The paper solves a counting problem about a specific combinatorial object called a "solid standard Young tableau." You can think of these as ways of filling a particular two-layer grid shape with numbers 1 through some maximum, following strict ordering rules in every row, column, and layer. The shape in question looks like two rectangular grids stacked on top of each other, and the question is: how many valid ways are there to fill such a shape when the grid size is n? Call this count g(n). A mathematician named Zeilberger noticed by computer experiment that g(n) satisfies a specific type of formula, a linear recurrence of order 2, meaning each value of g(n) can be expressed using the two previous values with coefficients that are polynomials in n. He challenged the mathematical community to actually prove this, not just observe it computationally.
The authors prove the recurrence by translating the counting problem into a completely different setting. Through a clever rearrangement procedure called a deletion-insertion bijection, they recast g(n) as a problem about counting certain paths on a grid, specifically paths that stay in the positive quadrant and follow movement rules associated with something called Kreweras walks. These lattice walk problems have a rich theory, and the authors use an algebraic technique called the kernel method to get exact formulas for the relevant path counts. Along the way, they discover several bonus results, including new closed-form expressions for related families of paths and an unexpected identity connecting different types of walks.
Once they have explicit formulas for the quantities involved, the authors explain structurally why the recurrence has to exist and why it has order 2 specifically. The key insight is that g(n) can be written as a combination of two so-called hypergeometric terms, meaning two well-behaved mathematical sequences with a specific multiplicative structure. Any function that lives in a two-dimensional space spanned by such terms must automatically satisfy a second-order linear recurrence, and the actual polynomial coefficients of that recurrence can be extracted using a standard linear algebra technique called Cramer's rule. This turns an empirical observation into a fully explained and proven mathematical fact.