In Isaac Asimov’s Foundation novels, “psychohistory” is a fictional science that predicts the large-scale behaviour of galactic populations. It works only for enormous numbers of people, and only if those people do not know they are being predicted.
Both conditions are instructive. Statistical regularities emerge from aggregates, not individuals. And measurement can change behaviour — teams adapt to being watched.
Engineering, not prophecy
Predictive engineering keeps the useful half of the idea: aggregate patterns, explicit models, falsifiable outputs. It drops the rest: no individual destinies, no hidden observation, no claims beyond the evidence.
Our unit of analysis is the project and the team configuration, not the person. Our standard is calibration against observed outcomes, not narrative plausibility.
This is a concept note. It reports no experiments and no results.