19
$\begingroup$

From Grace Note's comment in this answer, I was surprised to find out that the auditing system that is supposed to hinder robo-approvals is not active on math.stackexchange.

I remember seeing this a long time ago, after which I thought these were already in use.

(If this has already been discussed, my google searching skills must need some work...)


For those of you who might be unfamiliar with the scheme, here's a brief description. A review audit is a task in a review queue, just like any other item. However, if the user takes the wrong action (passing a bad post or punishing a good post, for example), they are penalized, and repeated failure causes their privilege to review to be temporarily suspended. The stats on audits can also be reviewed by moderators in order to investigate repeat offenders. It is designed both to catch poor style reviewers such as robo-reviewers, as well as to educate and realign folks who do genuinely want to help but are off-kilter to the standards of the community.

$\endgroup$
  • $\begingroup$ Personally, I would like to try it out, at least for some time period. It sounds like something that would be trivial to switch on/off, but if there are pitfalls associated with that I'd like to hear about them too. $\endgroup$ – rschwieb Aug 23 '13 at 13:24
  • 3
    $\begingroup$ To address the trial period - yes, we can certainly think of the first someever period of time as a trial period and at the end, if y'all feel it was a terrible idea, we'd deactivate. Or if y'all found it a great idea, we would leave it in. $\endgroup$ – Grace Note Aug 23 '13 at 13:27
  • 1
    $\begingroup$ Yes please! (We'll see if I remain as enthusiastic :).) $\endgroup$ – Lord_Farin Aug 23 '13 at 13:35
  • $\begingroup$ @Lord_Farin Yeah, so much of this is contingent upon how well the system works. I doubt we're going to be able to get much of a gauge on the number of bad edits prevented, but it seems like it should help. $\endgroup$ – rschwieb Aug 23 '13 at 13:39
  • $\begingroup$ Dear @GraceNote Thanks for polishing the description of the system. I was only working from sketchy reports :) $\endgroup$ – rschwieb Aug 23 '13 at 13:46
  • $\begingroup$ Interesting that this wasn't active up to now. It happened to me several times that I was convinced that I had just run into a review audit. For example some change to tags which looked completely random to me. $\endgroup$ – azimut Aug 29 '13 at 17:58
  • $\begingroup$ Sorry for bumping an old post; but if there are posts that should have the newly created (review-audit) tag; the post which lead to implementing them on this site is definitely one of them. (And perhaps bumping this post is not that bad - it can make a few more users aware of this.) $\endgroup$ – Martin Sleziak Sep 9 '13 at 8:56
19
$\begingroup$

I think this would be a very good feature. It happens to me too often that rejecting an edit is already too late - it was approved by other users. Even though there were much more obvious issues to address than removing a tag or adding two dollar signs.

An example from a couple of hours ago: an edit changed "Leibnitz" to "Liebnitz" and was quickly approved.

$\endgroup$
  • 5
    $\begingroup$ Yes! I'm also strongly in favor of enabling the feature. I regularly see approved edited that clearly shouldn't have been approved. $\endgroup$ – Ayman Hourieh Aug 23 '13 at 15:40
  • 2
    $\begingroup$ @AymanHourieh and OL: another thing you can do is flag that question for moderator attention. We've already had "talks" with some users who used to regularly approve bad edits, and most of them have improved after our talks. $\endgroup$ – Willie Wong Aug 23 '13 at 19:03
  • 1
    $\begingroup$ I think that's a bad example. I don't necessarily think such edits need to wait for someone who's good at spelling to show up and approve them. However, I also certainly don't want someone approving/rejecting mathematically substantial edits that they don't understand. $\endgroup$ – dfeuer Aug 25 '13 at 2:33
7
$\begingroup$

I'm happy to announce that review queue audits have now been enabled on Mathematics Stack Exchange.

I already edited in a description of what all they are into the question body, which leaves terribly little for me to put in my answer here. So, now that y'all know what you've got now, please enjoy!

$\endgroup$
  • $\begingroup$ I'm so going to camp the review queues right now :). Thanks. $\endgroup$ – Lord_Farin Aug 30 '13 at 19:08
  • $\begingroup$ Wow, I hit one on my third review today :) Hopefully they are not that thick all the time. $\endgroup$ – rschwieb Aug 30 '13 at 19:25
  • $\begingroup$ I just ran into one. I've posted a screenshot here: meta.math.stackexchange.com/a/10820/61691 $\endgroup$ – azimut Aug 30 '13 at 22:49

You must log in to answer this question.

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