How A Random System Can Actually Be Predictable #veritasium #science #physics #math

“randomness” inside a constructed constraint system is not randomness in nature. A Galton board is an engineered object which artificially enforces repeated binary branching, symmetry, constraints, paths and board geometry. You’re looking at something designed to Gaussianize, not some natural phenomenon. Gaussian distributions merely reflect structural constraints, almost always put there by man.

Watch the full video on YouTube

What would happen if you let each balls separately, rather than all together?

In this toy the balls are not random. They bump against each other which influences the randomness

If you say a stock price is equally likely to go up 100% or down 100%, then it'll quickly drop to zero. But if it's equally likely to double or halve — i.e. go up 100% or down 50% — then you'll get a random walk than never reaches zero.

Actually:

Except that's incorrect, and you should in epistemic honesty do a second video on Mandelbrot because CLT empirically does not work with price returns due to the distribution.

@grok 😉

Attn Mighty @FairParentingUK 🤔🤫

utter BS !! RETARD!! 👇

That transition from order to chaos isn't random; it's governed by precise bifurcation points. We mapped how systems reach that edge of criticality before they tip into complexity. The second layer gets very deep 🌀 deepscroll > doomscroll 👇

Then it wouldn’t be random, would it…why are we still having this conversation?

The pachinko machine of life, lol.

What if the starting point is to the left or right?

@nntaleb may have, thoughts.

You can filter data in the exact same way for more predictable and quicker outcomes.