Elliptic curves are mathematical objects central to modern cryptography, and two elliptic curves can be connected by special maps called isogenies. When you take all supersingular elliptic curves over a finite field and draw edges between curves connected by an isogeny of a fixed degree, you get what mathematicians call a supersingular isogeny graph. These graphs have remarkable properties, including being very well connected, which is why they have been studied as the basis for certain post-quantum cryptographic protocols. Within this graph lives a special subgraph called the spine, which consists only of the curves and connections that are defined over the smallest relevant finite field rather than a larger extension of it. Understanding the spine matters both for the mathematics and for assessing the security of cryptographic systems built on these graphs.
A previous paper had worked out the structure of the spine in most cases but got stuck on a particular combination of parameters: isogenies of degree 2 and primes of specific forms (those that leave a remainder of 71 or 119 when divided by 120). This paper tackles exactly those leftover cases. The authors carefully analyze the geometric and combinatorial shape of the spine's connected components by computing distances between nodes, how far each node is from the farthest point it can reach (eccentricity), and the overall diameter of each component. A key technical move is to temporarily remove one particular edge that behaves differently from the others, analyze what remains, then account for that edge's effect separately.
By working out these distance and shape measurements precisely, the authors can compute a kind of average diameter for the spine and show that this single number reliably distinguishes between the different structural types the spine can take. This gives a practical diagnostic tool: rather than mapping out the entire spine by hand for a new prime, you can compute the mean diameter and immediately know which structural case you are dealing with. The authors use this to build heuristics, meaning informed probabilistic predictions, about how the spine's structure changes as the prime number varies, providing a clearer overall picture of this previously murky corner of the theory.