Class strength estimates

In thinking about the 4A-3A games, it occurred to me that I really need a baseline measurement of the difference between 4A and 3A. I took the score data from 2001-2009 and then applied the team data for 2010. Thus the games played by Sioux City Heelan and Waukee, 2001-2009, are the basis for this comparison between 4A and 3A. Class A and Class 8 are similarly connected by the many more teams that move up and down this divide.

cl   mean    stdev
4A   87.44   15.05
3A   85.87   13.31
2A   72.57   13.50
1A   68.04   17.49
 A   61.98   15.33
 8   45.90   20.64

Initial thoughts:

1. Mean of class 4A and class 3A are very close in this time period. This is again primarily based on the schedules for Heelan (2001-2005 4A, 2006-2009 3A) and Waukee (2001-2005 3A, 2006-2009 4A). This seems like an underestimate of the difference between 4A and 3A. However, the time period is a seemingly large 9 years, with the pre and post time periods are very close in length (5 years pre-change, and 4 years post-change). I am very interested in seeing how Newton does in 3A this year.

2. Continuing with 4A and 3A, 4A has a larger standard deviation than 3A. This means that the best teams in 4A are better than the best teams in 3A. Additionally, and since the means of 4A and 3A are about the same, the worst teams in 4A are probably worse than the worst teams in 3A.

3. A large divide exists between 3A and 2A. A short-term reason for this may be that Solon and Boyden-Hull-Rock Valley are coming up from 2A. In using the 2010 team assignments, 3A now gets credit for all of Solon and R-H-RV’s 2001-2009 game history. However, this seems to me to be only a 1-2 point change, and thus fundamentally there seems to exist a large difference between 3A and 2A.

4. There is quite a divide between Class A and Class 8. I think the 8-man teams are simply much smaller than the Class A teams.