Collatz Triangles: why does φ appear in the triangles the orbits draw?

) iteration is usually visualized as the Ulam spiral or as a branching tree. This paper takes a different view: each orbit is plotted as a set of coordinate points 

, and the orbits of all starting numbers up to 

 are overlaid. The resulting figure fills a wedge that is tiled by a sequence of triangles. Closed forms are given for the vertices of the 

-th triangle, and hence for its three side lengths 

. The ratios of areas and perimeters of consecutive triangles tend to 

, and the side ratios tend to the clean limits 

. The main observation is that the combination 

). The evidence is computational and no proof is offered; a related golden-ratio structure in the 

 itineraries is noted (German, 2004). The phenomena, the constructions behind them, and the questions a proof would answer are recorded.

I decided to look at the Collatz Conjecture from a different angle. Instead of visualizing the iteration as the famous Ulam spiral, I wanted to treat the iteration as a group of points. The Collatz map sends 

 is odd; the conjecture is that every starting number eventually reaches 

The idea is simple: take the orbit of a starting number and plot its consecutive terms as coordinates. For the starting number 

   x        y
   3       10
  10        5
   5       16
  16        8
   8        4
   4        2
   2        1

I acknowledge that plotting Collatz orbits as points is not new — this kind of figure has been reproduced before by other means, and I reference the paper it appears in at the end. What follows is my own exploration of the geometry it produces.

I did this procedure for all starting numbers up to 

 using Desmos (the interactive plot is at https://www.desmos.com/calculator/wusydxtkkv). Overlaying the orbits fills a wedge in the plane.

Figure 1. The orbits of many starting numbers, each plotted as points 

Figure 2. A second view of the same plot after adding more starting numbers.

Zooming out shows the global shape — a wedge tiled by a row of triangles that grow along the diagonal.

Figure 3. A zoom-out of the plot. The orbits fill a wedge that is tiled by a sequence of triangles.

This document is an experimental report: it records the observation, the closed forms behind it, and the analysis I have so far. It does not contain a proof, and it is honest about that.

I wanted to investigate the area of the triangles formed. At first I used Heron's formula, but I knew that would be cumbersome, so I devised a plan to find expressions for the vertices and the sides of the triangles directly.

 Reading the triangles off the plot, I found that the vertices of the 

 gives Triangle 1, and so on. The figure below shows a single triangle with its vertices and the three sides 

 to the vertices gives the three sides in closed form:

. The triangles are easiest to see in a zoom on the first two, Triangle 0 and Triangle 1:

Figure 5. The triangles are more visible here. The first is Triangle 0 and the second is Triangle 1.

Note that nothing about the golden ratio has been used anywhere in this construction — it is built only from the Collatz orbits, the plotted points, and the distance formula. Where 

 appears, it appears afterward, when the side lengths are compared.

With closed forms for the sides I looked at how the triangles behave as 

Consecutive triangles converge to one shape.

 The ratio of the areas of consecutive triangles, and the ratio of their perimeters, both tend to 

. So the triangles do not wander through different shapes — they settle onto a single limiting form.

 In the same limit the three side ratios approach simple values:

, the limiting triangle is right-angled, with legs in the ratio 

 — consistent with the right triangle drawn in Figure 4.

 The surprise is what a particular combination of the sides does. Forming 

. So the golden ratio and its reciprocal sit on the two ends of this one side-combination — inside a construction that began with nothing more than plotting 

I am still not sure how the Collatz iteration is connected to the golden ratio. The appearance is clean and reproducible, but it is empirical: the vertex formulas were found by matching the pattern in the plot rather than derived from the Collatz map, and the ratios are observed limits, not theorems.

There is one reassuring piece of outside evidence. After finding this, I came across a paper that generates a golden-ratio structure from the 

 map by a completely different route — symbolic dynamics rather than plane geometry:

German, Dan-Adrian. "The Golden Mean Shift is the Set of 

 NKS 2004 conference. http://www.wolframscience.com/conference/2004/presentations/material/adriangerman.pdf

German shows that the set of Collatz itineraries (the even/odd symbol sequences the orbits produce) is exactly the *golden-mean shift* — the classic subshift of finite type whose forbidden word is "

. That the golden ratio also governs the itineraries makes me doubt that the geometric appearance here is a coincidence, even though I cannot yet connect the two pictures.

Honestly, though, the two could still be unrelated. The golden-mean shift is about which symbol sequences occur; my triangles are about the metric geometry of the plotted points. Bridging them — showing the 

 entropy are the same fact wearing two costumes — is exactly the part I cannot do.

Plotting Collatz orbits as points, rather than as a spiral or a tree, exposes a wedge tiled by triangles whose side lengths carry the golden ratio: 

. The evidence is computational and every appearance of 

 here is currently empirical. A satisfying account would answer three questions.

 Derive, from the Collatz map itself, why the 

-th triangle in the point plot has vertices 

 — rather than reading them off the figure.

 Show rigorously that the side ratios tend to 

 Decide whether this geometric golden ratio is the same phenomenon as the golden-mean shift of the 

 itineraries (German, 2004), or an independent coincidence.

Question 2 is the concrete, self-contained target — it is elementary algebra once the vertices are granted, and it is the natural first result. The project's bridge request asks for exactly this. If you know symbolic dynamics, question 3 is the deeper and more interesting one.

Collatz Triangles: why does φ appear in the triangles the orbits draw?

Abstract

Introduction

Body

Analysis

Results

Proofs

Conclusion