Sports Statistics and topicality

Question

The question below was closed, and probably correctly according to our FAQ (although I'm not sure what "source" exactly means). I'd like to open a discussion over whether questions like this should be on topic.

---

Some background (skip to the end if this is too long!):

Statistics in sports is a very hot topic right now; FiveThirtyEight, Football Outsiders, and their ilk are some of the most popular sports sites this side of, well, ESPN. Assuming this site intends to cover sports-as-fans and not just sports-as-participants (which is a valid question, I think!), it seems obvious that some discussions of statistics must be on topic here; and certainly some questions are, such as this question about dWAR which wasn't really a very well written question but still was certainly on-topic and generated one very good answer.

On the other hand, questions about specific statistics - "who holds the record for..." - are not really great on topic questions, as this isn't really a very good place to host things like that - although we've tended not to shut them down, either, and the harder ones seem like reasonable questions. However, in particular ones that might change frequently tend to be worse fits here, because of that fact - the answer becomes out of date but still is in our database and someone has to realize it's out of date to fix it.

---

This leaves, however, questions like the one that started this: questions about sources of information. These fall into two broad categories:

1. Questions about finding a source for a specific statistic, such as "I remember a game back in 1956 where Ernie Banks pitched in the seventh inning."
2. Questions about how to, in general, do research, including good sources for information.

1 should be off-topic as it's written; StackExchange isn't about asking someone to do research for you, it's about tapping people's already existing knowledge. However, it can be rewritten:

How do you identify a specific game that a position player pitched in the 1950s?

That is a very specific question, and should be on-topic. (The 1950s is important, as different eras have different kinds of information available.) Part of the answer to that question, though, would lead us to 2.

2 is basically what the initial question from this discussion is about: about how one can do research. It seems to me that the general topic - of how to do research - should clearly be on-topic; questions about sources for research are however more of a question.

Should we change the rules (or our interpretation) to allow questions about where to find statistics to be on topic? And if so, precisely where is the line drawn?

Answer 1

I think that yes, within limits, questions like this should be on topic.

Questions that ask where to find certain kinds of information - whether that be like the question at hand here, asking where to find salary information - are of sufficient scope; they don't necessarily require research (as some of us know them already); and they're useful to others performing the same research.

While they do have a degree of timeliness to them - certainly 50 years from now most of these won't have the same answer, though some will - many of these sites do have a relatively long term utility, and would be similar to asking on our Genealogy site how to find a particular kind of information.

As such, as long as the question is of sufficient scope to not be just one piece of information, but rather one kind of information, and is not asking for an opinion on quality of different sites, it should be on topic. Questions asking for "best" must be asking from the point of view of something concrete - "most complete" by some particular definition, for example . Questions asking for one piece of information - "how many home runs did Ernie Banks hit in 1956", "Who won the world series in 1980" - are off-topic.

Answer 2

I think Joe's analogy with the librarian is a good one, not sure how we can formulate and enforce it in more general terms though.

Also I have previously asked another question asking for ways to find a specific statistic, which is among the top 50 most viewed questions we have on the site.

I think we should be careful in not limiting the site too much, it might end up becoming a site for "explain this rule for me", to which the only feasible answer is to take a citation out of the rulebook.