Note
Some predictive judgements are formulated based on complex tables and analysis, while others rely on intuition and the system 1. On this second one, there are two subdivisions:
- Some are expert guesses based on experience
- Others, that are sometimes personally indistinguishable from expert ones, come from heuristics that switch with an easier question a harder question
Intuitive judgments are formulated with high certainty even when they are based on non-regressive valuations not based on proofs.
Take for example, trying to guess the GPA of a student based on the fact that she/he was a early reader. A high number of people will predict a high GPA, not taking in consideration other factors.
To produce a non-distorted prediction, follow this framework:
- start with an initial prediction of the average GPA
- determine the GPA that you got from the proofs
- evaluate the correlation between your proofs and the GPA
- if the correlation is 0.30, move 30% of the average towards the correspondent GPA
This framework will moderate your prediction making it regress towards the mean.
See also
- Inside View uses specific details, Outside View uses base rates — the outside view provides the base rate (step 1); the inside view provides the evidence-implied prediction (step 2); this framework formalises how to combine them
- Planning Fallacy, we predict ideal scenarios instead of consulting distributional data — the planning fallacy is what happens when step 1 is skipped entirely: no base rate, only evidence-implied optimism
- Optimism Bias drives entrepreneurial dynamism but distorts risk assessment — the failure to regress toward the mean is driven by optimism bias: we weight our specific evidence more than statistics warrant
- Non possiamo eliminare i bias, possiamo solo riconoscere le situazioni a rischio — this four-step framework is the practical implementation of “engage Sistema 2 when the stakes are high”