14
$\begingroup$

What percentage of your upvotes would you estimate occurs after, say, 24h have elapsed since you gave an answer?

My impression is that this percentage is very firmly in the single digits (~3-5%) in my case.

However, when I started out on CV, my understanding of this site was that it aims (surely among other things) to become a repository of answers to common questions which would be seen and appreciated by other readers than just the OP. Whether or not that is the case is not easy to say, but if it is, that does not seem to translate into upvotes that occur after a question has disappeared from the top of the stack of active questions.

$\endgroup$
  • $\begingroup$ Wonderful question. Yes, something needs to be done about it. Maybe a nice wiki, or a better ranking system for the feed? $\endgroup$ – Dawny33 Aug 27 '15 at 10:55
  • $\begingroup$ I don't know my percentages by question; but my feeling is your figures seem a bit low for me (I often get votes over several days) ... but I do see a large fraction of my daily reputation coming from votes on old posts (which may still be consistent with your percentages if one has enough old posts, I suppose). $\endgroup$ – Glen_b -Reinstate Monica Aug 27 '15 at 12:22
  • $\begingroup$ Thanks. That may also be a reflection of the difference in quality of the answers, I suppose :-). $\endgroup$ – Christoph Hanck Aug 27 '15 at 12:24
  • $\begingroup$ I think it's probably more to do with other factors. $\endgroup$ – Glen_b -Reinstate Monica Aug 27 '15 at 12:38
  • 5
    $\begingroup$ Some natural experiments are available. Eg, User 11032 was very active for a half year abruptly ending in mid-October 2012, at which point (if memory serves) he had 19K points. Since then his posts have received few downvotes or acceptances and few people have abandoned the site (which causes their votes to be eliminated). We can estimate from his current 24K reputation that the annual reputation increase from upvotes has averaged 8%. Since his mean rep/answer is much less than average, we would expect others to do better than that. $\endgroup$ – whuber♦ Aug 27 '15 at 14:12
  • 2
    $\begingroup$ For very good questions/answers there is a very long tail. I have a (very) few answers where votes are coming now and then, even if they are year old. The percentage of votes after 24h would be above 90% $\endgroup$ – kjetil b halvorsen Aug 28 '15 at 14:48
13
$\begingroup$

See this blog post I wrote for the CV site, Voting behavior and accumulation of old votes on CV (I know I need to work on my titles). As of 2012, there was non-trivial accumulation of upvotes for older questions, they are somewhat invisible to regular interaction though.

Here I have updated the query to return the aggregate counts of Vote Day - Post Day. Since the data dump only releases statistics on the day a vote was cast, a vote within 24 hours can either occur on the same day (a value of 0) or the following day (a value of 1). In total, these only account for 54% of all historical votes on the site.

Long story short, I don't see any real problems. But increasing visibility to older, high quality posts likely is not a bad thing. A good way to do this and simultaneously answer new questions is to point out duplicates.

$\endgroup$
  • 4
    $\begingroup$ I don't really agree with the premise that upvotes and being a repository of knowledge are the same thing either. Visibility (in terms of page views) and upvotes I imagine are highly correlated though. $\endgroup$ – Andy W Aug 27 '15 at 12:09

You must log in to answer this question.

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