What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 133 Stack Exchange communities.

Until I started using the new suggested-edit review system I wasn't aware that you were capped at voting approve/reject on 50 edits a day (having less than the required rep for the existing tool meant running into posts requiring approval were quite rare).

Since using the review system and taking a look at the stats page, I've noticed that there's a reasonable number of users that are reviewing more than the 50 edits that they can vote on. This can be seen in the screen grab I've borrowed from this question, where you'll notice that John has reviewed 108 posts in one day:

wat

This isn't an attack against John, I'm sure that he's doing good work and he certainly has enough rep to be making the edits by himself anyway but is this a desired effect? Is it an unexpected effect possibly motivated by the 'Top Reviews all time' list?

I have a few concerns (and some of these may be covered by tools available to mods/users with greater rep anyway).

As far as I can tell, when we chose to improve, we can select helpful/not-helpful for the edit we're improving and that approval/reject is marked as community. Is this still tracked against the individual making the decision (I can't see it in their suggested edit review page)? For example, it seems like there would be a reasonable amount of work involved in determining if one user was always selecting 'improve + unhelpful' to reject a particular editor.

When you're using the review system, you see a lot of posts from tags that you wouldn't necessarily normally look at. If you're making decisions on your own in those tags, it's less likely that somebody else will catch any mistakes you make (for example). We all make mistakes (I'm sure there are plenty in my review history), but the current double review systems mitigates that to some extent, a rise in unilateral decisions seems like a bad thing.

If people are regularly hitting the max number of votes, would it be better to increase the number of votes per day?

Would it be better to count improve decisions count in your votes per day, or would this just create an unclearable backlog?

share|improve this question
2  
I don't believe so. A lot of people (including me) have always done this, even before the review-beta stats tracking. –  animuson Jul 27 '12 at 8:15
    
Note that you have more options than Approve or Reject when reviewing edits, which do not count against your flags but do count as an action taken in the stats you have see. 108 actions is not the same as 108 flags. –  Martijn Pieters Jul 27 '12 at 8:19
    
@MartijnPieters: I'm not sure I follow, the options for suggested-edits appear to be (approve, reject, improve, not sure). The first two act as a vote + increase your reported stats, 'improve' only increases your reported stats, 'not sure' doesn't impact either. –  forsvarir Jul 27 '12 at 8:29
    
@forsvarir: Okay, indeed, on suggested edits there are no flag counts. With enough rep you can make unlimited edits, and cast unlimited reject or approve votes. Retracting my comment, I had that conflated with the other review categories. –  Martijn Pieters Jul 27 '12 at 8:33

1 Answer 1

I don't have figures, but I do not think your analysis is correct. Under both the old and the new interface, you can do a maximum of 50 approvals plus rejections every day, and improvements are unlimited. The only thing that's changed is that there is now an indicator that includes improvements as well as straight approvals and rejections.

Improvements are more work than straight votes: you need to fix something else. Even plain reviews are, let's face it, pretty boring. I don't doubt that there are reviewers who are doing it for the sake of the badges or the dashboard, but I the least effort for them is to do 50 a day and never improve.

I prefer that improvements be encouraged. I see plenty of suggestions that don't go far enough or do the wrong thing altogether. Improving is better than only rejecting the suggestion (or approving if good but incomplete).

There is less of a risk of error when improving because it involves reading the post more closely. The most common error I see in suggested edit reviews is approving crap edits anyway; when people bother to click “Improve”, they usually do improve the post. Besides, on every site except Stack Overflow, suggested edits are reviewed by a single person.

The number of votes per day was increased from 30 to 50 due to the huge backlog. Having done 50 suggested edit reviews (plus I don't know how many improvements as there's no easy way to count them; probably in the 10–20 range) almost every day for the past couple of months, I find that 30 was ok, 40 was a lot, and 50 was too much. (I only did it because someone has to do it.) I do not want the quota to be increased. In fact, I hope it can be decreased back to 30 (not counting improvements, in the interest of encouraging improvements) now that the review queue isn't almost always saturated.

share|improve this answer

You must log in to answer this question.

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