Expert Mental Modelling: Building a Coherent Worldview

## The Mental Model Library Charlie Munger advocated building a "latticework of mental models" from multiple disciplines. A bettor with a rich mental model library applies the right model to each situation — rather than forcing every situation through a single framework. ## The Core Betting Mental Model Stack **Probabilistic thinking:** Every outcome has a probability. Evaluate decisions by EV, not by outcome. (Foundation) **Base rates and reference classes:** Start with the outside view. Adjust modestly for specific information. (Calibration) **Regression to the mean:** Extreme observations are partially variance. They revert. Bet accordingly. (Correction) **Bayesian updating:** Update beliefs with new information proportionally to its diagnosticity. (Learning) **Second-order thinking:** What do other market participants think? Where does consensus diverge from probability? (Edge location) **Expected value in distribution:** Think in distributions, not point predictions. Calculate EV from the full probability curve. (Precision) **Efficient market awareness:** Know how efficient each market is. Concentrate analytical effort where efficiency is lower. (Strategy) **Opportunity cost:** Every decision excludes alternatives. Value the alternatives explicitly. (Resource allocation) ## Applying Multiple Models to One Decision A mature bettor does not pick one model for each decision — they apply multiple models simultaneously: "This looks like a value bet (EV thinking). But the base rate for this scenario is lower than I initially assumed (base rate correction). The market has sharp action supporting my direction, which reduces my edge estimate (second-order thinking). But my model is better-calibrated in this specific league (market efficiency consideration). I'll bet at 60% of my calculated Kelly stake to account for model uncertainty." ## The Expert Edge The bettor who can fluidly apply and integrate multiple mental models produces probability estimates that are more accurate than any single model generates alone. This integration — the synthetic intelligence of a rich mental model library — is one of the deepest and most durable competitive advantages available in betting.