Wiki
Wiki

Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.

Updated

Muyesser 2022 random hall paige conjecture

../

theorem_6_9: For a large group G, color sets C of size at least |G| - |G|^{1/2}, vertex sets V with |V| + 1 = |C| and a product condition in the abelianization, the division and multiplication digraphs contain directed rainbow Hamilton paths between prescribed endpoints; read here as the source of the very large range of Graham's rearrangement problem.


Alp Müyesser, Alexey Pokrovskiy, A random Hall-Paige conjecture. arXiv:2204.09666 (2022).

Theorem 1.1, the main result, states that for a group G of order n and p >= n^{-1/10^105}, with p-random subsets R^1, R^2, R^3 of G, with high probability every triple of equal-sized sets X, Y, Z whose symmetric differences from them total at most p^{10^30} n / (log n)^{10^30} elements and that satisfy the abelianized sum condition sum X + sum Y = sum Z in G^{ab} admits a bijection phi: X -> Y with x -> x phi(x) a bijection onto Z. Taking p = 1 and X = Y = Z = G recovers the Hall-Paige conjecture for large groups (originally proved in 2009 by Wilcox, Evans and Bray using the classification of finite simple groups) by purely combinatorial means, and Proposition 1.2 derives that the multiplication table of a large finite group contains a near transversal. Using Theorem 1.1 as a black box the authors characterize large sequenceable and R-sequenceable groups (settling problems of Gordon 1961 and Ringel 1974 for large groups and confirming Keedwell's 1981 conjecture that every sufficiently large non-abelian group is sequenceable), prove Snevily's 1999 conjecture on transversals of subsquares of abelian multiplication tables for large subsquares, in a strong form (Theorem 1.4), characterize large abelian groups partitionable into zero-sum sets of given sizes (a problem of Tannenbaum 1981 and a conjecture of Cichacz), and characterize large harmonious groups (a problem of Evans 2015). The proof mixes probabilistic absorption in the spirit of Keevash's design construction with the algebraic Hall-Paige obstruction in the abelianization. Problem 475 asks for orderings of subsets of F_p with distinct partial sums; the paper does not treat it, but its Theorem 6.9 yields the case of very large subsets, as recorded below.

The retained folder-name PDF is arXiv:2204.09666v3 (25 February 2025, "final version, to appear in Inventiones Mathematicae", 73 pp.), whose pagination is used here; the journal version is Invent. Math. 240 (2025), no. 3, 779--867, DOI 10.1007/s00222-025-01328-x (published online 5 March 2025; Crossref record read), not held and not compared, so the paper is refereed. Read status: claims checked for Theorem 1.1 (p. 3), the definitions of p. 51, Theorem 6.9 and Corollary 6.10 (p. 51), Theorem 6.12 (p. 52) and the opening of Lemma 6.22 (p. 57), text layer, on 2026-09-18; the proofs were not read. On Problem 475 the paper does not state the subset case itself: Theorem 6.9, read in the division digraph, yields that every S⊆Fp∖{0}S\subseteq\mathbb F_p\setminus\{0\} with ∣S∣≥p−p1/2+1|S|\ge p-p^{1/2}+1 has a valid ordering (a reading recorded on the result page), and Bedert, Bucić, Kravitz, Montgomery and Müyesser state the general form, for all finite groups and ∣S∣≥N−N1−γ|S|\ge N-N^{1-\gamma}, as their Theorem 7.1, proved from Lemma 6.22 here. Result page: theorem_6_9. The arXiv record (https://arxiv.org/abs/2204.09666, read 2026-10-02) names the Creative Commons Attribution 4.0 license.

Source: https://arxiv.org/abs/2204.09666.

Bears on. #475

Results to transcribe.

  • Theorem 1.1: Main random Hall-Paige theorem: for p >= n^{-1/10^105} and p-random R^1, R^2, R^3 in a group G of order n, whp any equal-sized X, Y, Z within p^{10^30} n/(log n)^{10^30} symmetric difference of them with sum X + sum Y = sum Z in G^{ab} admit a bijection phi: X -> Y with x -> x phi(x) bijecting X onto Z.
  • Proposition 1.2: The multiplication table of any sufficiently large finite group contains a near transversal (n-1 cells sharing no row, column or symbol), reproving a result of Goddyn and Halasz.
  • Conjecture 1.3 (Snevily): Statement of Snevily's conjecture on transversals of subsquares A x B of abelian multiplication tables, including the even-order cyclic exception.
  • Theorem 1.4: For all n >= n_0, a subsquare A x B of an abelian group G with |A| = |B| = n has a transversal unless A and B are cosets of a subgroup H isomorphic to Z_{2k} x H_odd, or, for H isomorphic to (Z_2)^k, A and B are translates of H minus two distinct elements, a_1, a_2 and b_1, b_2 respectively, with a_1 + a_2 + b_1 + b_2 = 0.
  • Applications (Section 1.1): Characterizations of sequenceable and R-sequenceable groups, of abelian groups partitionable into zero-sum sets of specified sizes, and of harmonious groups, all for large groups.