Combinatorial analysis uncovers exact path counts in multidimensional king graphs, highlighting polynomial structures and recurrence relations.
Let Kd,n = P_n⊠d be the d-fold strong product of a path, with opposite corners s and t. For k ≥ 0, let Cd,k(n) count simple s–t paths with n − 1 + k edges. Coordinate deficits show that every such path has at most dk moves other than the positive main diagonal. We prove that Cd,k(n) = Σ_{S ⊆ In+k−1,k} (−1)|S| Mn,k(S)^d. Consequently, as a function of d, Cd,k(n) is a finite linear combination of integer exponentials and therefore satisfies a constant-coefficient recurrence. For every d ≥ 2, we obtain closed forms for k = 0, 1, 2, 3; the k = 3 formula is accompanied by an independently executable Möbius certificate. A constructive skeleton-gap argument proves polynomiality for fixed d and k throughout the sufficient range n ≥ d²k² + (2d − 1)k + 2, with Qd,k(n) = 1/(k!)^d [n^(dk) + α_d(k)n^(dk−1) + O(n^(dk−2))], where α_d(k) = k(2k − 3) for d = 2, and α_d(k) = (3d/2)k(k − 1) for d ≥ 3. An independent nineteen-skeleton classification verifies the two-dimensional excess-two formula. Reflection gives the exact admissible one-coordinate word count, and in the growing-dimension regime Cn,k(n) ~ e^[3k(k−1)/2] (n^k/k!)^n for fixed k. The novelty claim is restricted to this fixed-excess structure; the graph family and the elementary partition by path length are not new.
No takes yet. Share an insight, caveat, or question.
Douglas da Silva Ferreira Brilhante (2026) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: