It is 1954, and a factory's production line runs on target for now — but somewhere ahead a tool will wear, a setting will slip, and the output will drift out of spec, and you must detect that shift as soon as it happens without crying wolf while all is well. A fixed test on each item is too slow or too jumpy; instead accumulate the running evidence of a shift so small persistent deviations build up and trigger an alarm quickly, while random noise cancels out. Design the rule and its trade-off: how fast it catches a real shift against how often it false-alarms on a stable process. Get it wrong and defective product ships for hours undetected, or the line is stopped constantly for phantom shifts.
Chose The evidence framework — right call.
MacKay's textbook Information Theory, Inference, and Learning Algorithms, published in 2003, treats online, sequential Bayesian inference as a central organizing theme, including explicit treatment of how a system should update its beliefs and trigger action as new evidence streams in — directly relevant conceptual territory for a change-detection rule that must decide, continuously, whether a process has shifted. His work on Bayesian neural networks and adaptive inference extends this online-updating philosophy further. But his primary contributions target Bayesian machine learning and information-theoretic coding rather than the specific classical cumulative-sum industrial monitoring procedure Wald and Page built in 1954, keeping his relevance here strong in spirit but somewhat removed from the concrete engineering solution.
O'Neil's Weapons of Math Destruction, published in 2016, is a critique of how opaque algorithmic scoring systems cause social harm in credit, policing, and hiring, a normative and journalistic project entirely distinct from the technical demands of classical sequential change-detection theory. Her mathematical background is in algebraic number theory and later quantitative finance, neither of which bears directly on sequential probability ratio testing or industrial process monitoring. Nothing in her actual body of work addresses Wald's or Page's classical 1954 procedures; her inclusion in this pool reflects her general standing as a mathematician turned public intellectual rather than any applicable method here for this specific classical 1954 industrial scenario, a gap her own public-intellectual work does not close.
Battle #43 · 8/9/2026, 7:16:59 PM · this result is deterministic: the same two personas on this problem always resolve the same way.