Second Edition, totally revised, new answer replacing a very highly upvoted one that doesn't answer the question as well

Question

Here's a shocker, the answer involved concerns avocados.

This Q&A has obviously been very rewarding: Browning Avocados - What Helps?

The answers to that question other than my own caused me to want to do further experimentation. I particularly wanted to explore ascorbic acid (Vitamin C) and propanethiol S-oxide gas from diced onions. I did those experiments with the same kind of photographic record keeping as the experiments in the original answer.

I couldn't be more pleased with the results of the second bout of experimentation, because what I tried, worked.

So now I have a real answer to the question, "What Helps?" What has been in that slot has been so highly upvoted because it was a good experiment with rather shocking results, not because it really answered the question. That answer says quite well what doesn't work. (Being on the "Hot List" for quite a while didn't hurt)

So my inclination is to replace the older, highly upvoted answer with the new (IMO better) answer, and move the old answer to a new question, "Do lemon juice, lime juice, vinegar or Fruit Fresh help keep avocado green?"

Early in the new answer, I will link to the old one and point out that all edits are transparent.

Now, to do it that way is not going to give me a lot of rep. Most regulars here have already upvoted that slot. So the only way they can upvote the new answer is to actually go to the old one and upvote it. That's not going to happen in droves. But I really don't care, I care far more that I'm treating the results in the best possible way for them to be found by future users.

My biggest concern is that some users won't necessarily think that through, and will see something cheesy in replacing such a highly upvoted answer with an entirely different answer.

So I bring this here for your input. Is there a better way? If so, I'd love to hear it. If the community is supportive of my handling it this way, I have the benefit of having someplace to send people who might think I'm trying to pull something cheesy.

If you want to see of a draft of the new answer, you can find it (with a few edits) in chat, here: http://chat.stackexchange.com/transcript/message/19422570#19422570

Answer 1

Bottom line, I think you are trying to turn "acid doesn't work" into "X Y and Z work, acid doesn't", and that's a strict improvement, so you should just edit - I would if it were my answer.

There are two groups of people to serve here: future readers and past voters. It's clearly best for the future readers if the best information is in the right place, so on behalf of them we should choose editing. For past voters, yes, hypothetically they might want to retract a vote. But honestly, given that they voted for the old answer, they'd surely vote for the new answer too. If a minority would seriously change their mind, then they should be checking all the posts they've ever voted on all the time, because people are editing those too. We shouldn't cater to them over our future readers, who are really the entire reason we're writing all these posts.

I would be sure to still convey all the information that the old answer conveyed. That doesn't mean much of the text has to survive; it could be heavily summarized. But that takes care of people who previously voted: the information they valued is still present. One thing to keep in mind: the overall tl;dr at the top should cover the material from both old and new answers, to make sure you haven't lost anything even for the impatient readers.

I think Aaronut's points are all in general valid, but on the other hand this is a specific case with known details, so I don't think we really have to worry so much about potentially offending previous voters. Sure, it's possible, but quite honestly, if people liked your first answer they're going to like the new one. And even if a few people for some reason don't, making their votes less accurate is a small price to pay for presenting information to future readers in a clean way.

Answer 2

We usually recommend not doing this, and simply posting a new answer. Updating the existing answer is fine if it's a minor edit (maybe adding a few lines).

Remember, people upvoted the answer, not you. That means that what you originally posted is what they found to be useful. Changing it to something entirely different is not really in the spirit of how our edit system is supposed to work. It is essentially a bait-and-switch, however well-intentioned it may be.

Some exceptions have been made in the past for answers that turned out to be completely wrong, even to the point of being dangerous, but those are few and far between.

Of course it's your answer, you can edit it however you want. IMO it's just not totally fair to the people who upvoted; and some users would be justified in undoing their upvotes (which is always allowed after a post is edited).

Answer 3

Jolenealaska, would it be an option to edit the original question to better fit the answer? For example, 'Does lime juice, lemon juice, or vinegar keep avocado from browning?'.

This would maintain the integrity of the original question and answer. Then you could form a new question to relate more to your latest experiment.

You could even refer to the previous experiment/Q-A by staring your answer with something like 'In the first experiment I tried lime juice, lemon juice,.....'.

Just an idea as I am not as familiar with all of the intricacies or how something may be perceived. Perhaps @rumtscho, @Jefromi, and @Aaronut could give some feedback on this idea.