To What Extent Is Foresight Luck? Implications for Cognitive, Evolutionary, and Chance Theories of Foresight
Type
working paper
Date Issued
2024-12-31
Author(s)
Abstract
This study assesses the relative plausibility of cognitive, evolutionary, and chance theories of foresight by quantifying the extent to which foresight is luck. Results from a regression-towardthe-mean approach applied to foresight estimates derived from a beta item response theory model using data from 265 forecasters who made 40,514 forecasts on 212 questions over two years in a forecasting tournament suggest that foresight is on average 32% luck. This means that the foresight of an individual forecaster regresses on average by 32% toward the mean foresight of all forecasters over one year. This regression toward the mean is stronger among top forecasters (35%) and forecasters with below average foresight (47%) but weaker for the average forecaster (23%). Despite this regression toward the mean, the top forecasters in year 1 did not display worse foresight than other forecasters in year 2. I discuss the implications of these findings for cognitive, evolutionary, and chance theories of foresight.
Keywords
foresight
cognition
evolution
luck
item response theory
regression toward the mean
File(s)![Thumbnail Image]()
![Thumbnail Image]()
open.access
Name
Chapter 4 Foresight Luck.pdf
Size
3.72 MB
Format
Adobe PDF
Checksum (MD5)
5f55541a59191fe71370680971b7bae2
open.access
Name
FL.pdf
Size
2.64 MB
Format
Adobe PDF
Checksum (MD5)
998003f3e93c9587be8c01e52108a08e