q-deformations of shuffling operators
Sarah Brauner
Abstract
The k-random-to-random shuffle is a Markov chain that works as follows: one removes k cards from a deck at random and reinserts those k cards into new, uniformly random positions. To understand the convergence behavior of this process, one would like to know its eigenvalues—a surprisingly difficult question that remained open for more than a decade. It was resolved by Dieker–Saliola (k = 1) and Lafrenière (k > 1) in 2019, but the formulas for k > 1 were abstract enough that the second-largest eigenvalue was only conjectured.
In this talk, I will describe work with Commins, Grinberg, and Saliola that gives a q-deformation of this process to the type A Iwahori–Hecke algebra, where one may understand the parameter q as encoding a probability. We prove that the eigenvalues are always polynomials in q with nonnegative integer coefficients and can be expressed as certain evaluations of supersymmetric functions. This allows us to verify—and generalize—Lafrenière’s conjecture for the second-largest eigenvalue.