Yesterday, I was notified of an edit of one of my answers and was asked to review it. However, the review was already approved 3:1. I went through the reviewers and the only one with somewhat similar programming interests did reject the edit.

I finally did my first rollback ; problem solved. However, what I do not see quite is whether or not there is some feedback mechanism to the approvers?


I believe that feedback should be primarily of non-punitive character. After all, people do this on a voluntary basis. The first page of review (https://stackoverflow.com/review/) shows only what is here to do. It does not directly show how I scored so far. This implicitly suggests: Everything I have done so far is perfect, and now here is some more work to you.

Ideally, both the number of edits and false positives (both absolute and as a percentage - with a link to those) would be shown.

  • 1
    There is not. And that's the cause for much teeth-gashing and consternation among other edit approvers (myself included) and many poor edits getting approved. – Michael Petrotta Nov 11 '12 at 0:23
  • 2
    @MichaelPetrotta: Just an idea: Insert inappropriate edit-suggestions into the edit queue in order to find out "serial approvers". – false Nov 11 '12 at 0:31
  • Or detecting outlier approvers (those who are the only approver or rejecter for a large number of edits. Then... do something, revoke their approver privs for a week. I'm pretty sure this has been discussed elsewhere on Meta. – Michael Petrotta Nov 11 '12 at 0:37
  • @MichaelPetrotta: Given the sheer (subjective) number of cursory approvals, your method might punish the more critical reviewers easily... – false Nov 11 '12 at 0:40
  • I've occasionally done the cheap-and-cheerful notification approach, like this exchange just now. I don't like attaching comments to random posts, but I don't know how otherwise to stop a reviewer making bad edits other than yelling at my screen. Stepping away is clearly not an option... – Michael Petrotta Nov 11 '12 at 0:41
  • Yeah, I know, and that'd probably include me, false. – Michael Petrotta Nov 11 '12 at 0:42
  • @MichaelPetrotta: I happened to have "constructively" edited one of sємsєм's web client-side tags.. – false Nov 11 '12 at 0:46
  • Hm, yes. Good change, but if you think the suggested edit is poor, remember to uncheck the "suggested edit is helpful" box when improving the post. That causes the same uptick in the user's "rejected" count as when you reject an edit outright, and with enough of those rejections, the user is prevented from making any edits for a week. – Michael Petrotta Nov 11 '12 at 1:08
  • Totally agree there are way too many approval trolls/badge hunters/whatever you call them, polluting the place with false approvals. On extreme cases I contact them by posting comment on old question they asked (smallest impact) but it's not really a good way. – Shadow Nov 11 '12 at 10:19
  • 1
    You had code marked up as comments and in <pre> tags. It was a minor, and probably unnecessary edit, but it wasn't a bad one. – ben is uǝq backwards Nov 11 '12 at 12:23
  • @Ben: It is not the <pre> thing but rather the > vs. >!. I agree that <pre> looks as minor and I would not have complained about that. I use <pre> such that I can use <s> and <b> within it which is necessary for so called failure slices – false Nov 11 '12 at 12:30
  • @false Why do you want to use spoiler tags? – CodesInChaos Nov 11 '12 at 13:57
  • @CodesInChaos: To permit the reader to think out - or better see - the answer without that help. E.g. the first spoiler is on highlighting certain relevant parts - evidently OP did not understand that. The second is after sketching the kind of reasoning to find the counterexample. The reader should get that with that reasoning alone. In traditional print you can only do this by revealing an answer some pages later - which is rather cumbersome online. The spoiler tag is ideal for this. Without it, you see the solution immediately. – false Nov 11 '12 at 14:04

You must log in to answer this question.

Browse other questions tagged .