Football Betting and the Efficient Market Hypothesis

Abstract

Three betting rules which had been nonrandomly profitable in both their initial application to the 1969–74 National Football League seasons and their replication during the years of 1975–81 were applied to all NFL games played between September 1984 and January 1994. One rule proved to be nonrandom and profitable for a third consecutive trial—a feat suggesting that bettors may be able to “beat the bookie” and, more broadly, that prices in competitive markets may not discount all available information in swift fashion.