34

Possible Duplicate:
The current review system encourages fake reviews; some people upvote everything rather than actually fixing problems

I just saw this question pop up in the "review edits" queue. Take a look at the edit history, specifically edits 3 and 4.

Not one, but two successive reviewers have come along, taken the slightly weak English of the original, and made it worse. It actually got harder to understand with each edit. It's telling that the edits are so minor, as well: They're both barely above 15 characters, and that's if you count whitespace.

As I shook my head and reached for the "reject" button, I got a message saying that the edit had already been approved. One of the voters was Community, but two real users gave the OK. This happens all the time, and I'm hardly the first to notice:

The current review system encourages fake reviews; some people upvote everything rather than actually fixing problems

There's no disincentive to messing up an edit. You can't downvote it, and once it's approved, there's no undoing it other than a rollback or another edit. But while you're doing that, whoever made the mess is off making six more -- and getting rep and badges for it.

I don't begrudge them the rep, but I don't want SO's quality to degrade, which it will if abuse is incentivized. Toss 50 junk edits in the queue, risking nothing, and some are bound to be approved. After all, approving them is incentivized too.

IMHO the whole incentive system for reviews needs to be removed or overhauled. If we can't do that, can we either:

  1. Allow voting (with reputation consequences) for suggested edits, or
  2. Allow flagging edits for moderator or community attention?
9
  • 21
    The number of meta posts about the broken review badges is getting out of control too Nov 13 '12 at 20:45
  • 1
    Maybe we should have a meta-meta-stackoverflow for that. ;)
    – J. Steen
    Nov 13 '12 at 20:52
  • 2
    This one is about users with either poor grammar themselves, or users who are too lazy to perform real improvement. Nov 13 '12 at 20:55
  • 3
    Maybe there should be a review queue for meta posts. Toss in a gold badge for "number of meta posts" for good measure. Surely quality would go through the roof!
    – McCannot
    Nov 13 '12 at 20:55
  • How is your question different from the meta question you link in your post, other than having a few additional concrete examples?
    – Servy
    Nov 13 '12 at 21:10
  • I was going to fix it, but there'll be so much attention to it now that I'm scared
    – juan
    Nov 13 '12 at 21:15
  • @Servy - This one is specifically about suggested edits, especially those attempting to correct another post's English/grammar. Edit suggestions are also unique in that people earn reputation for them. That's more of an incentive than badges, and it's why I'm asking that they risk reputation as well. Nov 13 '12 at 21:17
  • 5
    FWIW, this was a problem well before the Review system was implemented. See meta.stackexchange.com/questions/137784/… Nov 13 '12 at 21:18
  • @MichaelMrozek - Should I just have posted this guy? assets.nydailynews.com/polopoly_fs/1.474521!/img/httpImage/… Nov 14 '12 at 5:12