P5 Alternatives

Got Jukes? · Jul 11, 2017

cpthomas said: ↑

Good discussion. Perhaps of interest, in trying to come up with pre-season ratings for teams in order to project an end-of-season tournament bracket, the study I did (which was based on the 2016 season) indicated that the best forecaster of next year's ratings is the team's average rating over a number of years. The number of years varies with the longevity of the head coach, i.e., the longer the coach has been there, the greater the number of prior years to average: for coaches who have been there 9 years or more, I use their average over the last 8 years; for coaches there from 4 to 8 years, I use their average over the last 6 years; and for coaches there 3 years or less, I use their average over the last 3 years. I came up with this after trying a lot of different approaches, as the approach producing ratings that come closest to the end-of-season ratings. And, one of the possible approaches I considered involved evaluations of graduated players, returning players, incoming players, and coaches (not my evaluations, but ones that I consider very good). According to what I did, the coach is the biggest factor, in other words player turnover is not as big a factor as one might think.

When I look at teams that have had multiple appearances in the College Cup over the last 10 years:

Florida State 6 appearances -- same coach entire time
Stanford 6 appearances -- same coach entire time
UNC 4 appearances -- same coach entire time
UCLA 4 appearances -- same coach for first 3, current coach to some extent inherited from that coach
Notre Dame 4 appearances -- same coach for all 4
Duke 2 appearances -- same coach for entire time
Penn State 2 appearances -- same coach for entire time
Virginia 2 appearances -- same coach for entire time
USC 2 appearances -- two different coaches

The other teams, with 1 appearance each, are Georgetown, West Virginia, Rutgers, Texas A&M, Wake Forest, Virginia Tech, Boston College, and Ohio State.

The top list, of multiple appearance teams, comprises 80% of all College Cup participation over the 10 years.

Who the coach is appears to me to be the most critical factor, by quite a bit.
Click to expand...

Definitely agree with your point that the coach makes the biggest difference. If it was player turnover, then one would expect Stanford and UCLA to win every year with the number of YNT players they bring in each year.