Category
Forecasting & Calibration
Browse 22 free mental model guides in Forecasting & Calibration.
9 guides
Aggregation & Extremizing Playbooks
Calibration Training
A structured process for aligning your subjective confidence with your objective accuracy to make more reliable predictions.
Read guideIndependent Forecasts
Forecasts built separately produce better collective predictions than those developed together.
Read guideBrier Score
A mathematical tool that rewards confidence when you're right and penalizes you harshly when you're wrong—exposing the difference between feeling certain and being correct.
Read guideMitigating Correlated Errors
Strategies to prevent groups from making the same mistake, turning 'wisdom of crowds' into 'madness of mobs.'
Read guideSimple Forecast Averaging
Improve prediction accuracy by combining multiple independent forecasts into a single, simple average.
Read guideWisdom of the Crowd
Under the right conditions, averaging independent guesses from diverse groups often beats individual experts.
Read guideWeighted Averaging by Track Record
A method for combining predictions by giving more influence to sources with proven past accuracy, using scoring rules to measure and weight expertise.
Read guideForecast Calibration
The degree to which predicted probabilities match actual outcomes over time—the difference between what you expect to happen and what actually happens.
Read guideWhy Aggregates Are Conservative
Explores why averages and group data are more stable and less extreme than the individual pieces they are made of.
Read guide13 guides
Aggregation & Wisdom of Crowds
Brier Score
A method to measure the accuracy of probability forecasts by rewarding both correctness and confidence.
Read guideCalibration Training
A method to improve the accuracy of your predictions by systematically tracking your confidence against real-world outcomes.
Read guideConditions for Crowd Wisdom
An editorial checklist for collecting and combining judgments: varied information, private first answers, local knowledge and suitable aggregation. It does not guarantee accuracy.
Read guideCrowdsourcing vs. Crowd Wisdom
Distinguishes between getting work done by many people (crowdsourcing) and extracting accurate collective judgments from diverse, independent opinions (crowd wisdom).
Read guideExtremizing Forecasts
Moving a combined probability forecast away from 50% toward 0% or 100%, with the adjustment checked against past outcomes.
Read guideGalton's Ox Experiment
The surprising discovery that the collective average guess of a large, diverse crowd can be remarkably accurate.
Read guideInformation Cascades
How people make decisions by observing the actions of others, sometimes ignoring their own private information.
Read guidePrediction Markets
Markets where people trade on future events, turning scattered information into a shared forecast when participation and incentives are strong.
Read guideSimple Averaging of Forecasts
A powerful method for improving prediction accuracy by taking the simple average of multiple forecasts, which cancels out individual errors and often outperforms a typical single forecast.
Read guideSuperforecasting Principles
A set of empirically tested cognitive habits that distinguish elite predictors from ordinary experts, focusing on how they think, not what they think.
Read guideDelphi Method
A structured way to combine expert judgments through private questionnaires, repeated rounds and shared feedback, without requiring everyone to agree.
Read guideThe Wisdom of Crowds
The collective opinion of a diverse, independent group is often more accurate than that of any single expert.
Read guideWeighted Averaging
A method for combining judgments that gives more influence to more reliable or knowledgeable sources, improving accuracy over simple averaging.
Read guide