Kurotowa
Legend
Like calling frontliners "tanks", theorycrafting and DPR are imports from video games that mesh imperfectly with a TTRPG. When you're talking about fighting a raid boss in an MMORPG, theorycrafting an optimal build and rotation for maximum DPS is a legitimate goal. I mean, it's not a precise science because when you're not hitting a training dummy the encounter type and boss abilities matter, and you can't always use the mathematically optimal rotation because execution difficulty also matters when a human has to do it. But the conditions are controlled enough, and the randomness is limited enough, that it's a useful tool.Virtually all the theorycrafting is based on someone's very limited experience with their particular group and has then massaged the numbers to "prove" that this should be a problem at everyone's table and then will dig in and absolutely refuse to accept that the problems that they've "identified" through "math" are actually just self-justification and confirmation bias.
The fact that people will 100% argue that other people's actual play experience don't actually matter and then refuse to actually test their own math shows how entrenched the idea that you can simply mathematically model something as complex as an RPG and reduce it down to a couple of simple calculations.
Theorycrafting, like I said, can identify potential problems, but, is not proof of anything other than people's inbuilt, internalized biases..
Most of that all goes out the window with a TTRPG. The dice produce a wider range of results. The NPCs don't have to follow mechanically pre-scripted routines. The players too aren't limited to specifically allowed actions that aren't attacks. The encounter terrain is far more varied and interactable. Narrative input from all parties is encouraged, and can a substantial impact.
Is there are a place for theorycraft in TTRPGs? Yes, but only in broad strokes. It's helpful for pointing out glaring errors in class balance. It's good for turning up interesting new builds to try out. But it's not good for attempting rigorous optimization, and it's not useful when you're zooming in close for a difference of a few DPR.