Is theoretical charop on-topic or off-topic?

Question

I just noticed that the tag wiki for optimization says

While practical CharOp, solving specific build problems for a specific character in a specific game is encouraged here at RPG.SE, general theoretical optimization questions tend to be broad guessing games seeking things like "the best DPR." and usually are more suited to forum discussions.

(emphasis mine)

That's not been the trend lately, so I wonder if that text should be removed as contrary to community practice, or whether we've gone astray and it should be enforced.

Then

For context, that language was added April 7, 2012 by Brian Ballsun-Stanton after we ran into confusion over how the tag should be used. (See How should we improve character optimization question tagging?, dated April 6, 2012.) This comment exchange appears to the the genesis of Brian's tag-wiki edit. (This is in no way calling Brian out for this! Just laying out the tag's history.)

However, after the tag wiki was edited, we had another discussion in August 2012 about what kinds of charop questions were on-topic, since we were getting a lot of poor ones at the time. That original discussion is at Are character optimization questions on topic?

Notably, I don't find direct support for that added language in the two answers to that discussion. The closest is this in the highly-upvoted answer by wax eagle:

The last kind of question that I see as useful is the challenge question. These are the most borderline Op questions IMO, but they can also be the most fun. (Examples 1 2 3)

That "challenge question" obviously meaning theoretical charop, by context and the examples provided. In mild contradition of what was earlier edited into the tag wiki, this argues that they're very borderline in being on-topic, but are also the most fun (implying they're worth keeping more-or-less firmly on-topic).

Now

Given the recent concerns about the quality of our 5e questions, and given how much discussion theoretical charop questions have been needing in comments before they can be answered, we should be saying more clearly than this whether they're OK or not. To that end we should fix the tag wiki to match actual community practice and good site governance, whatever that is.

So are these on topic or off? Or are they somewhere in the middle, with some being off topic and some being on topic, and the difference turning on what point?

Answer 1

I think the wording is fine. "General Theoretical Optimization" is and should be at the very least tread carefully around. Questions like "who is the best at DPR" is not a very good question, and probably shouldn't be justified with an answer.

However, there are sets of more specific theoretical optimization questions that are good questions and are squarely on topic here. For instance, the question that prompted this question, while broad, is quite answerable. Yes, it's hard, But it solves a quite actual problem (which should I play Wizard 10/Fighter 10 or Eldritch Knight 20). Just because it requires a bit of work, and a good bit of expertise doesn't mean it's a bad question. While it doesn't ask it directly, by asking it the way it does, it gives sufficient metrics for the OP to make their own decision. A good answer should at least provide summary builds of all three characters and show how they measure up. This isn't a particularly large burden with the ease of character generation in 5e.

While often poo-poo'd by certain members of this community, theoretical optimization is an arena of RPG study, and should be respected here. There are good questions to be asked there, and we have several of them (the speedy pixie is a good one, plenty of other ones). Yes, it can be a bit of a minefield here, but there is no reason that there be a blanket prohibition as advocated in the linked post.

Point being, if a question is answerable, and on topic, whether or not it's theoretical is besides the point. There are hundreds, if not thousands of questions on this site that don't solve an "problem someone actually faces," that's not a useful metric (and hasn't been for some time). The site's policies are reflective of this (no one checks...or at least I hope they don't).

If we wanted to raise a proper meta discussion about the question that prompted this, we should ask whether class comparisons are on topic (pretty sure they are).

Answer 2

It's on-topic, but not how we've been doing it

I am most swayed by Brian's comment in the wiki text that theoretical charop questions involve a lot of guessing games...

Who produces good theoretical charop?

Theoretical charop has historically been the result of many eyes turning a deep problem into a shallow one, over a long period of time, in back-and-forth discussions in forums that range across many threads and flamewars.

The end process is a community consensus on the topic derived from an understanding of the subject that is broad, deep, and distributed in such a way that errors and invalid working models are minimised, eliminated, or discredited and marginalised.

… Not us.

That's doesn't describe our site's operation at all.

Our home-grown theoretical charop that hasn't been grounded in work already done elsewhere has been pretty anemic. The answers tend to be posted, challenged in comments, get edited, and then iterate for a little while. And then they drop off the front page or get a checkmark... but are they right? History says "unlikely", because the analysis and assumptions haven't actually been put through the ringer the way a forum does.

We don't want the forum-like dynamic necessary to do good theoretical charop. But more than that, our existing attempts to do a forum-like iteration process to generate consensus has been ineffectual because the site's design makes it hard iterate. Instead of intense debate that burns away all invalid ideas and approaches, people vote and move on, challenging in comments only if they disagree really strongly. A few voters and the OP's checkmark end up blessing the "right" analysis, which is a shallow kind of consensus completely inferior to that produced by the wild west of forums.

And even if we could cultivate a strong iterative dynamic, such argument in comments over strongly-held views is contrary to our best practices. Every part of the process necessary to do good theoretical charop is anathema to the site's design.

So we kinda suck at theoretical charop, and aspiring to produce anything like the theoretical brilliance produced elsewhere for D&D 3.5e and 4e is unrealistic. Because we don't have enough eyes to make the problem shallow or a site model that allows the necessary discussion, our conclusions are suspect and probably make RPG.SE look bad.

But theoretical charop can still be on-topic

The subject is perfectly on-topic. There's a straightforward parallel to our relationship with lists though. Creating lists here is almost always off-topic (because our site model sucks at it) but we will report on other people's lists happily.

So too, we should limit ourselves to using external theoretical charop conclusions that have already withstood the test of the public's flames, and prohibit questions that ask for us to produce bespoke theoretical charop. We just can't ever be sure that an answer is actually complete, in the same way that a list might never be complete, and we can't be sure that it's correct, reducing voting to a popularity contest of which opinion "looks right" to us.

Answer 3

Theoretical CharOp should not be on topic here. It violates the point of the site ("solutions to a problem you have") and is more of a just-for-fun exercise as a "challenge." Without real use planned, it's very hard (as you're seeing from the thrash on those questions) to devise good success criteria.

I don't think these questions are valuable enough to have a heavily curated set of specific criteria (like we do only semi-successfully for game-recs). So if it's not a real problem, save it up for a monthly just-for-fun posting or something.