Underestimating randomness: Outcome bias in betting exchange markets

Abstract

This paper examines whether the outcome bias harms price efficiency in betting exchange markets. In soccer, the match outcome is an unreliable performance measure, as it underestimates the high level of randomness involved in the sport. If bettors overestimate the importance of past match outcomes and underestimate the influence of good or bad luck, we expect less accurate prices for lucky and unlucky teams. Analyzing over 8,900 soccer matches, we find evidence that the prices are overstated for previously lucky teams and understated for previously unlucky teams. Consistent with the outcome bias, the betting community verestimates the importance of past match outcomes. Consequently, this bias translates into significantly negative betting returns on lucky teams and positive betting returns on unlucky teams. Based on this finding, we propose a simple betting strategy that generates positive returns in an out-of-sample backtest.