I started with the ESPN Football Power Index of each FBS team, then additionally computed each conference's average FPI to plot each team on an 2D plane. Then I used K-Means Clustering to categorize each of the 134 teams into similar clusters.
The teams subdivided nicely into twelve different groups, which I will detail below:
College Football Playoff Contenders (15)
College Football Playoff Crashers (9)
Power 4 Dark Horse (16)
Group of 6 Dark Horse (5)
Group of 6 Conference Contenders (11)
Power 4 - Solid (11)
Group of 6 - Solid (8)
Power 4 - Mid (13)
Group of 6 - Mid (10)
Power 4 - Basement (15)
Group of 6 - Basement (13)
Group of 6 - Rebuild (8)
Comments: This exercise was mostly done as an exercise to teach myself how to do K-Means Clustering. I have no idea how accurate these clusters will be relative to the launch-day team ratings in EASCFB25, but for now it may prove useful to you as a broad guide.
ESPN FPI strongly favors Power 4 teams across the board; only teams in the CFP Crasher cluster have stronger FPI ratings than any Power 4 team, and even then, none of the CFP Crasher teams have FPI ratings stronger than the lowest-rated team in the Power 4 - Solid cluster. Further, every team in the Power 4 - Basement cluster has a higher FPI than all G6 teams not in the CFP Crasher cluster. Given that the sport of college football is rather strongly defined by upsets — especially the further one moves away from the top of the sport — I don't believe this disparity between the bottom of the Power 4 and the second tier of the Group of 6 to be truly accurate.


							
						
 Your rebuild will be fun.
							
						


Comment