Foundations and evidence

Trace the foundations. Build beyond them.

Psychohistory begins as a fictional idea, but the practical method draws from probabilistic forecasting, behavioral research, statistics, habit science, and person-specific modeling.

Origin of the term

Psychohistory is the established English term

Isaac Asimov used psychohistory for a fictional discipline associated with the Foundation stories. Its central premise is that very large human populations may display regularities that support mathematical prediction of broad historical and social developments.

Asimov's idea supplies the name and direction, not the content. The Order does not reproduce protected characters, settings, plots, institutions, fictional equations, dialogue, quotations, art, logos, or screen designs. The movement, doctrine, Canon of Contingency, liturgies, interface, code, and practical method are original.

The Encyclopedia of Science Fiction: “Psychohistory”

A concise history of the term in science fiction and its association with Asimov's work. Used here to verify terminology and literary provenance, not as scientific evidence.

sf-encyclopedia.com/entry/psychohistory

Probabilistic forecasting

Scoring, training, tracking, and the outside view

The practice requires users to state probabilities, preserve them, and resolve them because forecasting skill exists only in a scored record. It begins with reference classes because comparable cases often predict more reliably than a compelling story about the present case.

Glenn W. Brier, “Verification of Forecasts Expressed in Terms of Probability” (1950)

The foundational paper for the Brier score. The local ledger uses the binary form: squared difference between the forecast probability and an outcome coded 1 or 0.

Monthly Weather Review, 78(1), 1-3

Barbara Mellers and colleagues, “Psychological Strategies for Winning a Geopolitical Forecasting Tournament” (2014)

Reports benefits associated with probability training, teaming, and tracking in a geopolitical forecasting tournament. It supports the Order's use of explicit probabilities, regular scoring, collaborative judgment, and disciplined revision.

doi:10.1177/0956797614524255

Daniel Kahneman and Dan Lovallo, “Timid Choices and Bold Forecasts” (1993)

Develops the contrast between an inside view centered on the current plan and an outside view grounded in comparable past cases. This informs the instruction to establish a reference class before writing a detailed personal story.

doi:10.1287/mnsc.39.1.17

Self-prediction and behavior

Intention matters, but context and friction remain powerful

The personal method gives repeated behavior more weight than current motivation. Research on self-prediction indicates that people may overweight present intentions, while work on habits and implementation intentions supports tracking stable cues and specifying the situation in which an action will begin.

Connie S. K. Poon, Derek J. Koehler, and Roger Buehler, “On the Psychology of Self-Prediction” (2014)

Examines optimistic self-prediction and finds evidence that current intentions can receive too much weight relative to their imperfect translation into future behavior.

doi:10.1017/S1930297500005763

Peter M. Gollwitzer, “Implementation Intentions: Strong Effects of Simple Plans” (1999)

Reviews plans that connect an anticipated situation to a goal-directed response. This supports the site's insistence that an intervention name a trigger and an action rather than merely repeat a goal.

doi:10.1037/0003-066X.54.7.493

Judith A. Ouellette and Wendy Wood, “Habit and Intention in Everyday Life” (1998)

Reviews multiple processes through which past behavior predicts future behavior, including the importance of repeated behavior in stable contexts.

doi:10.1037/0033-2909.124.1.54

David T. Neal and colleagues, “How Do Habits Guide Behavior?” (2012)

Investigates perceived and actual triggers of habits and informs the prompts about frequency, context stability, and environmental cues.

doi:10.1016/j.jesp.2011.10.011

Person-specific modeling

Within-person patterns may differ substantially across people

The site begins with a person's own repeated observations rather than assuming that a population average is the final description of that individual. Intensive longitudinal research shows why person, situation, and time may need to be modeled together.

Emorie D. Beck and Joshua J. Jackson, “Personalized Prediction of Behaviors and Experiences” (2022)

Uses intensive longitudinal data and person-specific models for everyday experiences and behaviors. The study supports attention to person, situation, and time, while its sample and outcomes limit generalization.

doi:10.1177/09567976221093307

Data protection

Useful records require disciplined control

Personal psychohistory becomes more useful as records deepen, which makes data discipline part of model quality. The Forecast Lab keeps its ledger in the browser. Deployed question and contribution routes process the material and delivery details submitted for human review, as described on the privacy page.

European Commission: Data Protection Explained

Official overview of the GDPR's processing principles and individual rights. Production compliance depends on the operator, jurisdiction, purposes, data categories, vendors, and technical design.

commission.europa.eu, Data Protection Explained

Open research problem

What remains to be proved

No existing discipline yet matches the full reach of Asimov's fictional psychohistory. The Order's central proposition is that forecasting science, behavioral data, social modeling, and sustained compute can be assembled into an increasingly capable practice. That proposition must earn confidence through resolved predictions.

The calculator is the Order's first transparent personal model. Its general weights have not been validated on a representative dataset, so the initial output is a prior to test. The long-term objective is to replace generic weights with patterns learned from resolved personal and collective records.

The research program must still confront sampling bias, measurement error, changing systems, feedback effects, causal confusion, overfitting, missing variables, rare events, distribution shift, and uneven costs of error.

Standard of proof: judge the Order by published forecasts, calibration, out-of-sample accuracy, reproducibility, and whether each generation of models outperforms the last.