For every football game there is a team that is expected to win by a certain number of points. In betting parlance, this is called the spread. If point spreads are accurate, we would expect about half of all games played to result in the favored team winning (beating the spread) and about half of all games to result in the team favored to not beat the spread. The accompanying data represent the results of 45 randomly selected games where a 0 indicates the favored team did not beat the spread and a 1 indicates the favored team beat the spread. Do the data suggest that sport books establish accurate spreads?