By Quentin Proud '26 and Grayson Segars '26
Introduction
On December 4th, 2017, Jimbo Fisher shocked the college football world by leaving Florida State to sign a 10-year, $75 million fully guaranteed contract to become the next head coach at Texas A&M. Fisher, just a few years removed from back-to-back college football playoff appearances and a national championship, was viewed by the Texas A&M’s administration as the man who would finally help the school’s football program make the jump from “great, but not elite, program” to “perennial championship contender”. In Fisher’s introductory press conference, Texas A&M Athletic Director Scott Woodward would claim that Fisher “is the right coach at the right time...” and that he was “confident that the opportunity for greatness is right around the corner.” Fisher echoed this optimistic tone, saying he would “try to take Texas A&M where it should be at the top of the college football world.”
Despite the high hopes for Fisher’s Texas A&M turnaround, his 6-year tenure led to zero double-digit win seasons and only one New Year's Six Bowl Appearance (during the pandemic-shortened 2021 season). Even worse, Fisher’s average annual win total of 7.67 was fewer than the 8.5 his (much cheaper) predecessor Kevin Sumlin posted. Finally, in the midst of his 5-7 2023 season, new Texas A&M Athletic Director Ross Bjork listened to Texas A&M’s fans’ wishes and fired Fisher. In order to void the remainder of his contract, Texas A&M paid Fisher a record $80 million dollar buyout. Fisher’s once-celebrated hire had gone up in expensive flames.
Texas A&M’s failed hire of Jimbo Fisher is not the only example of highly touted coaches being handed massive buyout checks. In 2023, a record $118 million was paid in buyouts to fired head coaches, smashing the previous record of $91 million in 2021. Is there a way to predict which coaches will pan out and which will cost the university and its boosters millions? Newly-hired head coaches typically come from a few select backgrounds: former NFL player, former NFL/NCAA Power 5 offensive or defensive coordinator, former NFL/NCAA Power 5 head coach, a prior connection with the program, or a mix of several of these backgrounds. Are coaches from some of the backgrounds more likely to be successful?
Methods
Our process to determine the success of certain backgrounds in coaching hires began by determining how successful these coaching tenures were. We used game results and recruiting rankings data going back to 1980 in order to analyze the success of all coaching tenures from 1990 up to and including 2021. In order to quantify the extent to which these tenures were successful, an algorithm was developed that gave each coaching tenure a score.
The goal of this algorithm was to capture the success a coach brought to a program relative to that of the program he inherited; 10 wins at Vanderbilt is not considered the same as 10 wins at Alabama. To this end, we used two proprietary metrics, Win Percentage above program Average (WPA) and Roster Strength above program Average (RSA).
WPA takes the average win percentage over a coaching tenure and subtracts from it the average win percentage of the program from the decade prior to the coaching hire (ex. Nick Saban’s win percentage at Alabama was compared to Alabama’s win percentage from 1997 to 2006). This number is multiplied by 100 to calculate the WPA. RSA takes the final roster strength of a coaching tenure, found by first averaging the recruiting rankings of the three years previous (a similar idea to 247’s Blue Chip Ratio). This number is subtracted from the average recruiting ranking from the decade prior to the coaching hire, just like WPA. The goal of this metric is to indicate the state of the roster left behind by the coach relative to the historical standard. RSA is incredibly useful in calculating the “unrealized” wins/losses of a coaching tenure by instead using the strongest predictor of future success in college football, recruiting. In our algorithm, we only use it to account for the two years following the tenure (thus, the denominators of the coefficients of WPA and RSA being the length of the tenure and 2 years) and not any years during the tenure because recruiting rankings are only useful when win/loss is not available, in our opinion. Why does one recruit but to win?
Finally, we added incentives for two undeniable markers of success, conference championships and national championships. These incentives help account for the added difficulty of continued improvement above a certain win threshold in college football. It is much harder to make the jump from 9-2 to 12-0 than from 2-10 to 5-7, and so the 12-0 team should be rewarded for achieving certain milestones that a 9-3 team may not be able to.
Results
| 10 Most Successful Coaching Tenures Since 1990 |
| 1. Nick Saban at Alabama, started in 2007: 59.33333333333333 |
| 2. Pete Carroll at USC, started in 2001: 37.05181818181819 |
| 3. Bob Stoops at Oklahoma, started in 1999: 35.37233333333333 |
| 4. Art Briles at Baylor, started in 2008: 34.97466666666667 |
| 5. Dabo Swinney at Clemson, started in 2009: 31.75555555555556 |
| 6. Dennis Erickson at Oregon State, started in 1999: 30.842222222222226 |
| 7. Brian Kelly at Cincinnati, started in 2007: 27.804666666666677 |
| 8. Mike Riley at Oregon State, started in 1997: 27.302083333333336 |
| 9. Chris Petersen at Washington, started in 2014: 26.96333333333334 |
| 10. James Franklin at Vanderbilt, started in 2011: 25.91 |
| 10 Least Successful Coaching Tenures Since 1990 |
| 328. Rich Rodriguez at Michigan, started in 2008: -21.45466666666667 |
| 329. David Beaty at Kansas, started in 2015: -21.68666666666667 |
| 330. Gary Andersen at Oregon State, started in 2015: -22.70533333333333 |
| 331. Kevin Steele at Baylor, started in 1999: -22.992222222222225 |
| 332. John Bunting at North Carolina, started in 2001: -23.074166666666663 |
| 333. Paul Wulff at Washington State, started in 2008: -23.709999999999994 |
| 334. Darrell Hazell at Purdue, started in 2013: -24.254444444444445 |
| 335. Tyrone Willingham at Washington, started in 2005: -24.742222222222217 |
| 336. Greg Robinson at Syracuse, started in 2005: -28.36 |
| 337. Chris Ash at Rutgers, started in 2016: -29.02 |
The backgrounds we tested to see if they contributed to increased success of coaches were: previous P5 head coaching experience, previous P5 coordinator experience, previous tie to the university as either a player or a coach, previous NFL head coaching experience, previous NFL coordinator experience, played in the NFL, and comes from an offensive or defensive background. To test these factors, we used the point biserial correlation test, with our coaching score as the continuous variable and each of the above factors as the dichotomous variables. The results are as follows:
Point-biserial correlation for Power 5 Coach: -0.01 (p-value: 0.90)
Point-biserial correlation for Offense/Defense: 0.06 (p-value: 0.54)
Point-biserial correlation for Tie to School: -0.07 (p-value: 0.47)
Point-biserial correlation for NFL Coach: 0.20 (p-value: 0.04)
Point-biserial correlation for NFL Player: -0.03 (p-value: 0.75)
Point-biserial correlation for NFL Coordinator: 0.09 (p-value: 0.35)
Point-biserial correlation for Power 5 Coordinator: 0.08 (p-value: 0.44)
Being a former NFL head coach is clearly the strongest predictor according to our testing, followed by being an NFL coordinator and being a Power 5 coordinator. Being a former NFL head coach was the only one of these factors that reached statistical significance, positive or negative.
Conclusions
While NFL head coaching experience was our most statistically significant predictor of coaching success, there is a lot to glean from the rest of these results as well. One interesting result was that of whether or not a coach had a previous tie to the school. Conventional wisdom surrounding coaching hires would suggest that having a tie to the school, or more broadly being a “culture fit” is critical to success as a head coach. A recent high-profile example of this was the hire of Brian Kelly by LSU. Critics suggested that his lack of experience in the south would make a poor hire, despite his proven track record. Kelly won the SEC West and beat Alabama in his first year at LSU. It wasn’t the first time LSU had taken a coach without roots in the south. In 2000, they hired a coach by way of Michigan State who had never coached a team below the Mason-Dixon line. His name was Nick Saban, and he won LSU their second national championship ever in 2003. Our findings would suggest that having ties to a school, or being a “culture fit” is not an important factor.
Further, trends in professional football over the last half-decade suggest that hiring offensive-minded coaches is ideal because retaining the team’s offensive play caller (rather than seeing a defensive-minded coach’s offensive coordinator leave for a head-coaching job) is of great importance. However, we found no statistical significance to hiring an offensive-minded coach as opposed to a defensive coach on the college level. The top three coaches in our model, Nick Saban, Pete Carroll, and Bob Stoops, all came from defensive backgrounds. At the same time, Chris Ash and Greg Robinson, our model’s bottom two coaches, also came from defensive backgrounds.
Lastly, it is generally believed that NFL coaches, used to accommodating older, professional players, can struggle to successfully coach college athletes. NFL players are viewed as being more mature and self-regulating whereas college athletes may require firmer discipline from their coaches. The role of a college football coach is more “CEO-like” than that of a pro coach. College coaches handle scouting, recruiting, and roster management while NFL personnel decisions are typically handled by the front office – though often through collaboration with the head coach. Despite these differences, College hires with an NFL background tend to perform better than those without one. This could be because former NFL coaches are generally more qualified than college coaches – given that the NFL has fewer teams and the jobs are generally considered more prestigious – and thus are comparatively better coaches than those from other backgrounds.
References
Anonymous. (2022, July 20). Recruiting Database: 1980-2022. The Key Play.
https://www.thekeyplay.com/content/2022/july/20/recruiting-database-1980-2022
Conway, T. (2018, January 5). Jimbo Fisher named Texas A&M HC after 8 years at
FSU; contract reportedly record. Bleacher Report. https://bleacherreport.com/articles/2746231-jimbo-fisher-reportedly-named-texas-am-hc-after-8-years-at-fsu
Jimbo Fisher: Introductory press conference. Texas A&M Athletics - 12thMan.com.
(2017, December 4). https://12thman.com/news/2017/12/4/football-jimbo-fisher-introductory-press-conference
Rumsey, D. (2023, November 27). Schools owe fired CFB coaches a record $118m in
buyouts (so far). Front Office Sports., https://frontofficesports.com/schools-owe-fired-cfb-coaches-a-record-118m-in-buyouts-so-far/
Texas A&M Aggies College Football History, stats, records: College Football at
Sports. Reference.com. (n.d.). https://www.sports-reference.com/cfb/schools/texas-am/index.html