Power 5 Against the Spread

The table below illustrates results of Power 5 conference series against the spread in the most recent 10 games of the given series. In some cases, fewer than 10 games in the series have been played so the full series record against the spread is shown. If a game was played in a year when one or both of the teams were not in their current conference I have still included them. All data was pulled from https://www.oddsshark.com/ncaaf/database which appears to contain results going back to the 1995 season.

The most interesting data points to me are the outliers, where one team has won at least 8 of the last 10 games against the spread. My question is why does Vegas err on one side of  the line in these series so consistently? Is this just coincidence and bound to happen when setting lines for such a large number of games, or is there something particular that happens for these teams that leads the results to skew one way? Of the 21 outliers identified, 14 of them have scheduled games in 2018. I will be tracking the results of those games this year to see if the outliers are more likely to remain skewed in that direction, revert to the mean, or if there is no obvious correlation at all and this is dictated by randomness.

Open table in new tab