Circular drawing characterization via forbidden type

Open Question on Circular Drawing Characterisation · arXiv:2601.12746

arXiv Informal medium confidence— first stated 2026-01-19

Authors: Eli Berger, Daniel Carter, Paul Seymour

Status open high confidence

The paper by Berger, Carter, and Seymour (arXiv:2601.12746, January 2026) fully characterises weightable digraphs and derives a polynomial-time algorithm for testing weightability, but explicitly leaves open the question of whether there is a forbidden-structure characterisation of digraphs that admit a circular drawing. No follow-up work addressing this specific open question was found in the four months since posting.

Reviewer notes. The source paper is very recent (January 2026). The open question asks for a Thomassen-style forbidden-substructure theorem that characterises circular drawings of sufficiently connected digraphs. The authors explicitly state they could not find such a characterisation despite fully resolving the weightability question. No follow-up was found after three targeted web searches.

Auto-reviewed 2026-05-14 with claude-sonnet-4-6 (web search enabled).

Informal. Is there a theorem that says ``every appropriately connected digraph $G$ contains no thing of type $X$ if and only if $G$ admits a circular drawing''?

Context

Thomassen's theorem gives a characterisation of circular drawings in the presence of a specific 2-vertex separator. The authors generalise this into a motivating question and, despite fully characterising weightable digraphs in the paper, explicitly state: ``We still have not come up with the characterization we hoped for of the digraphs with circular drawings.''

Notes. No labelled theorem environment; stated as a motivating open question in the introduction and explicitly left unresolved at the end of the introductory discussion.

Source paper

When all directed cycles have the same weight
Eli Berger, Daniel Carter, Paul Seymour · 2026-01-19
https://arxiv.org/abs/2601.12746 PDF source