Plotting Collatz orbits as points draws triangles whose sides give φ

Intuition

The wedge of triangles feels like a hidden coordinate system for the 3n+1 problem — the same orbits everyone plots as a tree, but the geometry reorganizes them into a self-similar tiling. The golden ratio falling out of the side lengths feels like the geometry is telling me something the arithmetic keeps hidden.

Parallels

It rhymes with the way φ appears in continued fractions and Fibonacci — the "most irrational" number showing up wherever a simple integer process settles into a fixed ratio. It is also the second time I have watched φ emerge from a plain dynamical map (the "golden map" bifurcation project is the other), which makes me suspect there is a common reason I am not seeing yet.

Self-Critique

The vertex formulas are something I found by fitting the pattern in the plot, not something I derived from the Collatz map itself — so the first thing a skeptic should ask is why those exact triangles are the right ones to read off the picture. Everything is empirical: the ratios are limits I observed, not theorems. The convergence to φ is only in the n→∞ limit; for small n the triangles are visibly not golden.

Hypotheses

A real answer would (a) justify, from the Collatz map, why the n-th triangle has exactly those vertices, (b) prove the side ratios tend to sqrt5, sqrt5/2, 1/2, and hence that (D2+D3)/D1 → φ, and (c) say whether this geometric φ is the same phenomenon as the golden-mean shift German found in the 3x+1 itineraries, or a separate coincidence.