2

I would like to avoid making suggested edits on poor edit candidate posts, as discussed in this post.

There does not appear to be much advice here on how to select a good candidate post for making a suggested edit. The general advice "don't polish a turd" is good, however it does not tell us what we should try to polish. I would like to come up with some simple rules that a <2K editor can follow to identify good edit candidates.

For the sake of this discussion, let's take it for granted that the <2K editor has some competence, etc., in the domain of the posts that they are thinking of editing.

Good Candidate Posts for suggested edits

Obviously, Nice Questions and Nice Answers (those with a score of 10 or more) are good candidates for editing as the community has acknowledged that they are useful.

But is it a good idea to make a suggested edit on a high scoring post?

An analysis of the Suggested Edits data in SEDE shows that the rejection rate for suggested edits in 2023 is as follows:

Analysis of Suggested Edits for 2023:

Score Posts Accepted Rejected Rejection Rate
0 46464 40925 5539 12%
1 15818 13653 2165 14%
2 6156 5082 1074 17%
3 3166 2523 643 20%
4 2081 1635 446 21%
5 1455 1128 327 22%
6 1102 836 266 24%
7 882 661 221 25%
8 737 552 185 25%
9 623 479 144 23%
10-19 3518 2559 959 27%
20-29 1618 1153 465 29%
30-39 1000 681 319 32%
40-49 707 478 229 32%
50-59 517 336 181 35%
60-69 384 257 127 33%
70-79 306 185 121 40%
80-89 281 182 99 35%
90-99 227 144 83 37%
100-199 1197 738 459 38%
200-299 537 326 211 39%
300-399 280 162 118 42%
400-499 173 91 82 47%
500-599 111 63 48 43%
600-699 102 51 51 50%
700-799 72 40 32 44%
800-899 38 21 17 45%
900-999 44 23 21 48%

From this analysis, posts with a score of 3 have a rejection rate of about 20%, whereas posts with a score of about 30 have a rejection rate of about 30% and posts with a score of more than 100 have a rejection rate of about 40%, reaching almost 50% at a score of 1,000.

It therefore seems to be a good idea to avoid making suggested edits on high scoring posts as the probability of rejection is far higher (unless you have gained some experience and already have a good acceptance rate).

The post score used in the above table is the current score whereas the score may well have been different when the suggested edit was made. I am not sure if it is possible to extract the post score at the time the suggested edit was made in SEDE.

Poor Candidate Posts for suggested edits

At the other end, posts with a negative score are most probably poor edit candidate posts. They are probably on their way to being closed or deleted.

Similarly, posts with a zero score may well be poor edit candidates. Since no one has voted them up or down yet (or users are voting both up and down), we may well assume that the community has not made up its mind about those posts, so it may be advisable to avoid editing them too.

Looking at the above table, we may wonder why the majority of suggested edits have been made on zero score posts (and then accepted at an acceptance rate of 88%).

What about posts with a score of 1? If only one person has upvoted a post then it is probably too early to assume that the community has made up its mind about the post.

So what else can we take into consideration?

  • The score of the question if the post is an answer
  • The score of the answer(s) if the post is a question
  • How many answers does the post have (and whether those answers have been upvoted)
  • The number of views of the question
  • The age of the post
  • Does a question have an accepted answer
  • Has an answer been accepted
  • How many comments does the post have

If the edited post is deleted then you will lose your +2 reputation, however this answer reminds us that you won't lose your reputation from a suggested edit if the post is 60 or more days old and has a score of 3 or higher.

We could also consider the fact that the Archaeologist badge is rewarded for editing posts more than 6 months old.

Finally, why should you (as a <2K user) avoid making suggested edits for a poor edit candidate post?

Here are several reasons (I am sure there are more):

  • You are wasting your own time on an unnecessary edit
  • You are asking two or three >2K users to review your suggested edit—which also takes time
  • If your suggested edit is approved, and the post is still a poor post then it may eventually be deleted in which case you will lose the +2 reputation you gained for the suggested edit

I would therefore like to ask how can a user with a reputation of <2K quickly identify a good edit candidate post and avoid making a suggested edit for a poor edit candidate post.

Following this analysis, as a <2K user, I have decided to only make suggested edits on posts that have a score of 4 or more and are more than 6 months old.

The above table was generated in Excel with the help of this SEDE query.

3
  • 3
    Answered questions where the author can't delete their own post are a good candidate. Commented Dec 16, 2023 at 18:37
  • "At the other end, posts with a negative score are most probably poor edit candidate posts." Or good candidates, given they might actually have been downvoted because of lack of clarity or irrelevant information (provided, obviously, that the question itself is on-topic).
    – Joachim
    Commented Dec 16, 2023 at 18:51
  • 1
    That being said, I primarily make suggested edits for the actual edit, not for the reputation. After all, I made over 300 suggested edits here and stood to gain nothing from them since I was an anonymous user. Commented Dec 16, 2023 at 20:43

1 Answer 1

2

Very simple? Multiple upvotes

You can find all posts scoring 2 or more in search with score:2... For better results, wait until the voting on the post has settled down.

Quite simply, these are posts that multiple people think are valuable, and as a result are not eligible for auto-deletion. Therefore, multiple high-rep users would have to vote to specifically remove the post or OP would have to lose reputation to delete it. (Note: Moderators can always delete a post unilaterally but they shouldn't be doing that at random.)

Auto-deletion is the most common reason a question is deleted on most sites by far. Answers are usually deleted for not following the rules (e.g., new users leaving a comment) but there are a non-trivial number of self-deletions too.

You can of course make things more complicated by memorizing the specific rules for auto-deletion and the reasons an author won't be able to delete their post, but that's really not necessary. As you get more experience on a site, you should be able to better predict intuitively which posts will be well received and which ones won't (and will be deleted).

Make good edits

This can be really complicated and can vary greatly between sites. The principles are the same: fix as many superficial errors in the post without changing the meaning—comments are the place to explain why an answer is fundamentally wrong.

Your best bet is to stick with edits that fix all the spelling and grammar mistakes (and are more than 6 characters without hacks). Fixing broken links with an archive version is also generally a safe edit. If you know how (and yes there are some gotchas), formatting should be fixed as well, but shouldn't usually be made as an edit by itself unless it's a major fix (such as putting unformatted code in code blocks).

2
  • Do you have an explanation for the phenomenon OP notes of higher-voted questions seeing higher edit rejection rates? Commented Dec 18, 2023 at 9:09
  • 2
    @KarlKnechtel People try to make different types of edits to high scoring posts. For example, the knowledge that a new answer will never outscore the old drives people to try to add new code as an edit instead.
    – Laurel
    Commented Dec 18, 2023 at 12:14

You must log in to answer this question.

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