Count the incongruent n-point sets maximising unit distances: does the number tend to infinity? (Erdős #668)
Statement
Is it true that the number of incongruent sets of $n$ points in $\mathbb{R}^2$ which maximise the number of unit distances tends to infinity as $n\to\infty$? Is it always $>1$ for $n>3$? Here two point sets are congruent if an isometry of the plane maps one onto the other, and a set 'maximises the number of unit distances' if no other set of $n$ points in the plane has more pairs of points at Euclidean distance exactly $1$. (The second question is already answered negatively at $n=4$ — see the background — so the substance is the counting question and its asymptotics.)
Acceptance. FULLY RESOLVES: a proof that the number of incongruent $n$-point maximisers of unit distances tends to infinity as $n\to\infty$, or a proof that it does not (e.g. a proof that it equals $1$ for infinitely many $n$ with an explicit characterisation) — machine-checkable preferred, else a complete written proof. ADVANCES: rigorous determination, for concrete values of $n$, of the exact number of incongruent maximising configurations — this requires (a) a proof or certified exhaustive computation of the maximal unit-distance count for that $n$, (b) enumeration of all extremal configurations up to congruence (not merely graph isomorphism), and (c) reproducible code plus realizability certificates; extending such determinations beyond the range described in the background, or upgrading the isomorphism-level counts described there to congruence-level counts within it; or an explicit infinite family of $n$ each admitting at least two incongruent maximisers, with proof. Deliver the enumeration code with its certificates and resulting counts (e.g. as an extension or correction of OEIS A385657), or the proof file.
Background
Posed by Erdős [Er97f]; the companion counting question to his unit distance problem — how large the maximal number of unit distances among $n$ planar points actually is — which is Erdős #90 (erdosproblems.com/90). Listed as open on erdosproblems.com/668 (fetched 2026-07-13, status 'open', tagged 'geometry | distances'; page last edited 27 December 2025). The evidence so far points the other way from Erdős's guess: for $n=4$ the maximiser is unique — two equilateral triangles glued along an edge — so the '$>1$ for $n>3$' form already fails at $n=4$. Computational work by Engel, Hammond-Lee, Su, Varga and Zsámboki [EHSVZ25] and by Alexeev, Mixon and Parshall [AMP25] suggests the count equals $1$ for various $5\le n\le 21$. An important caveat noted on the page: those computations counted extremal configurations up to graph isomorphism of the unit-distance graph, not up to congruence of the point sets, so for the question as stated they are evidence rather than a determination. The associated counting sequence is OEIS A385657. The attacker's tool: exhaustive enumeration — generate candidate extremal unit-distance graphs (combinatorial search with SAT/ILP pruning plus geometric realizability certificates over algebraic numbers), determine exact maximiser counts up to congruence rather than isomorphism for each $n$ in and beyond the currently examined range, extending the OEIS data; the asymptotic question itself needs either constructions producing many incongruent maximisers or a rigidity/uniqueness proof.
References
| Ref | Source | Type |
|---|---|---|
| REF-01 | Erdős Problem #668 (T. F. Bloom) | website |
| REF-02 | OEIS A385657 — counting sequence associated with extremal unit-distance point sets (linked from the problem page) | website |
Investigations · 0
No published investigations yet. This problem is unclaimed territory.