Below are the net vote data from the "X mi yodeya" series to date. This should be a sample of around 250 tokens of equal value and therefore a good sample to examine. Since the questions are (presumably) judged on their own merit, regardless of the quantity or quality of their answers, what explains the variation in upvotes? Is there something inherently much more compelling about 71 than 72 for example? Or 252 than 133? Or are these numbers so low (0-5) that they are accounted for mostly by randomness? Does this have implications for the rest of the voting system?

A graph of net votes per question for about 253 questions. Some have as many as five net upvotes, while others have as few as zero. The median seems to be one net upvote.

  • Do you have these data raw? – msh210 Jun 13 '11 at 3:11
  • @msh210 goo.gl/9M7Wi – WAF Jun 13 '11 at 3:22
  • 1
    @WAF, did you try correlating the votes with the number of answers? – AviD Jun 14 '11 at 1:37
  • @AviD Nope. Good idea! Coming right up. – WAF Jun 14 '11 at 1:50
  • ... or with the amount of time the question was open? – msh210 Jun 14 '11 at 3:07
  • @msh210 What do you mean by "open"? – WAF Jun 14 '11 at 11:45
  • Sorry. I meant without having had an answer accepted. – msh210 Jun 14 '11 at 15:55

I dont think that these are very indicative for the site, these are kind of "toy" questions, so you'll have mostly the same readers with occasional visitors - which explains the peaks.

On the other hand, it is possible that the additional votes reflect good answers - e.g. numbers that turn up boring gematria, are not really "compelling"... but if you learn something interesting from a certain number, then that question is worth an upvote too!
Just a possibility....

  • 1
    Re additional votes indicating good answers: That would also be a very interesting finding since it would show a more complex relationship between questions, answers, and votes than it seems at first glance. – WAF Jun 14 '11 at 1:51
  • @WAF oh I agree. I don't know if thats the case, it's just a theory... You could try comparing q'votes with answer votes (and maybe # of answers?), but thats a more complex comparison. Or, you could try reading through the higher voted q's and see if you find the answers especially compelling ;) – AviD Jun 14 '11 at 7:05

You must log in to answer this question.

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