Decision-Making under Risk: Effect of Complexity on Performance

Abstract

This paper examined the effect of complexity on performance in decision-making under risk employing a database of 1161 randomly sampled betting decisions made during 1987 in offcourse horserace betting offices situated throughout the UK. The setting for the study is characterised by important ecological advantages over laboratory or alternative naturalistic settings. Complexity is defined in terms of both the number of alternatives in the decision-maker's choice set (number of horses in a race) and the complexity of the set of attributes for each horserace (handicap vs nonhandicap races). Performance was not adversely affected by an increase in the number of alternatives and only to a limited extent by an increase in attribute-based complexity. The robustness of the results is reinforced by adjustments which acknowledge and control for the differential random chance of correct decision-making across groups of events with differing complexity.