# How to deal with an accidental chameleon challenge

## What happened

I was about to post Does base n have any Rotate-Left-Double numbers?. Kevin Cruijssen told me in the sandbox comment that he got a proof that significantly simplifies the problem. I posted it to main, excited to see the proof. (Seriously, I should've reconsidered posting to main at this point; it's my fault.) Then Kevin posted the proof along with a 6-byte 05AB1E answer

So essentially, the challenge text is asking for

Given the base $$\n≥2\$$, determine if there exists a Rotate-Left-Double number in base $$\n\$$.

Given a positive integer $$\n≥2\$$, determine if it is not a member of A056469 (the number 2 and the numbers in the form of $$\2^k+2\$$).

which definitely sounds easier, and has a super-simple formula n-2&n-3 (as pointed by xnor).

## The question: How to resolve this situation?

It is not that uncommon to see that a challenge asking for a complex mathematical property boils down to a very simple task, sometimes so much that having such a challenge is pointless. If we identified that in the sandbox, we can redesign the challenge to something different. (Again, this particular incident is my fault posting it to main after the simplification is found.) However, it might slip through the minds of both the challenge writer and sandbox reviewers, and get found by someone else shortly after posting.

What can I do about it?

I can think of a few possibilities:

• Do nothing.
• Close as a dupe (if applicable). Otherwise, post a new challenge clearly asking for the actual task, and close the OP as a dupe of the new one.
• Close with a custom reason, and refine it again in the sandbox.

I don't want to get the challenge deleted, because there is at least one answer whose author put a lot of effort to prove such a property.

• Not that it super affects the content of your question but your definition of chameleon challenge seems to differ signfigantly from the one I am familiar with. I don't think I would call what this is asking about a chameleon challenge. It might be wise to remove this sort of wording in the question as it gives an impression of a somewhat different scenario.
– Wheat Wizard Mod
Commented Apr 2, 2020 at 17:01
• @AdHocGarfHunter To quote your linked post: "Chameleon challenges look like they're about one thing but are really about another." Doesn't it agree with my definition? It's just that this case is different from other kinds of chameleon challenges (e.g. cumbersome I/O or input validation, which are highly visible) and the symptom isn't visible until an answerer posts a simplification. Commented Apr 3, 2020 at 0:49
• I would certainly call it a pretty central example of a chameleon challenge.
– xnor
Commented Apr 3, 2020 at 1:14
• @Bubbler I feel like you have been omitting the relevant part of the definition. The next sentence is "In doing the challenge, most of the effort is spent on something peripheral." which does seem to be the case for the challenge in question. Either way I think the chameleon challenge wording is detrimental to this question.
– Wheat Wizard Mod
Commented Apr 3, 2020 at 1:57
• @AdHocGarfHunter I could remove the phrase from the body, but I don't see how I could rephrase the title without making it way too long... Commented Apr 3, 2020 at 2:30
• @AdHocGarfHunter I'm open to suggestions on changing the chameleon challenges thing-to-avoid to make it clearer. I definitely did intend it to apply to situations like this, and if that doesn't come across, then I think I did a bad job there.
– xnor
Commented Apr 3, 2020 at 2:45
• @xnor I think having the distinction Garf is talking about is important. This kind of challenge is really more of an "accidental" unnecessary fluff challenge. Consider your challenge about fruit pies: I do not consider this a bad challenge at all, despite the explanation being similarly misleading about the optimal approach as in this question. The main difference here is that the resulting simplification is too simple to be an interesting challenge on this site. Commented Apr 3, 2020 at 15:53
• It's common for a function defined in a complicated way to be identical to one defined more simply. But spotting which simple function works, and proving the identity, might be hard. True, once anyone's posted an implementation of the simple function, others can do likewise. But isn't it always true that, once anyone's posted an implementation of an algorithm, anyone else may post another implementation of the same? Commented Aug 19, 2020 at 8:36