Any chance we might start displaying a user's Rep to Post ratio on profile pages and/or wherever rep is shown? While overall reputation should remain the "one true metric on the system" it would be interesting to see how much rep, on average, a particular use earns whenever they enter a question or answer into the system.

I was curious enough to whip up a fragile bookmarklet that should work on profile pages to display a user's ratio right after their rep

<a href="javascript:function n(i) {i = $.trim(i);return parseInt(i.replace(new RegExp('[^0-9]'),''),10);};var r = n($('td.summaryinfo div.summarycount').text());var q = n($('div#questions-table .summarycount').text());var a = n($('div#answers-table .summarycount').text());var ra = r / (q+a);void($('td.summaryinfo div.summarycount').html($('td.summaryinfo div.summarycount').html() + '/' + parseInt(ra,10)));">Get SO Ratio</a>
  • 4
    A better ratio would be upvotes to posts ratio. That is a little harder to calculate, but a more accurate metric. – jjnguy Jul 13 '09 at 23:25

This metric (rep/post) has just as little meaning as any other perceived reputation statistic. People that outperform in less popular tags are far less likely to achieve the ratios of more popular tags.

That being said, I have both:

  1. calculated this for myself against a couple of other users, just for kicks, and
  2. seen users go through and methodically delete all their zero-vote posts, presumably to increase this ratio.

The old saying "Garbage in, garbage out" really applies to this calculation ;)

  • Fair points, although all reputation statistics, including straight rep, could be said to suffer from the same lack of meaning :) – Alan Storm Jul 13 '09 at 23:38
  • 2
    point #2 is really dangerous, as some of the zero vote answers do have value, if not votes – Jeff Atwood Jul 26 '09 at 11:30
  • Agreed. I can point you to at least one userID that I remember, but you can probably just do a query to see how widespread this is. – womp Jul 27 '09 at 5:28

I'd just be happy to see the number of answer and questions (or a ratio) next to the rep, e.g.

2k 1 10 (50/10) or 2k 1 10 (5)
2,000 rep, 1 silver, 10 bronze (50 answers/5 question).

Now someone who abuses the site, their rep would looks something like this:

2k 1 10 (0/300) or 2k 1 10 (0)
2,000 rep, 1 silver badge, 10 bronze (0 answers/300 questions).

I know there are users out there who asks just as many questions as answers, so maybe the ratio isn't the best idea, but having the actual number of questions and answers show with the rep would be very informative.

  • I was actually thinking about this just the other day - seeing that a user with multi-K rep has such based on X hundred questions would be fairly illuminating. The problem I have with it is screen clutter. – bananakata Sep 22 '09 at 19:38

Posts that happen near or after the daily limit is reached would count as a negative to this metric. Now, not only would you not get rep. you would dilute your rep-to-post ratio as well.

  • 1
    See jinguy's comment on the question - the better ratio is votes per post as these aren't affected by the daily cap. – ChrisF Jul 26 '09 at 11:24
  • Yes and no. They normalize out score to the good, but also to the "bad". The votes per post ratio counts downvotes 5x more heavily than they are generally counted (for now). – JP Alioto Jul 26 '09 at 17:34
  • @JP: you make it sound like its a bad thing. – perbert Sep 22 '09 at 16:51

I think this is an excellent idea. All the counter advice above is based on fringe problems that can be solved.

It's like back at high school, the final note was the only thing that counted but what would have been a lot more interesting to see what is similar to "Power" in Physics: Power = Work / Time.

So if someone made a very good mark but had to put a lot of effort into it, it would be less valuable that someone who had a less good mark but took a lot less time.

By this measure as of today, Marc Gravell would "lead" SO over Jon Skeet by 19 to 17 :-)

Therefore I think this measure is even more important than the reputation, because it is not absolute but relative.

Only relative measures allow to compare...

However I think it's at least fun to check so I installed the bookmarklet from Alan.

  • 2
    "It's like back at high school," and this is a good thing? No. No it is not. – Jeff Atwood Jul 26 '09 at 11:31
  • Not in that sense, agree :-) – raoulsson Jul 26 '09 at 11:58

You must log in to answer this question.

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