Failed audit for trying to "improve" it

Question

I was trying to edit away the "Thank you" when I came across this audit. Obviously it was an audit, but I figured that improving it would not mark it as incorrect. Apparently, improving counts as a failure.

I would like to propose that improving an audit should not fail the audit, but allow you to pass.

Answer 1

I disagree.

The specific audit you referenced is laughably wrong. If you had even looked at the suggested edit, you wouldn't have been compelled to hit "Improve", but instead, "Reject".

Thank mfm filmmaker you

mfm filmmaker?

Seriously?

And you wanted to improve this edit, when the thing the edit was doing was not only hilarious, but completely the opposite from what it should've been doing?

Lesson learned (I hope!): if you're reviewing a suggested edit, before you click a button, make sure you actually read the suggested edit. The audits are designed to be hilarious and obvious to any human being, mainly with the purpose of defeating automated tools or "mindless" clicking of random buttons. Even if you have no idea what kind of subject matter is discussed on SuperUser, this should still be obvious to you.

I can't support changing the system to be more lenient when this is the only case I've heard of where a legitimate reviewer tripped over clicking the "Improve" button by mistake.

Here is the procedure you should follow when you review edits:

1. Glance over the things that were subtracted from and added to the post, highlighted in red and green, respectively.
2. For each thing subtracted and added, make a judgment in your head if those things make sense within the context of the post.
3. If you start seeing repeated instances of things that are totally nonsensical, you should just reject the edit outright, regardless of whether one little piece of the edit might be plausible. Don't try to salvage little bits of valid information from a 90% bogus edit. If the edit makes you or laugh, or go "huh?!", you need to reject it.
4. If you want to improve the post anyway as part of your review, just click the link to open the post, separately from the review queue, and edit it directly in there. Just go through the post line by line and make any corrections you see fit. Don't worry about the edits that were made in the bogus suggested edit.

In summary, you should have very little tolerance for incorrect, nonsensical, or off-topic additions or subtractions in suggested edits. Even if the post legitimately needs to be revised, and the suggested edit does something to improve it, you should reject the edit as soon as you see something that is clearly incorrect, nonsensical, or off-topic. If you're feeling up to the task, you may then, at your option, perform a proper edit on the post itself.

The only time when it makes sense to click the "Improve" button when reviewing a suggested edit, is when the vast majority of the suggested edit's content is totally fine, but there's one TINY little mistake in there. For instance, if the suggested edit adds four sentences, fixes five typos and (whoops!) accidentally adds one new typo in the new prose, you could click "Improve" to fix that typo, after determining that the changes in the edit -- the new sentences and the typo fixes -- are valid, warranted, and make sense within the context of the post. Context is key.

And anyway, a single failed review audit isn't the end of the world; you will continue to be able to review, as long as you don't repeatedly fail audits over and over. You should not worry about this as long as this is more of an aberration than the norm.