A comprehensive statistical evaluation of the Orion–Giza correlation

We present the first exhaustive statistical assessment of Bauval’s (1994) Orion Correlation Theory, testing whether the spatial arrangement of the three Great Pyramids of Giza encodes the shape of Orion’s Belt. Using Procrustes analysis against all 695,520 stellar triplets brighter than visual magnitude 4.0, drawn from the Yale Bright Star Catalogue and independently confirmed with Gaia DR3 proper motions, we find the Giza–Orion shape correspondence ranks 1,717th (p = 0.0025). This constitutes 8.7 bits of information content across two independent shape parameters (distance ratio and bend angle). The match uses pure rotation without reflection, neutralizing Fairall’s (1999) mirror objection, and Giza ranks first among 12 famous triangular monument sites tested against their culture’s top asterism (BH-corrected p = 0.030 across sites). The Procrustes scale factor is not statistically significant (p = 0.219), adding no information beyond the two shape degrees of freedom. However, every proposed extension of the correlation fails quantitative testing: the pyramid-axis orientation does not match Orion’s position angle at any epoch (p = 0.74–0.98); the four Great Pyramid shafts do not converge on a single target epoch (spread = 1,325 years); Menkaure’s offset from the Khufu–Khafre line is in the opposite direction to Mintaka’s; the Nile Valley extension collapses (p = 0.50); and a novel builder’s precision envelope analysis reveals the match sits at the 0.8th percentile of what intentional naked-eye encoding could produce. We conclude that the three-point shape correspondence is statistically significant but far more limited than traditionally claimed.