This could happen if the question is of the kind where “Don't do it” is a valid answer. Per se, there's nothing criminal in questions like this - one asks when one doesn't know the answer, nothing wrong with this. Even more, questions like this can be useful - these can help someone else pondering particular bad idea find that it's the wrong way.
Now, how could valid and useful question get downwoted? Easy-peasy. Since asker doesn't know, they may think it's actually a good idea and phrase their question in the way that promotes it.
Stubborn reviewers keep complaining about this method: float multiply(float x, float y) { return x / y; }
- how do I convince them to fork off of my code?
If the idea is bad enough, question could get downvoted pretty heavily. Thing is, even readers not knowing how to explain what's wrong there, may feel that it is bad idea and feel the need to somehow signal about this using downvotes.
Now, "bad idea" question gets a couple of downvotes and few answers explaining that it's the wrong way.
Eventually, asker "gets it", figures the right way and accepts an answer that explains it.
Reviewers are right, this code needs fix. To use method like this, one has to permanently keep in mind that its result differs from what one would expect, which is very cumbersome and error prone.
Would this help for the question to stop collecting downvotes? Unlikely, and here is why - readers do not perceive "accepted" checkmark as part of the question itself (which is not entirely wrong by the way since asker can unaccept any time they want). To reader, the question still reads as promoting the wrong way, and same reasons for downvotes as explained above still apply.
To stop downvotes one better rephrases the question to more neutral tone.
This code popped up at review: float multiply(float x, float y) { return x/y; }
- why is it considered bad?
In a perfect world, above would suffice to revert the voting direction. However, if the question is already heavily downvoted it can continue to "sink", although there are no objective reasons for that anymore. Thing is, highly voted posts have a certain appeal that kind of makes you compulsory follow the "majority vote". 1 Despite the edit, question may get downvoted by readers who just follow pack mentality.
If "unfair downvoting" happens to question after edit, the most reasonable course of action is probably to bring it on meta. Meta readers could vote up the question to restore balance, or help to convert question to CW if "pack voting" went off-hand.
For the sake of completeness, askers also have an option to flag their question for mod attention asking to convert it to CW. This is not the way I would personally recommend because I think cases like this are better fit for meta discussion than for communication over simple "flagging interface".
- Think of it, how would one phrase flag message? - Please convert to CW. - Why? - Please convert to CW because I don't want to continue losing reputation? - Everyone would want to, what makes you special? - Please convert to CW because I don't want to continue losing rep for the question where I changed my mind. - Why don't you just bring it on meta? You see, simple explanations carry risk of flag decline. And detailed enough explanation would make a reasonably good meta question anyway.