English

Large deviations in random Latin squares

Combinatorics 2021-12-23 v3 Probability

Abstract

In this note, we study large deviations of the number N\mathbf{N} of intercalates (2×22\times2 combinatorial subsquares which are themselves Latin squares) in a random n×nn\times n Latin square. In particular, for constant δ>0\delta>0 we prove that Pr(N(1δ)n2/4)exp(Ω(n2))\Pr(\mathbf{N}\le(1-\delta)n^{2}/4)\le\exp(-\Omega(n^{2})) and Pr(N(1+δ)n2/4)exp(Ω(n4/3(logn)2/3))\Pr(\mathbf{N}\ge(1+\delta)n^{2}/4)\le\exp(-\Omega(n^{4/3}(\log n)^{2/3})), both of which are sharp up to logarithmic factors in their exponents. As a consequence, we deduce that a typical order-nn Latin square has (1+o(1))n2/4(1+o(1))n^{2}/4 intercalates, matching a lower bound due to Kwan and Sudakov and resolving an old conjecture of McKay and Wanless.

Keywords

Cite

@article{arxiv.2106.11932,
  title  = {Large deviations in random Latin squares},
  author = {Matthew Kwan and Ashwin Sah and Mehtaab Sawhney},
  journal= {arXiv preprint arXiv:2106.11932},
  year   = {2021}
}

Comments

15 pages

R2 v1 2026-06-24T03:28:45.830Z