By RICHARD CHEEKS

One prevailing premise that global rating systems such as Sagarin use as a basis for their pre-season rankings is how a team finished the previous season. The new pre-season rankings presuppose that teams, on average, do not vary much, year to year. On average, I believe this is a valid presumption, but I believe we can all agree that for any specific team, the assumption of exact consistency is baseless. Some teams will perform in the new season at a more efficient level than they did the previous year, and some teams will perform at a less efficient level. Returning players gain maturity and experience. Teams replace their departed players with new recruits.

Is there an objective basis to predict whether a particular team will improve or lose efficiency, and is there an objective basis to quantify these likely efficiency changes? One indicator of this trend may be the number of returning starters that a team brings into the new season from the previous season. I do not know the answers to these questions, but I will use the 2011 season with Kentucky and its 12 regular season opponents to begin the analysis. Then I can assess whether this parameter has promise for further study, or whether some other parameter, such as returning cumulative playing time.

According to the UK Media guide for UK opponents, and the Tennessee Media guide for the three SEC West teams not on the 2011 schedule, these teams return between 6 [Auburn] and 19 [Vandy] starters from their 2010 teams. The average team in this group returns 13.9 starters. Eight of them return 15 to 19 players, while seven return 13 to 6 players, with one team [Central Michigan] returning the group average 14 players.

I have monitored football efficiencies and believe that the Net Game Efficiency, in Points Per Possession, is an excellent measure of the team’s strength. An increase in a team’s NGE, one season to the next, corresponds to better play, and a stronger team. Conversely, a decrease in a team’s NGE, one season to the next, corresponds to poorer play, and a weaker team. At this point, the change in NGE is shown a zero for all 13 teams, but as the season unfolds, each team’s NGE will update in this table in real time. At the end of the 2011 season, I will use the data to test the following hypothesis.

Hypothesis:

Teams with the lowest number of returning starters will perform at a lower NGE in 2011 than they did in 2010, and the teams with the highest number of returning starters will perform at a higher NGE in 2011 than they did in 2010.