Evolution of optimal L\'evy-flight strategies in human mental searches

TL;DR

This paper explains how human Levy-flight search strategies in online auctions can emerge naturally through an evolutionary process where search ability influences reproductive fitness, without assuming initial strategies.

Contribution

It introduces an evolutionary model that accounts for the emergence of optimal Levy exponents in human search behavior without strong initial assumptions.

Findings

Empirical data shows humans adopt Levy flight strategies in online auctions.

An evolutionary model reproduces observed exponents based on search ability and fitness.

The approach generalizes previous models and suggests new research directions.

Abstract

Recent analysis of empirical data [F. Radicchi, A. Baronchelli & L.A.N. Amaral. PloS ONE 7, e029910 (2012)] showed that humans adopt L\'evy flight strategies when exploring the bid space in on-line auctions. A game theoretical model proved that the observed L\'evy exponents are nearly optimal, being close to the exponent value that guarantees the maximal economical return to players. Here, we rationalize these findings by adopting an evolutionary perspective. We show that a simple evolutionary process is able to account for the empirical measurements with the only assumption that the reproductive fitness of a player is proportional to her search ability. Contrarily to previous modeling, our approach describes the emergence of the observed exponent without resorting to any strong assumptions on the initial searching strategies. Our results generalize earlier research, and open novel questions in cognitive, behavioral and evolutionary sciences.