Is there a guidance about that?
In particular can 3 comments discussing a single answer already qualify a debate?
This is useful to know, because some questions are closed as not-constructive when they are likely to be dabated, even if (to my untrained eye) they look fine: very few answers, very few comments, on topic, no more.
So I'm trying to understand the measures (or the logic) used to define (or to forecast) a "debate", in the hope to avoid it at stackoverflow.
A "debate" isn't some property that's measured by a certain number of something. A "debate" is simply a discussion of differing and/or opposing views about a certain topic, which on our site typically happens in the comments for some reason.
You have to read the comments in order to determine if the commentators are debating or not. You can't use the number of comments posted to figure out if there's a debate going on.
For what it's worth, the comments under your question or my answer may or may not be an example of a debate. It depends largely on how you interpret it.
There seems to be some confusion about when questions will be closed as "not constructive" and whether debate over certain things (such as the validity of an answer) would qualify a question for closure as not constructive.
The full text when voting to close as "not constructive" is this:
As it currently stands, this question is not a good fit for our Q&A format. We expect answers to be supported by facts, references, or specific expertise, but this question will likely solicit debate, arguments, polling, or extended discussion. If you feel that this question can be improved and possibly reopened, see the FAQ for guidance.
Voting for closure as "not constructive" happens, among other reasons, when a question is asking a question that doesn't have a single, verifiable correct answer (usually such questions are asking for opinions). If I was to ask "Is C++ better than Java?" then that's going to solicit debate; people who prefer Java are going to post that Java is better, people who prefer C++ are going to post that C++ is better. Then they're (probably) going to go back and forth in a (vain) attempt to prove to the other people that they're correct.
That debate may never happen. In fact the point of closing the question as "not constructive" is to prevent it from happening.
However, it's possible for debate about a question to happen without the question itself being considered "not constructive"; I've seen questions about certain aspects of web development trigger a debate about whether they'd be considered "front end" or "back end". I also regularly "debate" with other users about the validity of answers to questions. Neither of those debates have any bearing on whether the question itself is "not constructive", though.
To sum up: There isn't a useful metric to determine whether a debate is occurring, and debate doesn't necessarily mean that a question is "not constructive".
It's not about whether a debate will happen; it's about whether the subject of the question is debatable. And that's not particularly subjective.
What is an anti-pattern is a matter of debate. People argue these all the time. Some people think X is an anti-pattern, others think Y is one. And even if most programmers agree that Z is one, it's still a matter of debate. It's still subjective.
Also, your question is a list question. So it's double not-constructive.
Closing posts as not constructive is not based on an actual count of comments, but on the way the question is phrased.
As is already answered in your question - and as Andrew Barber says in one of the not-accepted answers, the question is not whether it has resulted in a debate. The question is whether it is likely to result in a debate.
