To predict such movement you would have to be aware of all properties of involved molecules. I wonder if the Heisenberg's uncertainty principle would demonstrate such detailed knowledge is not possible.

How does not predictable mean not random?

Some studies on Brownian motion: [aip.scitation.org]

Not quite as predictable. Complex Systems (Chaos) Theorist have some algorithms that do a pretty good job. But it is not so predictable that you can create an exact repitition of the previous model, molecule by molecule. Enough randomness to always be different.

You're confused with the terms complicated and complex:
Building a helicopter from parts is complicated but doable i.e. predictable
Something complex is very difficult to predict the outcome
A problem is termed complex when it has interrelated, interdependent parts and has features and parts that are constantly changing and unpredictable.

[differencebetween.com]

I believe, not certain but Einstein showed Brownian movement to diffuse in the space or volume it was contained in. Though chaotic the overall effect is we can predict the thermal movement and that is fast.

You're missing a lot actually....... pool table events are macroscopic, therefore predictable. Particles are so mind boggling small that nature behaves differently at those tiny levels. In other words, you are not comparing apples to apples here. Now, on a serious question though.... are you serious?

Chaos theory?