How to fix incorrect math in old accepted answers?

A couple of weeks ago this question was posted, and the question contained a simple typo in the equations for coordinate rotation. The first answer spent two lines politely correcting this. Then, as it happened, I noticed that the same typo was present in the accepted answer to this linked related question from three years ago. So I thought, maybe the error in the new question was propagated from the old answer to the new question? And then I made a simple edit to the old answer in order to correct the typo. The edit changed the incorrect equation $$y = x'sin\,\omega t + x'cos\,\omega t$$ into the correct one: $$y = x'sin\,\omega t + y'cos\,\omega t$$.

My edit was rejected since it allegedly didn't improve anything, despite being objectively correct while the unedited answer still contains incorrect math due to the typo. However, after learning about the rules for editing I now understand that edits are not intended for things like fixing trivial typos (however game-changing they are) in accepted answers.

So my question is: what is the recommended procedure for fixing math errors in old answers, to avoid propagating such errors?

• Just leave a comment in the answer pointing out the typo and move on with your life. Sep 23 '18 at 19:33
• $sin$ and $cos$ is still wrong because they are both named functions and should be written as \sin to get $\sin$. Sep 23 '18 at 20:46
• Some of this advice is relevant here.
– rob Mod
Sep 23 '18 at 21:03
• And some of this advice Sep 23 '18 at 21:05

3 Answers

This is a little complicated because your edit did two entirely separate things:

• It fixed a typo that created an error in the math. This is good.
• It added extra useless content to the post (the periods at the end). This is bad.

Adding the extra useless content is bad enough that the harm it causes to the post outweighs fixing the mistake in the math, and that's presumably why your edit was rejected - or at least, that's why I would have rejected it if I had been reviewing it.

I do want to clarify that

...edits are not intended for things like fixing trivial typos (however game-changing they are) in accepted answers.

is not correct. Fixing typos, especially when they impact the meaning of the post (such as the one in this case), is a perfect reason to edit. The catch is that we want to keep the total number of edits down, so you shouldn't just fix the typo itself, but you should take a look at the rest of the post and see if there's anything else that needs to be fixed, and bunch it all together into a single edit. The 6-character requirement is a crude way to "enforce" that. Once you get 2000 reputation, you'll presumably have learned to edit everything that needs to be fixed at once, and at that point you gain the ability to make edits with less than 6 characters' change.

So what you should have done, in this case, is look for something else to fix in the post to go along with your change of $$x$$ to $$y$$. Read through it carefully and see if there are any missing words or misspelled words, missing or inconsistent math formatting, etc. All of that you can use to build up to the 6-character minimum. It simply needs to be the case that every individual change you make improves the answer. Of course, sometimes you'll find a post that is really well written and has nothing else to change, and in that case you just have to try to get someone else to make the change for you by commenting on the post and/or dropping a note in chat.

TL;DR: If you can't make the minimum 6 characters to suggest an edit, don't do it.

Personally, I'd have deemed your edit as harmful due to the use of the periods at the end of the post, even though the $$y\to x$$ was correct. Since your edit was really for a single character, what AccidentalFourierTransform said is correct: leave a comment indicating the incorrect point. Don't waste other people's time just to farm 2 rep, it's really not worth it.

While I don't see any typographical errors in the post peterh links, I do see that you could have replaced all italics d's with the correct, romanized d's (i.e., $$\mathrm{d}x$$ instead of $$dx$$), which would have easily jumped you well over the required 6 characters.

• Thanks for this answer. It reinforces what AccidentalFourierTransform said. I'll add a comment that points out the mathematical error, and move on with my life. I don't understand the subtleties of farming reputation points, so I didn't know that my edit attempt was wasting people's time. Maybe this aspect of editing should be made more clear for newbies like me. Sep 23 '18 at 22:14
• @CuspyCode anything that is reviewed by someone else is using someone's time. When rejected due to some harmful items to the end, it could be viewed as wasting their time since it'll be quickly rejected. When approved, I don't think it's wasting time. Sep 23 '18 at 23:09
• Getting the +2 for an edit is not "just farming 2 rep", it is the deserved reward for finding a bug in a post. (Not I was the down.) Sep 24 '18 at 0:21
• @peterh anything that earns you rep for doing something & you continually doing it is farming: asking questions, answering questions, etc. Sep 24 '18 at 0:26
• @KyleKanos Yeah, but you used “rep farming” with a negative connotation. If everything counts as rep farming, it’s a rather meaningless charge, isn’t it? Sep 24 '18 at 13:05
• @knzhou except since I don't view rep farming negatively (because that's the incentive to contribute), you're putting in your bias where it doesn't belong. That is to say, I didn't put any negative context in here. Sep 24 '18 at 13:29
• @KyleKanos Check this, maybe it might be a rare point where we agree. Of course the community reacted as it was pretty well predictable. Oct 4 '18 at 1:10
• @peterh i don't know that we agree here because I don't understand what you're saying in that post. The community seems to have DV'd and closed it because of (a) it not actually being a question and (b) it not making much sense (even with your comments). Oct 4 '18 at 10:03

This is the review record of your rejected edit.

I think the reasons of the rejections were these:

• Some of the reviewers represent more the negative voters of the site (gives more down as up, prefer to vote to close, rejecting edit suggestions and so on). You had bad luck in this sense.
• You wanted to edit the post of a high-rep user. Such edit suggestions are typically not very HQ.
• Being a <2k user, you can only initiate revisions if they change at least 6 characters. You circumvented this rule by inserting 6 periods to the end of the text.

However, the problem what you fixed was a real and very obvious one.

If I had reviewed your suggestion, I had probably clicked "Improve edit", and I had removed your added points at the end. In my opinion, it had been the correct thing to do, but also the reviewers are human and sometimes they have to handle a large mass of crap.

Writing a comment to the post, as AFT suggests, could work, but you have no way to know, when will anybody see it ever. Furthermore, you lose the +2 reward with it, which is your deserved reward for finding a bug in a post.

You can also find other ways to reach this 6 character limit. If there is really nothing in the post what could be done better, then improve its tagging (if it is a question). If you can't do that, then this post is non-editable for you and only AFT's idea remains.

Circumventing rules by tricks is generally unwelcomed and serves as a strong argument for the reviewers to click "Reject". Btw, if you had done it at least as a HTML comment (<!-- .... -->), then it had been at least invisible in the rendered version. However, it had been still visible in the markup and counted as circumventing rules.

In such an obvious case, I had either posted a comment, but only after I had looked for other ways to reach the 6 char limit.

The comment what you write to the edit suggestion is super-important. Write clear and convincing arguments to your edit suggestions, the reviewers see and check that!

If the case is very obvious and you are sure that the reviewers mishandled it, then it is not very bad if you suggest the same edit again (this time with more convincing comment). Last time I fixed the formatting of an SU mod on this way (rejection, accept). But never do it 3 times.

I edit the post of high-rep users very rarely, despite that now I could.

Coming to the meta is also a good idea. You might also try the site chatroom.

• The statistics for the reviewers of this edit are available by clicking (more) on the review record, and in their light, your first bullet point is factually inaccurate. Can you comment on what evidence you have to back it up? Similarly your second bullet point raises a statistical observation; what evidence did you use to come to that conclusion? Sep 23 '18 at 21:10
• @EmilioPisanty You might also regularly check other review statistic pages, and also activity histories of the reviewers. As 10k+, you might also get extra info in the review reviews functionality. I do these (except the last) regularly since years, to understand the main site dynamics here. My statements are based on my this experience, although I could construct SEDE queries for that if you really want. However, SEDE has typically very few information about rejected reviews and the review votes (they are available on the user and review entity pages). Sep 23 '18 at 21:16
• @peterh Your post makes a very clear accusation of bias against the reviewers and it very clearly asserts that it was reviewed by users with a record of rejecting more edit suggestions that they accept. That accusation is in clear conflict with the facts. I'm unsure why you levelled it and why, after its factual inaccuracy has been pointed out, you've chosen to retain it without backing it up. Sep 23 '18 at 21:29
• @EmilioPisanty There are significant differences in the lenience of the reviewers, it is not an accusation but a fact. The site concept to handle these differences is that multiple reviewers handle all the cases. It significantly decreases this random factor, but doesn't eliminate it. If you check the review page (I linked in the first line of the post), I think from high school level math we can see that it is typo. Simply rejecting the suggestion and not improving it, and that both reviewers did it so, was the effect of this random factor. This is what I tried to say, and I think I did. Sep 23 '18 at 21:34
• I think that there isn't anything wrong with rejecting a lot of edits: I've rejected twice the number of edits I've approved (about 50 approved and 90 rejected), but what's important is that I very frequently use the reject+edit thing and write the reason why that edit is bad (like I hate it when people add superfluous links just for the 6 char limit, and I mention that in the edit log with a link to the rejected edit). I also use 'improve' frequently, which doesn't count to the number of edits I approved. You shouldn't portray such people as antagonists or something.
– user191954
Sep 24 '18 at 13:34