What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 134 Stack Exchange communities.

I'm keeping a consistent ~10% rejection rate on my suggested edits (not that I agree with some of these rejections):

a

I'm just wondering what the community average is, across all users on Stack Overflow?

share|improve this question
1  
@ToonKrijthe I'm referring to the other way around. I.e. for me as an editor, not a reviewer. –  Danny Beckett Mar 21 '13 at 13:56
    
Ok, that explains the 10%. Try to fix anything you can in a post. –  Toon Krijthe Mar 21 '13 at 13:57

4 Answers 4

up vote 21 down vote accepted

Here you go: using Data Explorer

SELECT 
COUNT(*) AS TOTAL
, SUM(CASE WHEN ApprovalDate IS NOT NULL 
           THEN 1 ELSE 0 END) AS Total_Approved
, SUM(CASE WHEN RejectionDate IS NOT NULL 
           THEN 1 ELSE 0 END) AS Total_Rejected
, SUM(CASE WHEN ApprovalDate IS NOT NULL 
           THEN 1 ELSE 0 END) * 100 / COUNT(*) AS Average_Approved
, SUM(CASE WHEN RejectionDate IS NOT NULL 
           THEN 1 ELSE 0 END) * 100 / COUNT(*) AS Average_Rejected
FROM SuggestedEdits;

Average suggested edit rejection rate

share|improve this answer
11  
Great answer, thanks!! tl;dr: 18% –  Danny Beckett Mar 21 '13 at 14:03

Not an answer, but a serious advise. I have re reviewed some of your suggestions, and I can advise you to go for the bigger fish

  • This edit only adds some backticks, these are suspicious. And often rejected.
  • This edit only corrects the title, that is too minor.
  • This edit changes only part of the post, for example the bad formatted table could have been fixed, and again some backticks.
share|improve this answer
1  
You've chosen 3 of my accepted suggestions, so I'm not sure why you've chosen to give me "advise" on these, rather than my rejected suggestions, but here are my comments: 1. Regardless of whether suggestions like this are "suspicious", I edited everything that needed editing within that post. 2. Consensus is to remove tags from titles; there was nothing else to change in the post. 3. I have no idea what you're talking about - the whole post was thoroughly cleaned. –  Danny Beckett Mar 21 '13 at 14:10
2  
This is not an attack, I try to give you some advise. –  Toon Krijthe Mar 21 '13 at 14:15
    
Thanks for your advice Toon, but I respectfully disagree (it seems like 9 other reviewers also do). –  Danny Beckett Mar 21 '13 at 14:17
2  
@DannyBeckett Well, if you want to talk about rejected edits, that one really is too minor. And insulting reviewers in your edit message is also not the best approach. –  Bart Mar 21 '13 at 14:35
    
@Bart Can you please tell me, once and for all: is removing a tag from a title, where there is nothing else to edit in the post, given that the consensus is to do so, really too minor? –  Danny Beckett Mar 21 '13 at 14:37
3  
@DannyBeckett Our definition of "where there is nothing else to edit" differs quite significantly. –  Bart Mar 21 '13 at 14:40
    
@Bart You're right, I apologise, I just saw your revision. But is the answer to my previous comment, no? –  Danny Beckett Mar 21 '13 at 14:48
    
@DannyBeckett If there is absolutely nothing else to fix, I will personally not reject such an edit. But as always, the focus should be on the post as a whole when editing. Not the single offending items you noticed. –  Bart Mar 21 '13 at 14:49
1  
@Bart Thanks for your help! You too Toon. I do appreciate the feedback, although I might not always come across like so. –  Danny Beckett Mar 21 '13 at 14:50
1  
No hard feeling ;-). –  Toon Krijthe Mar 21 '13 at 14:59
2  
Not only would I reject all 3 of those edits, I advocate audits to punish people who approve them. Suggested edits should address multiple problems with a post; leave the tiny details to people who can edit directly without clogging the review queue. –  Wooble Mar 21 '13 at 14:59
    
@Bart Can you comment on this suggested edit I made that was just rejected? I don't feel it should have been rejected. Thanks again. –  Danny Beckett Mar 21 '13 at 15:32
1  
@DannyBeckett That I would have accepted. But I can see why it was rejected. At first glance my thought was "why did he remove all that content and destroy it?". Primarily with regards to the first code block. More carefully reading through your edit I realized that your edit was not destructive at all, but that was not my first impression. Unfortunate case I'd say. –  Bart Mar 21 '13 at 15:40
1  
@DannyBeckett the only real thing I can think of for such extensive edits is to clearly describe it in your edit message. But then again, some might simply skip that. Get the required rep to no longer have your edits reviewed I'd say. ;) As for punishing people, I'd rather punish those who accept stuff that shouldn't be. The occasional mistaken rejection I don't mind. ;) –  Bart Mar 21 '13 at 15:44
6  
@DannyBeckett Believe me, false rejections are annoying, but by no means as big a problem as false acceptances. And this one got approved now. That said, please keep your edit summaries friendly. People are easily triggered on the internet. As anywhere else on the site, be nice. No matter how frustrated. ;) And sorry for the persistent comments Toon. I'll stop here. –  Bart Mar 21 '13 at 16:02

So, I completely misread the question but leaving this here anyway as it give the other point of view.

hims056's answer gives the rejection rate of suggested edits from the point of view of the suggester; to do it the other way around.

In order to calculate the average rejection rate of the user who approves edits the SQL would look something like this (assuming I've read how SuggestedEditVotes works correctly)

with user_votes as (
select userid
     , cast(sum(case when votetypeid = 2 then 1 
                     else 0 
                end) as float) as approved
     , cast(sum(case when votetypeid = 3 then 1 
                     else 0 
                end) as float) as rejected
     , count(*) as total
  from suggestededitvotes
 group by userid
       )
select cast(avg(approved * 100 / total) as numeric(5,2)) as [Approval Rate]
     , cast(avg(rejected * 100 / total) as numeric(5,2))as [Rejection Rate]
     , cast(avg(approved) as numeric(5,2)) as [Average Approved]
     , cast(avg(rejected) as numeric(5,2)) as [Average Rejected]
     , cast(avg(total) as numeric(5,2)) as [Average Votes]
  from user_votes

The answer is 7%.

Removing outliers1, those people who approve 100% of everything and those who have voted on 10 or less the average rejection rate rises to 17.57% and SQL becomes:

with user_votes as (
select userid
     , cast(sum(case when votetypeid = 2 then 1 
                     else 0 
                end) as float) as approved
     , cast(sum(case when votetypeid = 3 then 1 
                     else 0 
                end) as float) as rejected
     , count(*) as total
  from suggestededitvotes
 group by userid
having count(*) <> sum(case when votetypeid = 2 then 1 else 0 end)
   and count(*) > 10
       )
select cast(avg(approved * 100 / total) as numeric(5,2)) as [Approval Rate]
     , cast(avg(rejected * 100 / total) as numeric(5,2))as [Rejection Rate]
     , cast(avg(approved) as numeric(5,2)) as [Average Approved]
     , cast(avg(rejected) as numeric(5,2)) as [Average Rejected]
     , cast(avg(total) as numeric(5,2)) as [Average Votes]
  from user_votes

1. Definition randomly made up by me with thought but without proper analysis.

share|improve this answer

So that's a good question and I appreciate that you're thinking critically for how to best contribute to SO, but that's not really the point of suggested edits. When you reach a certain reputation (2k, is it?) you can edit anything you want without it needing to be approved. The suggested edit queue up until that point serves a primary function of training new users to edit properly.

So you will graduate from your edits needing approval regardless of what "grade" you finish with. That being said, really, your goal is to strive for all edits being approved. And certainly understand why your rejected ones are rejected. If you disagree with some, then that's fine, you won't need approval in time and will be on equal standing in the community with your rejector on this issue.

FWIW - on the "too minor" ones, I go with the rule "Did this increase how fast I was able to process the question?" Most do so I approve most, and I nearly always approve title corrections since those impact the presentation of the post to the site and internet at large.

share|improve this answer
    
Thank you for your comments, but I've downvoted it due to not answering the actual question. I didn't ask the purpose of the privilege restriction etc. It was a simple, statistical question. Thanks though. –  Danny Beckett Apr 1 '13 at 21:44

You must log in to answer this question.

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