The Technology of Coaching Searches
Stephen Prather loves to hear the reasons athletic directors give for hiring football coaches. Prather laughs especially hard when an AD raves about a coach's energy and enthusiasm. "If you want a guy with energy and enthusiasm, well, the guy you fired probably had that, too," the co-founder of SportSource Analytics says. "And what about Bill Snyder? Would he strike you as an enthusiastic, energetic guy?"
By bringing up the coach who, when adjusted for degree of job difficulty, might be the greatest of all time, Prather drives home his point. Athletic directors often make decisions with millions of dollars and dozens of jobs (including the AD's, ultimately) hanging in the balance based on gut feeling and a set of abstract criteria for which there is no empirical data. "You're not looking to hire the guy that energies well," Prather says. "You're looking to hire the guy that coaches well."
Prather and his three colleagues at SportSource would like to change that. They aren't trying to replace gut instincts or search firms in college football coaching searches, but they are offering ADs hard data that can help narrow their candidate pool or supplement their interview process. For a fee—Prather wouldn't divulge specifics, but he said it's a fraction of what most search firms charge—athletic directors can have access to a coach search tool designed by the same company that provides analytics for the College Football Playoff selection committee. Prather believes his company's service can help ADs move beyond the abstract to make the best possible hire. What is the best possible hire? That's pretty simple, actually.
"Your goal," Prather says, "ought to be to hire the coach who can win the most games who will say yes."
Does an AD want a sitting mid-major head coach or a hot Power Five coordinator? He or she can use this tool to see which types have had more success moving to similar programs. If it's a coordinator, should it be an offensive or a defensive one? If it's a head coach, do any coaches in the candidate pool suffer from what Prather calls Coordinator Dependency Syndrome? Has the coordinator's offense trended up or down? Did the sitting head coach have a great season because he built a sustainable program or because he lucked into one great group of seniors? SportSource has created a web-based tool that can slice and dice all that data from any web browser.
Courtesy of SportsSource Analytics
ADs could try to find all this information on their own, but most don't have time and might have to pay their staff about what they would pay SportSource to collect it. They won't get that info from a search firm, either. Those companies serve to check candidates' backgrounds and allow for plausible deniability as schools gauge potential candidates' interest through back channels. With 11 FBS jobs already open and more likely to come open soon—hi, Rutgers and Virginia—subscribing to the SportSource service could be a valuable investment for ADs trying to create a pool of candidates from a group that has a few obvious names at the top but becomes murkier the deeper it goes. "You saw a lot of good up-and-coming coordinators get hired in the last two years," Prather says. "The pool is definitely different. It's not as big. You could see some guys get high-level jobs and you'll be like, 'Whoa. Really?'"
SportSource, a four-year-old company whose employees live in Atlanta and Nashville, Tenn., remains a second job for its principals. Prather, for example, also sells commercial real estate. The company's primary business is selling analytics packages to football programs to help coaches parse data to determine how they can win more games. SportSource also sells similar packages to agents who represent coaches. The idea for the coaching search tool came about during a meeting in 2014 between Prather and one of the firm's first clients. Patrick Strong, an agent who teams with partner Russ Campbell to represent coaches for Balch Sports in Birmingham, Ala., suggested a tool that would help agencies better use data to promote the coaches they represent.
"The data helps bring into focus things like: (1) The value of coordinating/head coaching/overall coaching experience in a head coaching search and (2) How much current position, age, and ties to the hiring school move the meter for decision-makers," Strong wrote in an email. "We pair the hiring data with other proprietary market-based tools we've developed to eliminate some of the unknowns in what can be a stressful process that often unfolds in a tight window."
Since SportSource developed the platform, Strong and Campbell have found several useful ways to employ the data. "We've used the data to respond to a school's initial hiring criteria—both in support of and in challenge to their approach," Strong wrote. "We've used the data to distinguish final candidates at the end of a search and even in the off-season where a school thinks their coach may be in play in the near future. The bottom line is that our analytics help us prepare for varying scenarios, giving our clients a competitive edge. We don't guess, we show."
Prather and partners Drew Borland, Marty Couvillon and Scott Prather thought the idea could also work for the athletic directors on the other side of the hiring process. After they created the platform last year, three schools (Houston, Oregon State and SMU) used it to aid in their coaching searches. SMU's choice of Chad Morris seemed obvious. Morris is a former Texas high school coach who had great success as a coordinator at Clemson running an offense similar to what most of the players SMU recruits run in high school. Gary Andersen was a shocking choice at Oregon State, because he already held what is perceived nationally to be a better job at Wisconsin. Bob De Carolis, Oregon State's AD at the time, was sifting through a pile of potential coaches but kept hearing Andersen might be open to a discussion. "A lot of it, we figured, was probably rumors," De Carolis told ESPN.com's Kevin Gemmell. "But you talk to people and you hear things. Maybe things aren't as rosy as they should be. A lot of it was probably rumors, but you follow up on it to make sure if there is some truth to it." It became clear Andersen had interest, and Oregon State had its man.
Houston's search, however, was wide open. A lot of coaches wanted the job because it had launched Art Briles and Kevin Sumlin into Power Five gigs. Tony Levine hadn't worked out in the position, but he hadn't run the program aground, either. The Cougars still had good players capable of winning and a brand new on-campus stadium. Then-AD Mack Rhoades had plenty of options, and he used the SportSource tool along with his own contacts to help narrow those options. Rhoades eventually chose to hire Ohio State offensive coordinator Tom Herman. Rhoades wasn't necessarily interested in hearing about potential candidates who might also happen to be SportSource clients, but the Houston group loved using the data tool to test theories and sharpen their focus on particular candidates. The data—especially the numbers for first-time head coaches—were also useful after Herman was identified to help Rhoades sell his choice to search committee members. Herman is now 10–0 at Houston after Saturday's comeback 35–34 win over Memphis.
Thearon W. Henderson/Getty Images
Rhoades, meanwhile, only worked with Herman briefly before leaving to become the athletic director at Missouri. When Rhoades took the job, he probably thought he wouldn't have to search for a new football coach for a long time. Gary Pinkel was coming off consecutive SEC East titles in 2013 and '14, and Pinkel's status as the man who revolutionized Missouri football was secure. But life isn't fair. Pinkel was diagnosed with Non-Hodgkin's Lymphoma earlier this year, and he announced last week that he will retire at the end of this season. So, Rhoades must now seek another coach.
Rhoades has already examined the data carefully, so he knows how many myths it can bust. But Prather hopes other ADs will soon learn the same thing. During a recent demonstration for the tool, Prather offered to test the well-worn trope that a coach who attended a particular school is better equipped to succeed at that school than a coach who matriculated somewhere else. "They understand the culture," boosters and fans will say. But does that help them win more? Prather asked the tool to show all of the coaches hired between 2001-14 who played for or graduated from the schools that hired them. There were 15. Only four (Ralph Friedgen at Maryland, Mike Gundy at Oklahoma State, Pat Fitzgerald at Northwestern and David Shaw at Stanford) registered better winning percentages than the programs had posted in the previous five years. Prather then pointed out something else. Gundy, Fitzgerald and Shaw had something in common. They were internal hires. Gundy was promoted after Les Miles went to LSU. Fitzgerald was promoted after Randy Walker died. Shaw was promoted after Jim Harbaugh went to the San Francisco 49ers. They had all taken over programs that were already reasonably successful. So, did these coaches win because they were good fits who understood the culture of the school back to their undergraduate days? Or did they win because they were good coaches inheriting good staffs and good players who understood the (more recent) culture of success by virtue of already working for these teams?
Prather knows Harbaugh's name will come up as evidence that an alumnus can improve a program, but Prather prefers to look at Harbaugh's more recent data. "They didn't hire Jim Harbaugh because he went to Michigan," Prather says. "They hired him because he's a good football coach."
Sometimes, ADs make their choice too complicated. They worry about whether a coach will be a good public face of the program. They worry about whether he can help raise money. They worry about his institutional fit. They worry about winning the press conference. "You can go 1–0 in the press conference. That's all you can ever do," Prather says. "The second you kick off, all of that stuff is worthless." Prather doesn't understand why some ADs can't see that coaches who win games are most likely to make fans happy and help raise money. When considered that way, the criteria are fairly simple to determine.
Prather doesn't believe hard data should replace an AD's gut instinct completely, but he does think data can help narrow the focus and sharpen that instinct. It can help an AD ask concrete interview questions about topics that actually affect winning. That way, when an AD finally does pick a coach, it will be easier to select the right one: the person who will win the most games and say yes to the job.