On products of permutations with the most uncontaminated cycles by designated labels

Abstract

There is a growing interest in studying the distribution of certain labels in products of permutations since the work of Stanley addressing a conjecture of B\'{o}na. This paper is concerned with a problem in that direction. Let $D$ be a permutation on the set $[n]=\{1,2,\ldots, n\}$ and $E\subset [n]$. Suppose the maximum possible number of cycles uncontaminated by the $E$-labels in the product of $D$ and a cyclic permutation on $[n]$ is $\theta$ (depending on $D$ and $E$). We prove that for arbitrary $D$ and $E$ with few exceptions, the number of cyclic permutations $\gamma$ such that $D\circ \gamma$ has exactly $\theta-1$ $E$-label free cycles is at least $1/2$ that of $\gamma$ for $D\circ \gamma$ to have $\theta$ $E$-label free cycles, where $1/2$ is best possible. An even more general result is also conjectured.