Hume's critique of causation

This is what makes a theory a theory. Just because we observe event B happening EVERY time after event A happens, doesn't mean that event B has to happen.

Example: if billiard ball A hits billiard B, billiard B will move.
The only reason we expect billiard ball B to move is because of precedent. We don't KNOW for sure if it will continue to do so in the future.

This is where I draw the line and say statistics & scientific method can be used. Just for the sake of having stuff to build up arguments with (and practical ones, for that matter).
Construct a hypothesis: If A, then B.
Test hypothesis: Let A happen in 100 independently run experiments.
If hypothesis is shown to be correct at least 95% of the time, it is statistically significant. If it happens 100% of the time, then you're even better off.

Note that this is NOT refuting Hume, it just posits a method for making conclusions using a different method. Course, some would argue that this departs from philosophy and enters the realm of science. After all, how do you construct 100 independently run experiments concerning the nature of reality?

I think there might be a way. Someone just has to be creative.

This sciencing up of philosophy is called "logical positivism," and is by no means original.