Analysis and Historical Decision-Analysis Reconstruction, 2–3 July 1863

This article tests whether probabilistic expected-utility models and Historical Decision-Analysis Reconstruction (HDAR) yield independent, historically defensible judgments about Joshua Lawrence Chamberlain’s charge on 2 July 1863 and Robert E. Lee’s renewed offensive on 3 July. It distinguishes documentary constraints, model-based probability scenarios, and empirical calibration, which the surviving record does not yet permit. Chamberlain’s report substantiates an ammunition emergency and restrictive orders; Lee’s reports and surviving July 3 artillery messages establish different, time-dependent information environments. A transparent reanalysis shows that the preference for Chamberlain’s charge reverses when its assumed chance of securing the position crosses a model-specific threshold, while Lee’s centre assault remains inferior in the baseline model but can change rank with different assumptions and feasible alternative sets. Independently constructed HDAR findings qualify the numerical models: withdrawal by Chamberlain and immediate manoeuvre by Lee cannot be treated as fully demonstrated, executable alternatives merely because they fit a decision table. The comparison therefore finds a division of analytical labour rather than a proven superior method. HDAR tests historical admissibility; quantitative analysis locates conditional reversals. The principal contribution is an auditable, falsifiable protocol for bringing these judgments into contact without mistaking simulated precision for historical fact.