Betting strategies based on the presence of home-underdog bias in the NFL have been shown to produce returns in excess of those predicted by market efficiency in some situations. Dare and Dennis (2011) attribute this bias to bettors underestimating the scoring ability of home underdogs. Using a more recent sample of data, we find contradictory results. We challenge the assumptions of the Dare and Dennis (2011) model and use detailed betting data to offer an alternative rationale for the homeunderdog bias. We illustrate that bettors have clear and predictable tendencies for betting on the best teams, and sports books do not appear to be operating as suggested by the balanced-book hypothesis.