Erdős Problem #958 — Configurations with consecutive distance multiplicities
For a finite set , let , let be the set of distances determined by unordered pairs of distinct points of , and let the multiplicity of a distance be the number of such pairs at distance . Is the following assertion false? For every and every finite set with , if and the set of distance multiplicities is , then consists either of equally spaced points on a line or of points on a circle with equally spaced angular increments. More explicitly, the line condition is that there exist with such that , and the circle condition is that there exist and with and such that .
References
Primary source
Additional references
Pinned Formal Conjectures source, Apache-2.0.
Progress summary
A recent paper gives a counterexample, so the proposed characterization is false even though the two standard families have the stated distance pattern.
Erdős conjectured that, for sufficiently large , the pattern with multiplicities forces the points to be equally spaced on a line or circle.
2025 counterexample
Clemen, Dumitrescu, and Liu construct, for every , a set consisting of the center of a unit circle and equally spaced points on an arc subtending less than . It is neither collinear nor cocircular, but its distance multiplicities are ; hence the equivalence is false. The paper also leaves open whether its displayed configurations exhaust all such sets for sufficiently large .
Current status (as of May 2025): The proposed equivalence is settled false by the Clemen–Dumitrescu–Liu counterexample; classification of all configurations with this multiplicity pattern remains open.
Solutions 0
No solutions have been posted yet.