Sports biz: Hey, Buffaloes: Meet big data

Stewart Schley //January 10, 2013//

Sports biz: Hey, Buffaloes: Meet big data

Stewart Schley //January 10, 2013//

Want to make a few million bucks? Get hired as the University of Colorado’s head football coach. Then get fired.

That’s the path that has enriched the last three coaches who guided the Buffaloes’ football program. Gary Barnett walked away with $3 million when he was fired in 2005. Dan Hawkins collected $2 million when he was shown the door in 2010. In November, CU paid former coach Jon Embree $1.5 million when it terminated his three-year contract a year early.

That’s $6.5 million paid out to three individuals in exchange for NOT coaching the CU football team. Colorado is always looking for growth industries to bring good jobs to our state. By gum, I think we’ve found one.

It wouldn’t be so farcical except for two facts. First, colleges fire football coaches with regularity, not just at CU but across the college gridiron kingdom. From 1997 through 2010, 117 Football Bowl Subdivision universities – we used to call them Division 1-A teams – gave the axe to coaches on 150 different occasions for performance reasons. That’s according to a provocative study published in October in the academic journal Social Science Quarterly.

Second, as the same study indicates – and this is where that $6.5 million starts to sting – replacing coaches doesn’t make college football teams any better than teams that stick with their existing coach.

You read that right. As tempting as it is go invest great hope in the next new coach, an exhaustive analysis of data tracking 1,643 cases over a 14-year period shows that coaching replacements had little impact for poor-performing teams compared to other struggling teams that didn’t replace their coach. For average teams, the study found replacing a head coach actually resulted in worse performance compared to teams that kept their coach.

Ouch.

It so happens that the study’s lead author is employed by CU. E. Scott Adler, an associate professor of political science at CU’s Boulder campus, is a University of Michigan graduate and a Wolverines football fan who specializes in the study of organizational leadership succession. His new book, “Congress and the Politics of Problem-Solving,” examines how the electoral replacement of lawmakers does (or doesn’t) impact policy change.

Adler notes there are often extenuating influences that lead to a coaching change, like sagging alumni support or luxury boxes that go empty. “It’s very hard for universities like CU to operate without that sort of revenue certainty,” he acknowledges.

But from a win-loss standpoint, the record doesn’t support the notion that teams are one great leader away from turning things around. “It’s just not in the data,” Adler told me a couple of Saturdays ago. He thinks a big-data injection is overdue for a world of sports that has long operated on inspired hunches and intuition.

For CU fans suffering Buffalo buzz kill at the idea that the new coach won’t produce a miracle, there may be some modest hope in the study Adler conducted with colleagues from CU/Denver and Loyola University. With a 4-21 record over the last two seasons, CU clearly falls into the “poor performing” category, suggesting the university may see a slight improvement after its coaching change. The problem is that there’s no evidence to suggest a poor performing team that replaces its coach will do better than a poor performing team that doesn’t replace its coach.

That’s the dismaying subplot here. Spending millions of dollars to terminate coaching contracts makes sense if it results predictably in on-field improvements. But the data says it doesn’t. CU athletic director Mike Bohn may have other, legitimate reasons for wanting to see a new coach running the team – a fundraising revival, an obligation to the Pac-12 conference or, what the hey, just a better vibe at pre-game tailgate parties. But if it’s more wins the team is after, the record suggests Embree would have done just as good as the new guy. And CU would have saved $1.5 million.