What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 130 Stack Exchange communities.

I often see very good posts on SO which go unacknowledged because of bad grammar / lack of clarity. Unfortunately, if the errors aren't fixed quickly the posts tend to die. I would like to see some way for suggested edits to be processed more quickly.

One solution would be to allow bypassing of the queue for "established editors" - users having high "edit weight" (by analogy with high "flag weight").

Right now, every edit must be vetted by a user with 2k+ rep. While I agree that a 2k reputation threshold for immediate editing is a good idea, it seems like it may be missing the point slightly: reputation and editing ability may be correlated, but a user with high reputation isn't necessarily qualified to correct grammar, fix formatting, etc, and users with low rep aren't necessarily unqualified.

This is going to be a problem for any privilege, of course, and the art is usually finding the appropriate threshold. In the editing world, however, we have another figure of merit: suggested edits. It seems like a user who's suggested 400 edits with a 98% acceptance rate is more deserving of editing privileges than someone who's asked 15 good questions and answered another 15 particularly well.

What I propose is some kind of short-track to editing privileges for "established editors", and maybe a corresponding badge. The advantages would be:

  • continued incentive for users with 1k+ rep to edit
  • Less burden on 2k+ users to vet edits
  • a potentially cleaner way to discriminate between 'good' and 'bad' editors
  • a way of rewarding editors without actually giving them rep (I imagine they get so little now because we don't want anyone to become a trusted user just by editing)

Of course, there are disadvantages to this proposal. The decoupling between privileges and reputation would be completely new to SE, and quantifying users with more than one number may be too complicated. But editing is important: it keeps SO professional, teaches new users how to post, and improves overall quality. It seems worth bending the norm if the result is better editing.

share|improve this question
2  
wrt rewarding editors without actually giving them rep - are you aware of Editor, Strunk & White, Archaeologist and Copy Editor badges? –  gnat Apr 20 '12 at 7:50
    
@gnat sure, but the Strunk & White badge is the only one that actually unlocks anything, and while badges are cool I think gaining privileges is cooler. –  Shep Apr 20 '12 at 7:55
3  
well 2 points rep gained with every accepted suggested edit moves one up the ladder to unlocking privileges in the regular way doesn't it? This works up to 1K threshold - isn't it already good enough, unlocking cool stuff up to established user? –  gnat Apr 20 '12 at 8:04
    
@gnat this isn't about unlocking cool things being "good enough". It's about improving the quality of posts via a more balanced set of incentives. There are a number of good questions on SO which get downvoted and / or not answered because the phrasing is unclear or the english is bad, and I'd like to see more reward for those who take the time to improve these. –  Shep Apr 20 '12 at 8:24
    
are you looking for something like bypassing edit review queue for "established editors" who didn't reach 2K reputation to unlock? –  gnat Apr 20 '12 at 8:29
1  
@gnat that's exactly what I'm proposing –  Shep Apr 20 '12 at 8:37
    
I see. This is not quite apparent from your question title / text - I would say it could benefit from some editing –  gnat Apr 20 '12 at 8:45
1  
@gnat edited, let me know if it's still unclear –  Shep Apr 20 '12 at 9:52
    
rev 4 looks good to me –  gnat Apr 20 '12 at 10:00

2 Answers 2

...quantifying users with more than one number may be too complicated

Not really complicated. Stack Exchange already has something like that implemented and working fine.

I mean flag weight feature that is used to arrange flag reviews queue to simplify handling flags submitted by users with proven ability to do it right.

  • I for one can easily imagine edit weight number / feature that would arrange suggested edits queue so that suggestions from "established editors" tend to flow to top.
     
    It takes only a minor effort to rephrase current description of flag weight to make it:

    Edit weight is a measure of how well a user proposes edits. Users with a history of suggesting edits helpfully — as judged by reviewers who act on those suggestions — earn high edit weight, and vice versa.

    All users start with edit weight of 100. Scores can fall to as low as 0 and rise to as high as 750.

    Edit weight is not displayed in users' profiles due to sustained community disapproval of similar feature that has been there in the past for flag weight, though it is used behind the scenes for ordering edits in each site's edit queue.

    User profiles show a history of suggested edits in "activity -> suggestions" tab.

Weight-based queue reordering is pretty close to "bypass" - simply because of the difference between being 200th and 5th in the line of suggested edits waiting for review.


As for your other note,

decoupling between privileges and reputation would be completely new to SE...

I think above is a very strong point in favor of keeping things coupled as they are now.

To justify a radical change like that, one better gets very compelling data to back it up.

share|improve this answer
3  
I really like this solution: I wasn't aware of flag weight when I posted my question, but I think something similar would serve the same purpose as my suggestion. –  Shep Apr 20 '12 at 9:45
    
The flag weight idea is not a bad one, and it could be used to actually reward editors and not just place them higher in a small(ish) list. Even if you are at the bottom of the list, your suggestion will be processed soon enough... –  Yannis Aug 31 '12 at 12:06

I don't really like this.

Depending upon the reviewer, even minor suggested edits are accepted. That doesn't imply that the user is trusted by the community. Suggested edits are small things.

On the other hand, questions/answers are a good measure of how trusted you are. For example, I can suggest spellchecks all I want on smugmug.SE--but that doesn't mean I'm qualified to make arbitrary changes to actual post content.

but a user with high reputation isn't necessarily qualified to correct grammar, fix formatting, etc, and users with low rep aren't necessarily unqualified.

It's not qualification based. It's trust based. The high rep users who have bad grammar will be disciplined enough to not make those kinds of edits. They probably know that their grammar is so-so, and will be expected/trusted to not meddle in grammar issues.

Low rep users may or may not be qualified, but they aren't trusted.

It seems like a user who's suggested 500 edits with a 90% acceptance rate is more deserving of editing privileges than someone who's asked 15 good questions and answered another 15 particularly well.

Same issue here--Q&A is a better way to gauge trustworthiness. Suggested edits aren't.

Also, a 90% accept rate isn't that good IMO. If you want the privilege just from your edits, then it ought to be close to 100%--you need to show that you can make valuable, good edits all the time.

Anyway, if you've suggested 500 edits with a 90% rate, then you got 900 rep from it. You're nearly halfway there towards full edit privileges. I suggest you do a bit more posting, and you'll get the privilege soon enough :)

share|improve this answer
    
I can recommend this answer. Would read again. –  Daniel Fischer Apr 20 '12 at 7:59
    
The accept rate and number of edits are completely debatable, the point is that if someone has made lots of good edits, I trust them to continue making good edits. I trust them to do that (and only that) more than I trust someone who has posted a lot of good questions with questionable grammar. –  Shep Apr 20 '12 at 8:08
    
@Shep: No, I would not-- Saw my smugmug.SE point? Editing isn't all about grammar. It's about improving the post content as well(clarifications, etc). And that requires subject knowledge. –  Manishearth Apr 20 '12 at 8:11
    
Very good point, and I'm sorry that my post has come off as overemphasizing grammar. But the objection seems to center around an assumption that someone who has edited well in the past is less likely to edit well in the future than someone who's asked / answered questions well in the past. It seems like asking / answering questions a rather indirect way of proving one's editing abilities, whereas editing (well) seems like pretty direct proof. –  Shep Apr 20 '12 at 8:19
1  
@Shep: Trust is trust. There's no editing-trust and posting-trust. Trust is gained from your cumulative actions. This is not about editing "abilities", this is about trust. And, like I said, posts gauge trust more than edits. (Also, posts are seen by a larger number of people). –  Manishearth Apr 20 '12 at 8:25
    
I understand the SO definition of trust. I know that reputation is a gauge of how much the community trusts you. And I recognize that what I'm proposing is a slight departure from the "trust is trust" way of thinking. But in reality people are multidimensional: you may trust your grandmother, but it doesn't mean you'll let her debug your code. –  Shep Apr 20 '12 at 8:39
1  
@Shep: Your analogy goes in the opposite way-- what you're proposing is "trusting the guy who debugs your code to be your grandmother for a day" (or something like that :P, twisting your analogy makes it strange ). As I've pointed out before, high reps won't edit if they aren't good at it--"You can trust your grandmother to abstain from debugging your code" –  Manishearth Apr 20 '12 at 8:44
1  
No, I'm saying the two types of trust are independent. There would be no more rep gained by editors, only editing privileges. –  Shep Apr 20 '12 at 8:50
2  
@Shep: Yes, but editing can be plain 'ol grammar--which has no indication of how much you know the site matter. Once you get editing privs, you can make potentially wrong edits to the main matter. Slightly paranoid, but 2k rep isn't that big a deal IMO. –  Manishearth Apr 20 '12 at 8:51
    
let us continue this discussion in chat –  Shep Apr 20 '12 at 8:57

You must log in to answer this question.

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