Optimal Setting of Point Spreads



We develop a model of competitive gambling markets addressing two empirical puzzles. First, why do bookmakers not set unbiased lines that try to equalize betting on both sides, and thus profit from commissions with minimal risk? Second, why is there little evidence of bookmakers competing through lower commissions? We show that the interaction between bookmakers' and gamblers' private information can induce biased lines even when all players are maximizing their chances of winning. We also offer an explanation for persistently high commissions charged by seemingly competitive bookmakers; these commissions are necessary to compensate books for assuming the disadvantage of moving first.