What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 129 Stack Exchange communities.

As the title of this question asks, should I? There's a particular user I've come across via this suggestion that has approved 239 edit suggestions and rejected none. The pattern of their activity log is also a ton of reviews per minute, sometimes one every other second.

What can one do about that? If anything?

And, yes, I know about the ten thousand other posts about the problem of bad reviews. I wouldn't mind some guidance in this particular case, though.

share|improve this question
12  
Name and shame. Name and shame. –  TRiG Dec 4 '12 at 15:47
7  
Behaviour like this is why I never perform reviews in the queues anymore. There's just no point. –  J. Steen Dec 4 '12 at 15:53
2  
@TRiG: I guess it's time for me to publicize my hit list of robo-reviewers (context). –  BoltClock's a Unicorn Dec 4 '12 at 15:55
3  
239:0 is not a solid proof of abuse. Reviewers who just skips slippery stuff and only approves safe suggestions, could in theory legitimately have score like that. It would be better to find a proof of blatant unambiguous abuse –  gnat Dec 4 '12 at 15:57
2  
@gnat For me it's blatant, but that's because I know how many bad edits there are, and because I don't skip bad edits. Personal behaviour, so it doesn't apply to them. But this user definitely approves bad edits, and sometimes within 2 seconds of each other (the minimum time for the button to activate). –  J. Steen Dec 4 '12 at 16:00
    
@J.Steen "2 seconds" I see, that makes a more solid proof. BTW it's a shame such robots are allowed to stamp their crap so easily, there's even a feature request to stop this: Drop delay for “Skip” and increase it for “action” review buttons –  gnat Dec 4 '12 at 16:04
    
At least they're only badge whoring and not rep whoring. That gets a pass right? –  user7116 Dec 4 '12 at 16:06
1  
And again! Amazing. –  J. Steen Dec 4 '12 at 16:16
2  
now you would like to see this stackoverflow.com/review/suggested-edits/1044307 nikhil agarwal –  NullPoiиteя Dec 4 '12 at 18:39
5  
@NullPointer clicking "more" at this suggestion opens "Nikhil Agrawal has approved 1000 edit suggestions and rejected 0 edit suggestions". Mission accomplished :( –  gnat Dec 4 '12 at 19:26
2  
Again proving that high reputation and years of experience at the site are absolutely no guarantee for good reviews. –  J. Steen Dec 4 '12 at 19:28
1  
See also #2 in the Late Answers review queue: stackoverflow.com/review/late-answers/stats. I've e-mailed team@stackexchange.com for their upvoting of everything, including opvious spam and non-answers, but nothing seems to have been done. –  LittleBobbyTables Dec 5 '12 at 16:34

3 Answers 3

up vote 20 down vote accepted

If you can find one or two really bad edit approvals (blatant mistakes, ridiculously trivial edits or ones that make a post substantially worse), flag the first instance for review using the normal flag control under each answer (or question). Then flag the second bad edit approval and link to your suspicions like you have above.

If there are no bad edits, then perhaps we should send them a T Shirt. If there are bad edits, the moderator team will appreciate a heads up to help correct the account or look into whether it's a systematic weakness.

As a user - it's not only allowed, but encouraged to politely challenge individual actions and point out / ask for clarification if you see isolated instances of bad edits (or bad whatever). When there is a pattern, it's much better to ask once or twice and then point out the pattern to the moderator team. They have either better tools and/or more practice in correcting patterns of actions. When the community isn't jumping all over one person but instead focusing on isolated actions - it makes the moderator efforts at pattern correction more effective. My experience is having one or two user corrections yields better results for everyone since the user base and mod team are reinforcing each other without being overly negative or repetitive by flagging all instances of a pattern.

share|improve this answer
2  
How would I flag a review? Just flag the question/answer that's been edited, pointing to the review-number? –  J. Steen Dec 4 '12 at 16:01
2  
Yes - you can flag the answer or question that was edited and then link to the revision or just the edit history. Mods will know what to do. I've checked several and this looks like a lot of grammar / trivial edits that are getting approved. That's a bit of a hard case - guidelines as to what is "substantive" is always a little grey rather than black and white. –  bmike Dec 4 '12 at 16:06
    
I've truffled-sniffed out some rather bad ones, so no worries. =) –  J. Steen Dec 4 '12 at 16:07
    
Sorry 'bout changing the answer, but a moderator's said nothing can be done (sic), so I won't be flagging. =) –  J. Steen Dec 5 '12 at 16:06
    
No apologies needed - you can always edit my answers - choose another as the best. I disagree with CasperOne's blanket no since a moderator like to get flags when someone finds patterns on my site. Mods can't be everywhere and can't see everything and flags work very well to let users offload something that's troublesome to them and let the "pattern matching" "human exception handlers" chew on patterns of action so users can be in the here and now and address single actions and single edits. I do agree that in many cases we dismiss flags as helpful and often don't act immediately on patterns. –  bmike Dec 5 '12 at 16:33
    
My flags were unfortunately declined, but I may try again with a more blatant case of someone else working hard to destroy the quality of the SE network. –  J. Steen Dec 6 '12 at 6:42

There is no way you can see 200+ edits, reject none of them, and be making correct decisions. A certain amount of skipping as a new reviewer is to be expected. But after a while, Skip (even when it comes from an abundance of caution) is the wrong decision for certain suggested edits. One sign of thoughtful review is thinking - and that means pausing betweeen button clicks. Another is getting around to rejecting things that are bad, and not skipping everything that looks like it might call for a decision to be made.

Skip is not the right thing to do to an edit that adds a tag while leaving Thanks, a bunch of mispellings, and some misformatted code in place. Skip is not the right thing to do when someone has added a paragraph saying "this answer doesn't always work" and continuing on to basically embed another answer in the answer. Skip is not the right thing to do when someone edits a question to say "I have this problem too doesn't anyone have an answer?"

I just saw this: http://stackoverflow.com/review/suggested-edits#suggested-edits/1126186?&_suid=135464032377509624612113081006 - the question consists of a screenshot of the IDE with (apparently - the font is too small for me) some linker errors showing. The edit: add a tag and "Need some help on it." while leaving several ... in place. Golly, I wonder what the right thing to do is for this one? Perhaps it's borderline and I should Skip? Well two people approved it. Heaven help me.

Or this one: http://stackoverflow.com/review/suggested-edits/1126492#./1126492?&_suid=135464436606804926754819074999 the edit consisted in its entirety of adding [solved] to the title - something that is never right. Yes, I checked, ADDING. Removing might have qualified for "too minor" but this edit-suggester took a perfectly ordinary post and made it worse. 3 people rejected it, but one approved it. Anyone who thinks a high acccept-to-reject rate is probably ok needs to spend some time in http://stackoverflow.com/review/suggested-edits/history to see just how robo some folks are.

People who never reject are making the site worse. They should lose their reviewing privileges. We should care more about the effect of these people on reviewers. (see my earlier question: The robo-approvers are killing my will to review edits ) it was closed as a dupe though I think it is not a dupe: it contains a feature request I haven't seen elsewhere. and it focused on motivating good reviewers to keep reviewing.

But anyway: tell a moderator. Maybe some of them can be slowed at least a little.

share|improve this answer
1  
I've flagged a couple of reviews. That's the best I can do with the current system. But thank you for validating my own feelings about the expected quality of the content and community of the SE sites. =) –  J. Steen Dec 4 '12 at 16:13
    
And I did read that particular question, back when it was posted. I was at that time also reviewing queues less and less often, so I wholly agree and sympathise. –  J. Steen Dec 4 '12 at 16:15
    
Re: the edit: Luckily, Community rejected the post. What the automatic reasoning was, I can't figure out. –  J. Steen Dec 4 '12 at 17:05
1  
It looks like Commmunity rejected because the OP hand edited (to add another screen shot) but the "Need some help on it" text is in the question now, as is the new tag, so I'm not sure how rejected it was. –  Kate Gregory Dec 4 '12 at 17:07
    
@KateGregory, +1 perhaps with the Review Suspension reading, "you make really bad decisions". –  user7116 Dec 4 '12 at 17:19
    
@KateGregory robo-approvers question is maybe false-duplicated (for the record I voted reopen) because of the last paragraph, "Can we do something like etc..." This part somehow shifts the focus into how to punish bad reviewers from original problem description, which to me reads more like how to keep good reviewers motivated –  gnat Dec 4 '12 at 17:22
    
When commenting - please focus on the behavior - not the person. People are more open to listening/changing when criticism isn't directed at their person and instead one isolated action. Let moderators handle patterns. –  bmike Dec 4 '12 at 17:23
1  
"There is no way you can see 200+ edits, reject none of them, and be making correct decisions." I disagree, especially for a newer reviewer. I know when I started reviewing I only approved edits I knew were correct, and didn't vote on others for a while until I was able to learn what is and isn't considered appropriate on the site. When I found clearly borderline content I would either make a meta post asking about what should be done, or just favorite the links to see what others did. Taking the time to learn the system shouldn't be discouraged. That's not the case here, but still. –  Servy Dec 4 '12 at 18:18
    
@servy - for 200+ edits? You needed that long to gain the confidence to start rejecting the kind of things I've been unable to resist linking in here - adding [solved] to a title, adding a random "need some help here" to a truly awful question, and so on? If so, I think you're very much an edge case, and that most people would start to reject sooner than that. –  Kate Gregory Dec 4 '12 at 18:27
    
@KateGregory 200+ would be unusual, but in no way a problem. As long as none of those 200+ edits should have been rejections I wouldn't have any problem with it at all. Now, for this particular user, it sounds like there are a number of already found clearly wrong reviews, but in the general case a reviewer is only "bad" if they have reviews for which they have taken the improper action. On the flip side of things, if a reviewer skips items that are obvious approvals and uses their daily votes just to reject items (that should be rejected) that would be fine too. –  Servy Dec 4 '12 at 18:29
1  
@servy, fair enough. I believe that a little poking into the specific reviews done by these very one-sided reviewers will ALWAYS reveal at last one very bad review, but I will grant you that it is at least theoretically possible to achieve that pattern while being a good reviewer. I'll edit my first sentence slightly. –  Kate Gregory Dec 4 '12 at 18:34
1  
@KateGregory If you look into the history of even very good quality reviewers you'll surely find at least one review that's wrong and/or questionable, if from nothing else than a mis-click. I would say that having 200 approves and 0 rejects is a warning sign, and it should result in the user being investigated as it's a probable sign of abuse. That's different from actual abuse. –  Servy Dec 4 '12 at 18:37

Yes, you should.

We're kicking blatant abusers out of the queues automatically now, but if someone isn't doing a good job, the mods can give them some guidance... Or at least give them a time-out from /review (currently, they have to pass this on to me, but at some point soon they'll be able to just do this directly).

As a general rule, whenever you see someone using the site in bad faith - whether that means tons of spammy answers or pages of lousy edits or scores of thoughtless reviews - raise a red flag. The moderators might not be able to do anything directly, but if not they'll get in touch with someone who can.

share|improve this answer
2  
I'm rather glad you're invalidating what's been said to me before in a flag-decline-message - that nothing can be done about this - and will take this as acceptable proof that the first answer I got - where I was urged to flag - is correct. –  J. Steen Jan 29 '13 at 8:47

You must log in to answer this question.

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