extremizing algorithm
A statistical aggregation method that, when diverse independent forecasters agree on a probability, pushes the combined estimate further from 50% to reflect the incremental information
As defined in Philip Tetlock on Superforecasting
Where it comes up · 2
- Philip Tetlock on Superforecasting1 claim
- Superforecaster Will China Invade Taiwan How Will The Ukraine War End Nick Hare1 claim
Most important claims · 2 of 2Search all claims for “extremizing algorithm” →
- When advisers each offer the same probability (e.g., 70%) but draw on diverse, independent sources of evidence, the mathematically correct aggregate is more extreme than 70%—because each estimate carries incremental information—whereas if they are clones drawing on identical information, the answer stays 70%; the IARPA extremizing algorithm exploited this and outperformed 99% of individual superforecasters.Philip Tetlock on Superforecasting
- Aggregating forecasts using an extremizing algorithm rooted in information theory outperforms simple averaging, because when two independent forecasters both say 80% there is likely shared information making the outcome more likely, so the aggregate should be pushed higher than the plain average.Superforecaster Will China Invade Taiwan How Will The Ukraine War End Nick Hare