be possible to have in the API? This data is already available in the form of the leaderboards on the tag stats page - e.g. c# - could and should it be available on the user profile and potentially elsewhere?

The API brings out some possibilities for using this data elsewhere. Imagine php.net reborn with StackOverflow reputation information from PHP-related questions sitting next to posted code samples in the comments section (some of which is awful, old deprecated code). Same for blogs, issue trackers and the like. Imagine if you are working on a Java library - you log into your bug tracker and sort the newly-reported issues by StackOverflow reputation for the java tag. You'd get a rough idea that the issues reported are likely to be from people who don't need to be pummelled with a clue-by-four and can address them first (alternatively: if they are so smart and have all that reputation, they ought to fix it themselves and send a patch). Ought it be possible for external sites to be able to use this kind of data?

3 Answers 3


using the v1 api, simply pull a users questions or answers for a tag and sum the score.

by screen scraping you can check, for instance, https://meta.stackoverflow.com/tags/api/topusers and scrape pre-ranked data.


If you want to gain more fine-grained knowledge, you'll have to examine a sampling. Relying on a number as course as reputation is more likely to lead you down the wrong path. For example, posting many questions is relatively easy to do, steadily increases rep, and makes it look like I'm heavily active in that tag (i.e. increases the count shown).

You can do this from the profile page by following links for any of the tags. For example, [area51] on my Meta profile takes you to a search showing all of the questions or answers I posted under that tag. You can sort that by newest or highest-voted, through additional tags, or by the other advanced search options.

  • But even if you do ask questions, that's not a bad thing. My point is simply that people are using reputation for a variety of things. If it is useful in the aggregate, then surely it would also be useful to see tag-specific totals for users?
    – Tom Morris
    Jun 7, 2010 at 3:06
  • @Tom: I'm not surprised people are using rep in lots of ways, but that doesn't mean it's accurate. Once being correct isn't required, it's just a simple matter of gaming the system (which is a modified quote about program optimization, but I can't recall the source). I would find the top 5 answers someone picked out much more valuable. For example, my highest voted SO answer is .. really trivial. It has 83 votes and shows absolutely nothing other than I know how to quote Wikipedia.
    – Gnome
    Jun 7, 2010 at 5:38

It's possible for other sites (and people that want to find it out) to retrieve this data from the dump (e.g. with this query), although it will be out-of-date by up to a month.

You must log in to answer this question.

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