3
$\begingroup$

Related: this
So I have just run a query at SEDE, and get the following data:

  • Upvotes cast by rep 101 users: 66626
  • Upvotes cast by all users: 1777003

Hence by simple calculation, we get $$\frac {66626} {1777003} * 100\% = 3.7493\%$$ Which is quite a lot.

Note: ironically, I am also a 101 user, and has cast 5 upvotes (and posted 1 answer, in which I messed up) but I promise that I have at least some ideas about what I am voting on.

$\endgroup$
10
  • 2
    $\begingroup$ Also choosing exactly 101-rep users and at most 101-rep users makes the result slightly different. $\endgroup$ May 9, 2020 at 4:29
  • $\begingroup$ @MartinSleziak My code looks like SELECT SUM(UpVotes) as upvotes FROM Users WHERE Reputation = 101... I think this is exactly 101 rep users. $\endgroup$ May 9, 2020 at 5:33
  • 1
    $\begingroup$ Top voters among 101-users and Top voters in total. (I only counted upvotes.) Of course, if we do not want restriction on reputation, we can simply check the list of voters on the users page. One thing visible there is that there is a lot of votes help by Community user - with reputation 1. $\endgroup$ May 9, 2020 at 6:06
  • 4
    $\begingroup$ What makes you say that this is a lot? Actually, if I can see a bias in upvotes, I see it in the way posts from the first months of MO were highly upvoted (were there less restrictions to be allowed to vote?). Some statistics (e.g., number of non-cw questions or answers with $>50$ upvotes) could confirm that it's not just an impression, and to which extent it's true. $\endgroup$
    – YCor
    May 9, 2020 at 8:06
  • 7
    $\begingroup$ A lot of them are coming from the Hot Network Questions list; despite the diverse subjects, the Stack Exchange largely consists of programmers and other people working in the IT business. They generally have some (or quite a lot of) affinity with mathematics, but they won't be able to grasp most of the advanced mathematics here. (I myself am also part of that group.) 'Simple' questions and their answers tend to get more upvotes, especially from 101-rep users. $\endgroup$
    – Glorfindel
    May 9, 2020 at 8:48
  • 2
    $\begingroup$ Among the votes by "low-rep" users, 4582 are by Ali Taghavi, who is "low-rep" only because of having offered bounties worth a total of 25600 points. $\endgroup$
    – Stefan Kohl Mod
    May 10, 2020 at 15:56
  • 1
    $\begingroup$ @StefanKohl Yes, and there are also 20k votes by the community user. For these reasons, looking at users with exactly 101 reputation probably gives a better idea about votes of users who have reputation only from association bonus. $\endgroup$ May 12, 2020 at 5:21
  • $\begingroup$ @Glorfindel This suggests me it would be interesting to try a Pagerank-like voting system: upvotes from high-reputation users are worth more. $\endgroup$ May 16, 2020 at 19:57
  • 2
    $\begingroup$ @FedericoPoloni: This might tilt the system towards inbreeding, not unlike what is happening in some highly rated mathematics journals (which tend to accept papers the easier the more existing papers in the same journal they relate to) or the domination of reddit by a dozen or so highly active users. $\endgroup$ May 17, 2020 at 19:47

0

You must log in to answer this question.

Browse other questions tagged .