12

A recently savaged meta post (Disable privileges if a user has been inactive) seems to have been inspired by the OP's bewilderment that so many votes were cast (and not for him) in an election for DIY.stackexchange.com... He was trying to figure out why so many people participated in such a low-activity exchange, chalking it up to long-inactive users who suddenly showed up for the election. In the course of the conversation, someone pointed out that

Users who are eligible to vote in an election receive a notification in their global inbox letting them know it's voting time

Given that eligibility to vote is 150 rep points, and that you're automatically awarded 100 points when you join a new stack, it's likely that many - if not most - eligible voters on low-activity stacks may be only occasional participants (answered or asked one or two questions)... this may weaken the voting power of regular and active community members, and may tilt elections towards candidates with more recognizably from other exchanges (SO, etc) even if they haven't been particularly active in the exchange for which they are candidates.

So 3 dependent questions:

  1. Do people feel that this is occurring (what do the actual voter stats tell us? Are elections in low-activity sites dominated by relatively inactive members?)

  2. If it is happening, is it an issue? Maybe it's good to favor candidates with broader SE presence

  3. If it is happening, and it is an issue, what's a solution? I know that any suggestion of weighting votes will send the meta crowd into a down-voting frenzy, but what about making the amount or rep required to vote in low-use exchanges higher (though this might make the voting community too small)? What about not broadcasting a specific election to eligible voters who have low-rep in that exchange AND haven't visited the site in X months? They can still vote if they show up, but they're not necessarily encouraged to do so.

20
  • 7
    Good job turning a horrible trainwreck into a constructive discussion.
    – yannis
    Aug 8, 2012 at 20:56
  • 6
    Hm, I haven't noticed this problem; the 150 rep does require a teeny bit of activity on the site itself. If the requirement was 100 rep I'd certainly say it should be made stricter.
    – Ben Brocka
    Aug 8, 2012 at 20:57
  • 7
    My immediate gut feel is that considering the low levels of participation in elections on smaller sites, the last thing we want to do is to make it harder for people to take part.
    – Adam Lear StaffMod
    Aug 8, 2012 at 20:59
  • I feel there should be some minimum amount of time that you need to have been a member before you can start voting. Sure you can hop into a site and gain 200 reputation that same day, but how well do you really know that community if you've only been a member for one day?
    – animuson StaffMod
    Aug 8, 2012 at 20:59
  • It's worth discussing what the negative consequences are, even if all of the assertions are true. Is it really a bad thing if someone who has a very positive reputation (not the number, but the support of community members) on another site has an advantage in a moderator election of a smaller SE site? I would think that if I know X from Y, that will only help me vote more intelligently, which is good.
    – Servy
    Aug 8, 2012 at 20:59
  • 3
    @YannisRizos don't mention the war
    – Ben Brocka
    Aug 8, 2012 at 20:59
  • @animuson I got the required rep to vote in the latest SU election the day (or a day before) the election started.
    – yannis
    Aug 8, 2012 at 21:01
  • On a related note, Kevin Montrose wrote a blog post a few days ago exploring community elections: kevinmontrose.com/2012/08/04/…
    – Adam Lear StaffMod
    Aug 8, 2012 at 21:01
  • @animuson - I tend to agree, though if I logged 200 rep in a day 2 years ago and haven't been back, I'm even less qualified to make good choices... some other activity indicator may be necessary if you're going to go that direction.
    – Ben D
    Aug 8, 2012 at 21:01
  • 1
    @BenD The thing is, though, that even if I have 10,000 (or any other number) rep on a site, that's not a guarantee of my (continued) participation or familiarity with the candidates. Maybe I only vote for candidates whose name starts with E. You can't proof against that. And on the flipside, someone having low rep or low activity doesn't mean they didn't look at each candidate's activity, comments, etc. to make an informed choice. I'll have to try to dig up some user stats.
    – Adam Lear StaffMod
    Aug 8, 2012 at 21:04
  • This feels slightly like early revisions of the first question posted by the OP of the "recently savaged meta post." That question also got downvoted, but not quite as heavily.
    – Pops
    Aug 8, 2012 at 21:10
  • @AnnaLear I agree to a degree (hence the second question), but I doubt that there isn't a correlation between rep and interest/involvement/stake in the exchange in question... some voters will no doubt be exceptions, but the rules that govern the election process will likely make a difference to the outcome (otherwise why not have anyone, regardless of rep or activity vote?). Regardless, I agree that rep might not be the most important indicator... and that it might not be bad to have low rep users dominating elections.
    – Ben D
    Aug 8, 2012 at 21:13
  • 7
    As a side note to that, notice that rare posting does not imply rare viewing. There are site where I am essentially a lurker. Aug 8, 2012 at 21:17
  • 5
    Association bonuses in general have outlived most of their usefulness. It worked well when there were only a handful of closed-related sites, but now you have tons of people with voting privileges on sites they know nothing about. I haven't looked at this too critically, but my impression is that this causes a strong vote skew towards questions that appeal to non-experts on smaller sites, because of the influence of 101-rep guests. I've reduced my own voting, feeling guilty. Anyway, elections are a much bigger deal and the association bonus definitely shouldn't count towards voter eligibility.
    – Jeremy
    Aug 8, 2012 at 21:21
  • 2
    @everyone - what about the suggested solution in the last question? Don't change the rules, just don't actively encourage inactive users of an exchange through the notification system? If they haven't been seen in six months (say) AND they have never been significant users, let them vote if they show up, but don't ask them to.
    – Ben D
    Aug 8, 2012 at 21:36

1 Answer 1

8

I could get behind the suggestion that Ben makes in the comments:

Don't change the rules, just don't actively encourage inactive users of an exchange through the notification system? If they haven't been seen in six months (say) AND they have never been significant users, let them vote if they show up, but don't ask them to.

It is clear that the system tracks the presence of users (profiles say "last seen"), so use that as part of a condition on the delivery of the notification.

To vote, users would need

(rep>X) && other_conditions

To be notified, users would need

((rep>Y) || ((rep>Z) && (last_seen < W))) && other_conditions

You must log in to answer this question.

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