There is a "problem" with the type of theorization we do here. The "problem" has to do with determinism. Chemist, by comparison, have it easy. A chemical reaction should always held the same result. If it does not, we can check for contamination. Chemicals lack free will, they cannot "choose" not to interact as expected.

The b1tch with any human prediction is that humans have this pesky little thing called free will. While it is true that if we took a huge sample of humans, the behavior would appear somewhat determined. However, any given human is somewhat unpredictable. Human can just choose to not interact as expected for any or no reason at all. Keep this in mind. Just because something doesn't hold true doesn't mean it's not a good rule of thumb, and just because something holds true doesn't make it a good rule of thumb.

For every action there is a equal and opposite action. What we do is all a game. If you got skills you get the reaction you want.

it doesn't matter

there are a ton of scientific applications and areas of study you can't mathematically model to 100% accuracy (IE artificial intelligence, guided missiles, bioinformatics, stock market) but they're still useful even though they have a probability for failure.

most math models are just approximation anyway - taking the more important rules and ignoring the ones that hold less weight. if you wanted to mathematically model PUA concepts (why? I don't know), it would be an approximation

why does this bother you anyway?