Erdős Problem #982 — Distinct distances from a vertex of a convex polygon
For an integer , let be distinct points that are the vertices of a convex -gon. For each vertex define its number of distinct distances to the other vertices by
The question is whether, for every such -gon, at least one vertex has at least distinct distances; equivalently, whether
holds for every and every set of vertices in convex position. Erdős Problems #982 Kominers (2026 preprint)
Background
This is a classical conjecture attributed to Erdős and associated with his work on distance sets; the problem bibliography includes his 1946 paper and the later joint paper with Fishburn, “A postscript on distances in convex -gons” (1994). The regular -gon shows why the proposed threshold is natural and sharp: each vertex determines exactly distances, one for each possible separation around the polygon. Erdős Problems #982 Kominers (2026 preprint) Erdős (1946) Erdős and Fishburn (1994)
A useful route to lower bounds counts good edges: an edge is good when its perpendicular bisector contains at most one point of the configuration. Such an edge can be the base of at most one isosceles triangle with another point as apex. Double-counting these incidences converts a quadratic lower bound on the number of good edges into a linear lower bound for the number of distinct distances from some vertex. In the convex setting, the smallest enclosing circle and its at most three support points divide the polygon into arcs, or caps, where distances from either end vary monotonically; counting pairs within and across these caps is the key geometric input. Kominers (2026 preprint)
Convexity is consequential. The half-bound does not extend to all planar sets with no three collinear: the preprint describes an eight-point configuration, consisting of two concentric squares in a specified relative position, for which every point has exactly three distinct distances, fewer than . Kominers (2026 preprint)
Progress summary
Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.
Solutions 0
No solutions have been posted yet.