The paper tackles a counting problem in combinatorics involving permutations, which are ordered arrangements of numbers. Researchers had previously noticed a surprising pattern: when you count how many permutations contain a given smaller pattern a prescribed number of times, in a very structured way where each occurrence overlaps the next by a fixed amount, the answer is often a "double factorial" number. Double factorials are products of every other integer, like 1, 1, 3, 15, 105, and so on. The open question was whether you could always find explicit, concrete examples of patterns that produce these double factorial counts in a uniform way across all possible overlap sizes and pattern lengths.
The authors answer this question by constructing a specific permutation for every valid choice of overlap length and pattern length. They prove that counting arrangements containing exactly a prescribed number of non-overlapping copies of their constructed pattern always yields a double factorial. The key to the proof is understanding the structure of a certain ordering called an "occurrence poset," which tracks how different copies of the pattern can coexist. They show this poset has a clean, predictable shape built from two simple components: a chain (a totally ordered sequence) and a comb (a structure that branches like teeth on a comb). Counting the valid linear orderings of the comb component, using a clever insertion argument, gives precisely the double factorial numbers.
A notable feature of the paper is that it includes a formal computer verification of the main results using Lean 4, a proof assistant software. This means the mathematical arguments were not just checked by human readers but were encoded into a system that mechanically verified every logical step. This level of rigor is still relatively rare in combinatorics and signals a growing trend toward machine-verified mathematics, giving extra confidence that the results are correct.