Regarding your first point, that's not what I meant and you know it. Dark matter is the current theory explaining many different astronomical observations. It explains quite a lot of observations, so something pretty conclusive will need to happen for people to abandon it. Either new observations will disprove it or a new theory will come about that makes some prediction that can be observed that cannot for with current theories. I hope something like that happens! New science would be so much cooler than dark matter, but at the moment it doesn't seem likely.

I'm not sure I see how your second point is relevant. You seem to imply that I think that academics are infallible in all areas not even related to their expertise. Do you truly see no difference between non-statistics researchers incorrectly calculating statistical values and dozens of unrelated experiments all leading to the same result?

Flexibility is a main feature of the dark matter explanation, so it is relevant. It amounts to putting otherwise undetectable halos of mass wherever is needed to explain deviations from the model. There are further constraints people put on it for now, but none are crucial and will be quickly discarded (or more dark matter added) if required.

>"I'm not sure I see how your second point is relevant. You seem to imply that I think that academics are infallible in all areas not even related to their expertise. Do you truly see no difference between non-statistics researchers incorrectly calculating statistical values and dozens of unrelated experiments all leading to the same result?"

This is series of strawmen... I'm saying widespread confusion can, and currently does, exist on a topic even amongst the experts. Therefore it is not unbelievable that it does on other topics as well.

Also, dark matter is an interpretation of a result, not a result itself, so an experiment cannot lead to it directly. Finally, I don't think there are any actual experiments that have supported dark matter, only astronomical observations. Correct me if wrong but I found this after a quick search: https://www.sciencenews.org/blog/science-ticker/results-slew...

>"non-statistics researchers incorrectly calculating statistical values"

Actually another point is that this is not the issue. The calculations are fine, they are just calculating something other than they think: p(Data | Hypothesis) when they want p(Hypothesis | Data).

That is very similar to what the OP claimed about calculating a model that assumes a sphere to describe a disc (no idea if they are correct on that).