The multipartite orientation conjecture for the multiplicity of
The multipartite orientation conjecture for the multiplicity of
Let be an orientation of a complete multipartite graph, and let be the number of its parts having odd size. Write for the Eulerian polynomial of , and let denote the multiplicity of as a root of this polynomial. Multipartite orientation conjecture. One has
This conjecture would generalize the paper's result that every -vertex tournament has as a root of multiplicity exactly . Its status is not resolved in the supplied text.
Progress summary
A posted argument claims a complete proof of the conjecture, but no independent verification has been found.
Celano, Sieger, and Spiro (2023) conjectured that the multiplicity of in the Eulerian polynomial of every orientation of a complete multipartite graph equals , where is the number of odd-sized parts.
Known results
- Celano, Sieger, and Spiro (2023): for every tournament on vertices, the multiplicity is exactly .
Posted attempt
An argument claims a complete proof for every orientation and all part sizes, using a pairing expansion at and parity of minimal pairings; it also claims an exact -adic valuation for the first nonzero coefficient. The argument has not been independently verified.
Current status (as of August 2026): the tournament case is established, while the multipartite conjecture has a complete-proof claim but no independent verification or corroborating publication.
Sources
Sources & referencesView supporting material
Primary source
Kyle Celano, Nicholas Sieger and Sam Spiro, “Eulerian Polynomials for Digraphs”, arXiv:2309.07240 (2023).
Solutions 1
Sign in to submit a solution.
Proof for every orientation, with the exact 2-adic valuation of the first nonzero coefficient.
Let , let be an arbitrary orientation, and put
Write
Partition each ordering into consecutive unordered two-vertex blocks, followed by a final singleton when is odd. For an ordered block configuration , let count blocks whose vertices lie in different multipartite parts, and let count descents on edges between distinct blocks. Summing over the two internal orders of every block gives the exact identity
Indeed, an internal nonedge contributes , while an internal edge contributes .
Every odd-sized part must provide an unpaired vertex unless it contains the final singleton. Consequently , and hence . Equality holds precisely when every even part is paired internally and every odd part contributes one mixed-block vertex, except that for odd one odd part instead contributes the final singleton.
Set . Formula (1) gives
In a minimal configuration, two distinct mixed blocks use disjoint multipartite parts. For any two blocks , their number of connecting edges is
which is even: if either block is homogeneous, every nonzero summand is even; if both are mixed, their parts are disjoint. Exchanging adjacent blocks therefore changes by an even integer. Thus all orderings of a fixed unordered minimal pairing have the same sign, and
The number of unordered minimal pairings is
with . Every factor is odd. Therefore
In particular , and (3) proves the stronger conclusions
Both hold for every orientation, all part sizes, and all boundary cases . The first equality is precisely the conjecture.
Source: Celano, Sieger, and Spiro, Eulerian polynomials for digraphs, Combinatorial Theory 5 (2025), article 15, Conjecture 6.4, https://doi.org/10.5070/C65165026 ; https://arxiv.org/abs/2309.07240 .