Statistical Analysis of the Cerro Macareno Tablet: Non-Random Structure, Spatial Gradients, and Astronomical Exploration

We present a quantitative analysis of the Cerro Macareno Tablet, a fired-clay artefact (8 rows × 14 columns = 112 cells) housed at the Museo Arqueológico de Sevilla. Previous work by Sáez Uribarri (2006) demonstrated non-randomness using discriminant analysis. We extend that baseline with modern spatial statistics, spectral methods, and astronomical hypothesis testing. Our central finding is strong evidence against uniform or independent mark placement: Fisher’s exact test yields OR = 7.58 (p = 7.0 × 10−6), confirming a pronounced left-to-right gradient, and runs tests reject sequential independence across three distinct linearisations. A spectral peak at period T = 14 cells is noted but cannot be attributed to temporal periodicity independently of the columnar gradient. Regarding astronomical hypotheses, a structured-null permutation test (fixed-margin, 9,999 replications) detects a residual structured association between a Venus elongation model and the tablet beyond the row-plus-column baseline (p = 0.034). However, a leave-one-row-out specificity analysis comparing Venus against eight pseudo-periodic families plus a jittered Venus control shows that Venus performs worse than all substantive periodic controls, ranking 9th out of 10 families tested (Venus+jitter ranks last), with no significant advantage over the best control (pbootstrap = 0.38; descriptive bootstrap only). We conclude that the tablet exhibits robust non-random spatial structure, but the data do not identify Venus as the unique or best-supported periodic astronomical explanation.