Does the OP have to explicitly state the winning criteria?

Question

If a challenge is tagged with code-golf or fastest-code, do you really have to say that that is the winning criteria in the question? It seems pretty obvious, just by looking at the tags, right? After all, we don't say "this is a discussion" on questions tagged with discussion, do we?

However (just playing devil's advocate here), if you don't explicitly state how to chose a winner, people may not notice, leading to a whole lot of confusion as to why someone's post has been closed and did people not actually notice this and so on and so forth.

If a question looks something like this

Easy Prime Question

code-golf prime number

Write a program that outputs a prime number.

Does the obligatory "Shortest code wins" really have to be included, along with the code golf tag?

Answer 1

Yes, the winning criterion needs to be specified in the challenge

First, tags like code-golf where the winning criterion is obvious in theory. You're right that it's probably not entirely necessary, but especially with newer users it's often not clear whether they just picked a suggested tag without even being aware of the concept of winning criteria. To avoid that doubt, and so that people don't have to scan the tags to get the full specification of the challenge, the winning criterion should be mentioned in the challenge.

Regardless of obvious winning criteria like code-golf, something like fastest-code absolutely needs the winning criterion specified in the challenge. For the latter it's not clear just from the tag how exactly the challenge will be scored. Timing on a given test input? Largest test input that can be done within a given time? Who is going to measure the submissions how? Etc...

Answer 2

No, as long as the winning criterion is already clear from the tags

Most challenges on PPCG are code-golf. A lot of distraction, duplication, and possibly inconsistency is caused by having to re-explain in each of them what code-golf means. On the other hand, the probability and the cost of a new user mistakenly using the tag are small, as that is usually caught quickly and then the user learns and doesn't repeat the mistake (assuming good intentions).

Due to the nature of fastest-code, it does require an explanation of how performance will be measured.

In practice we don't have this problem with other tags, such as koth or cops-and-robbers. They are rare and authors tend to explain the winning criterion well.