2016 Projected Wins: NCAA FBS Edition

Another friend sent me a message this past week in response to my last article. He noted that the same regressions could possibly be deciphered for college athletics if there were only a way to monetize a salary. Since “students don’t receive payment to attend a particular university”, I had to find another salary. This led me to research head coach salaries for FBS schools.

Luckily, I was able to find data for this year and the last three years’—thanks mostly to USA Today’s Coaches Salary Database—worth of data. I then took that data and arranged it alongside the wins, winning percentages and SRS numbers for each team. (SRS data supplied by www.sports-reference.com/cfb)

Once I had everything sorted out and aligned, I was then able to run some correlation tests on each of the sets of data and found some interesting things.

1. SRS data—because football-reference.com is so amazing—has an 87.4 percent correlation with win percentage.
2. Head Coach Compensation only has about a 41.5 percent correlation with winning percentage. It is also more negatively correlated to points allowed than it is to points scored. Maybe this means that money can only buy good defense, on average?
3. Compensation has a 68 percent correlation with SRS. This was significant, and seeing as SRS had a high correlation with WPCT, I decided that I could use this route Compensation–>SRS Rating–>Win Percentage. That made the computations much simpler.
4. Seeing that the correlation with compensation was still lower than, say, a more acceptable rate of 85 percent, I decided to add in a trend factor. When running a regression for trend on WPCT and Wins I didn’t feel that either correlation rated high enough from year to year. So, I went back to SRS and found that it had an 86.7 correlation rate from year to year SRS.
5. Once I arrived at projected data for wins and SRS (from both the Compensation and Trend data) I was able to then give a weighted (2:1 Trend) average of each university.
6. These numbers are based off of current head coach salaries as well as five year trend data for SRS statistics. Not all schools had reported HC salary—mostly the private universities—and for them I took simply the weighted average of the two projected win trends.
7. The below chart details a “Top 25” of projected wins based on the noted data pools.

 

Rank	School	Conf.	Weighted Average Wins	Projected Wins (HC Salary)	Projected Wins (Trend)	Projected SRS (HC Salary)	Projected SRS (Trend)
1	Ohio State	Big Ten	12	9	14	10.93	26.69
2	Alabama	SEC	12	10	12	14.25	20.93
3	Mississippi	SEC	11	9	13	7.71	23.86
4	Stanford	Pac 12	11	—	11	—	16.38
5	Clemson	ACC	11	9	12	7.96	19.82
6	Michigan State	Big Ten	11	9	12	8.86	18.63
7	Baylor	Big 12	10	—	10	—	14.67
8	Florida State	ACC	10	9	11	9.48	16.94
9	Texas Christian	Big 12	10	—	10	—	14.36
10	Mississippi State	SEC	10	9	11	7.71	17.27
11	UCLA	Pac 12	10	9	11	7.84	17.15
12	Auburn	SEC	10	9	11	9.48	15.55
13	Southern California	Pac 12	10	—	10	—	12.45
14	Oklahoma	Big 12	10	9	10	10.85	12.68
15	Tennessee	SEC	10	9	10	7.71	13.96
16	Notre Dame	Ind.	10	—	10	—	11.52
17	Georgia	SEC	10	9	10	8.19	13.11
18	Washington	Pac 12	10	9	10	8.19	12.86
19	Utah	Pac 12	9	8	10	5.75	13.91
20	Oregon	Pac 12	9	8	10	3.98	14.34
21	Duke	ACC	9	—	9	—	10.21
22	Louisville	ACC	9	9	9	8.42	11.07
23	Arkansas	SEC	9	8	10	7.20	11.64
24	Memphis	AAC	9	6	11	-1.72	15.75
25	Louisiana State	SEC	9	9	9	10.42	9.14

 

The below chart is a look at how the SEC would fare this year according to this compiled data.

EAST	Weighted Average Wins W’s	Projected Wins (HC Salary)	Projected Wins (Trend)	Projected SRS (HC Salary)	Projected SRS (Trend)
Tennessee	10	9	10	7.71	13.96
Georgia	10	9	10	8.19	13.11
Florida	9	9	9	8.42	8.58
Missouri	8	9	8	7.84	5.18
South Carolina	7	9	6	9.48	-2.20
Kentucky	7	8	6	4.92	-2.02
Vanderbilt	4	—	4	—	-7.52
WEST	Weighted Average Wins W’s	Projected Wins (HC Salary)	Projected Wins (Trend)	Projected SRS (HC Salary)	Projected SRS (Trend)
Alabama	12	10	12	14.25	20.93
Mississippi	11	9	13	7.71	23.86
Mississippi State	10	9	11	7.71	17.27
Auburn	10	9	11	9.48	15.55
Arkansas	9	8	10	7.20	11.64
Louisiana State	9	9	9	10.42	9.14
Texas A&M	9	10	9	11.25	7.83

 

Professional Sports Salaries and Fan Expectations

The other day a good friend sent me a link for an article about the LA Dodgers, Moneyball and how the tactful art of underspending may not be the most effective way of building a title contender.

While the points espoused in the article fall short of becoming gospel, it was successful in getting my statistical juices flowing. The article’s main theme, crudely stated, was essentially: more average money = more average wins. In baseball which has no hard salary cap, the potential exists for large pocketed clubs to overspend and create “stacked” programs, as do teams like the Yankees and Red Sox. In other sports, such as football and basketball, there are more concrete caps on spending.