Complexity as a guide to understanding decision bias: A contribution to the favorite‐longshot bias debate

Abstract

This paper investigates the origins of a widespread decision bias in betting markets, the favorite-longshot bias (FLB); in particular, whether it is caused by cognitive errors on the part of bettors or by the pricing policies of bookmakers. The methodology is based on previous literature, which has suggested that: (i) races, as decision tasks for bettors, can be distinguished by their degree of complexity and their attractiveness to those with access to privileged information (insiders), (ii) cognitive errors increase as complexity increases, and (iii) bookmakers set odds in a manner to protect themselves from insiders. The degree of FLB was examined in races of varying complexity and attractiveness to insiders using a dataset of 8545 races drawn from the parallel bookmaker and pari-mutuel markets operating in the UK in 2004. The results, interpreted in the light of the cognitive error and complexity literature, suggest that neither bettors' nor bookmakers' cognitive errors are the main cause of the bias. Rather, bettors' preferences for risk and the deliberate pricing policies of bookmakers play key roles in influencing the bias in markets where bookmakers and pari-mutuel operators coexist.

The Credit Crisis and De Nova Mimicking in Security Analysis

“…One series of studies finds that with increased complexity, jury awards become more unpredictable (Horowitz et al [2001], Horowitz et al [1996]). Another study observes that cognitive errors increase as the complexity of the decision increases (Sung et al [2009]). Some researchers find that as tax rate complexity increases, there is an accompanying increase in investment errors (Rupert et al [2003], Boylan and Frischmann [2006]).…”

Heuristic decision making in network linking

“…As humans are boundedly rational (Camerer, 1998), they cope with complexity in decision making by simplification, which commonly involves assessing a judgment object (e.g., linking choice option) using only the subset of properties of the object that are most accessible, that is, that come most readily to mind, rather than using all relevant properties (Gigerenzer et al, 1999), as long as this leads them to a satisfying situation (Simon, 1956). This is clearly illustrated in the literature about the effects of task complexity in several other contexts, like job candidate selection (Timmermans, 1993), audit judgment (Bonner, 1994), consumer choice (Swait & Adamowicz, 2001;Dellaert et al, 2012), and betting markets (Sung et al, 2009), but no empirical research to date has addressed such effects in making complex network formation decisions.…”

“…The objective of the current paper is to investigate whether systematic heuristic shifts occur in individual decision making in network formation as a function of complexity in the network linking setting. Such complexity effects have been studied in several other choice contexts (e.g., Timmermans, 1993;Bonner, 1994;Sung et al, 2009;Dellaert et al, 2012).…”

Heuristic Decision Making in Network Linking

“…As humans are boundedly rational (Camerer, 1998), they cope with complexity in decision making by simplification, which commonly involves assessing a judgment object (e.g., linking choice alternative) using only the subset of properties of the object that are most accessible, that is, that come most readily to mind, rather than using all relevant properties (Gigerenzer et al, 1999), as long as this leads them to a satisfying situation (Simon, 1956). This is clearly illustrated in the literature about the effects of task complexity in several other contexts, like job candidate selection (Timmermans, 1993), audit judgment (Bonner, 1994), consumer choice (Swait & Adamowicz, 2001;Dellaert et al, 2012), and betting markets (Sung et al, 2009), but no empirical research to date has addressed such effects in making complex network formation decisions.…”

“…Therefore, plexity effects have been studied in several other choice contexts (e.g., Timmermans, 1993;Bonner, 1994;Sung et al, 2009;Dellaert et al, 2012).…”