Is it the coach or the program? Ranking CFB coaches while factoring in expectations

Back in May, ESPN's team of college football reporters voted on the sport's best coaches for 2025. The results were about as you would expect: Start with the three active guys who have most recently won national titles (Georgia's Kirby Smart, Ohio State's Ryan Day, Clemson's Dabo Swinney), move on to guys with recent top-five finishes or national title game appearances (Notre Dame's Marcus Freeman, Texas' Steve Sarkisian, Oregon's Dan Lanning, Alabama's Kalen DeBoer, Penn State's James Franklin), then squeeze in a couple of long-term overachievers at the end (Utah's Kyle Whittingham, Iowa State's Matt Campbell).

The rankings made plenty of sense, but I couldn't help but notice that the top eight coaches on the list all work for some of the richest, most well-supported programs in the country. There are some epic pressures associated with leading these programs -- just ask Day -- but there are also major advantages. It might only take a good head coach to do great things in those jobs, while at programs with smaller alumni bases or lesser historic track records, it might take a great coach to do merely good things. They're such different jobs that it's almost impossible to even know how to compare the performance of, say, Matt Campbell to Steve Sarkisian. Could Campbell have led Texas to back-to-back CFP semifinals? Could Sark have brought ISU its first two AP top-15 finishes?

The May rankings made me want to see if there were a way to apply stats to the conversation. If you think about it, we're basically measuring two things when we're gauging coach performance: overall quality and quality relative to the expectations of the job. I thought it would be fun to come up with a blend of those two things and see what the results told us.

Performance versus expectation

Gauging overall performance is easy enough. You could simply look at win percentage, and it would tell you quite a bit. From 2015 to 2024, the active coaches with the best FBS win percentages (minimum 30 games) were Day (.870), Lanning (.854), Swinney (.850) and Smart (.847). All ranked high in the May rankings. I tend to want to get fancy and use my SP+ ratings whenever possible, and they tell a similar tale. Looking at average SP+ ratings for the past decade, the top active coaches are Day (30.4), Smart (27.0), Lanning (22.3), Swinney (21.9), Franklin (20.3) and Freeman (19.0). They're all in the May top 10 too.

Again, though, all of those coaches are employed by college football royalty. (Granted, Swinney gets bonus points for helping Clemson turn into college football royalty, but still.) Isn't it more impressive to win 11 regular-season games at Indiana, as Curt Cignetti did in 2024, than to go 10-4 like Swinney did? Isn't it probably harder to finish 12th in SP+ at SMU, as Rhett Lashlee did in 2024, than to finish fifth like Franklin did?

I've begun to incorporate teams' performance against long-term averages into my preseason SP+ projections, and it seems we could use a very similar concept to evaluate coach performances. For each year someone is a head coach, we could compare his team's SP+ rating for that season to the school's average from the 20 previous years. (If the school is newer to FBS and doesn't have a 20-year average, we can use whatever average exists to date. And for a program's first FBS season, we can simply compare the team's SP+ rating to the overall average for first-year programs.)

By this method, the 10 best single-season coaching performances of the past 20 years include Art Briles at Baylor in 2013-14, Jim Harbaugh at Stanford in 2010, Mark Mangino at Kansas in 2007, Bobby Petrino at Louisville in 2006, Greg Schiano at Rutgers in 2006 and Jamey Chadwell at Coastal Carolina in 2020 -- legendary seasons of overachievement -- plus perhaps lesser-remembered performances such as Gary Andersen at Utah State in 2012, Matt Wells at Utah State in 2018 and Brian Kelly at Cincinnati in 2007.

As far as single-season overachievement goes, that's a pretty good list. And if we look at a longer-term sample -- coaches who have led FBS programs for at least nine of the past 20 years -- here are the 15 best performance versus baseline averages.

(Note: I'm looking only at performances within the past 20 years, so Nick Saban's work at LSU (2000-04) or Michigan State (1995-99), for instance, isn't included. I also went with nine years instead of 10 so Smart's current nine-year run at Georgia could be included in the sample.)

Best performance vs. historic baseline averages for the past 20 years (min. nine seasons):

1. Chris Petersen, Boise State (2006-13) and Washington (2014-19): +12.8 points above historic baseline

2. Art Briles, Houston (2005-07) and Baylor (2008-15): +12.8

3. Gary Pinkel, Missouri (2005-15): +12.5

4. Nick Saban, Alabama (2007-23): +10.7

5. Jeff Monken, Army (2014-24): +10.3

6. Willie Fritz, Georgia Southern (2014-15), Tulane (2016-23) and Houston (2024): +10.0

7. Lance Leipold, Buffalo (2015-20) and Kansas (2021-24): +9.5

8. Bobby Petrino, Louisville (2005-06), Arkansas (2008-11), Western Kentucky (2013) and Louisville (2014-18): +9.5

9. Gary Patterson, TCU (2005-21): +8.6

10. Jim Harbaugh, Stanford (2007-10) and Michigan (2015-23): +8.5

11. Blake Anderson, Arkansas State (2014-20) and Utah State (2021-23): +8.5

12. Steve Spurrier, South Carolina (2005-15): +8.2

13. Greg Schiano, Rutgers (2005-11 and 2020-24): +7.8

14. Jeff Brohm, Western Kentucky (2014-16), Purdue (2017-22) and Louisville (2023-24): +7.7

15. David Cutcliffe, Duke (2008-21): +7.7

If we are looking for pure overachievement and aren't in the mood to reward coaches for winning at schools that always win, this is again a pretty good list. Petersen was spectacular at both Boise State and Washington, while Briles, Pinkel, Monken and Patterson all won big at schools that hadn't won big in quite a while. (Monken, in fact, is still winning big.) Blake Anderson's presence surprised me, but most of the names here are extremely well regarded. And Saban's presence at No. 4, despite coaching at one of the bluest of blue-blood programs, is a pretty good indicator of just how special his reign at Alabama was.

Still, looking only at performance against expectations obviously sells coaches like Saban and Smart short. Saban is probably the best head coach in the sport's history but ranks only fourth on the above list. Meanwhile, Smart has overachieved by only 6.0 points above the historic baseline in his nine seasons at Georgia thanks to the high bar predecessor Mark Richt set. But he has also won two national titles, overcoming Georgia's history of falling just short and at least briefly surpassing Saban as well. If our goal is to measure coaching prowess, we need to account for raw quality too.

The best coaches of the past 20 years

If we combine raw SP+ averages with this performance versus baseline average, we can come up with a pretty decent overall coach rating. We can debate the weights involved, but here's what an overall rating looks like if we use 60% performance versus baseline and 40% SP+ average:

I always like to say that numbers make great starting points for a conversation, and this is a pretty good starting point. Anyone reading this would probably tweak this list to suit their own preferences, and while it probably isn't surprising that Pinkel is in the top 20, seeing him fourth, ahead of Meyer, Harbaugh and others, is a bit jarring. (I promise that this Mizzou alum didn't put his finger on the scales.) Regardless, this is a fun mix of guys who won big at big schools and guys who won pretty big at pretty big schools. That was the goal of the exercise.

Maybe the most confusing coach in this top 20 is Dabo Swinney. Clemson had enjoyed just one AP top-five finish in its history before he took over 16 years ago, and he has led the Tigers to 2 national titles, 6 top-five finishes and 7 CFP appearances. And while they haven't had a true, title-caliber team in a few years, they've still won two of the past three ACC crowns. How is he only 10th?

The main culprit for Swinney's lower-than-expected ranking is his recent performance -- it has been inferior to both national title standards and his standards. Since we're using a team's performance against 20-year averages, a lot of this rating is basically comparing Swinney to himself, and he hasn't quite measured up of late.

From 2012 to 2020, Swinney's average rating was an incredible 17.0, which would have ranked second to only Saban on the list above. But his average over the past four seasons is only 3.6.

Part of what made Saban so impressive was how long he managed to clear the bar he himself was setting in Tuscaloosa. Per SP+, his best team was his 14th -- the 2020 team that won his sixth and final title at Bama. While Swinney was basically matching Saban's standard 12 years into their respective tenures, Saban continued at a particularly high level for at least three more years while Swinney fell off the pace.

Comparing Saban, Swinney and Smart year by year, we see that Smart was hitting Saban-esque levels seven seasons into his tenure, but his rating has fallen off each of the past two seasons. Even Saban slipped starting in Year 15, even though he still had nearly the best program in the sport for a couple more years.

The best coaches of 2025

Six of the top seven coaches on the list above are either retired or coaching in the NFL now, so let's focus our gaze specifically on the guys who will be leading college teams out onto the field in 2025. Using the same 20-year sample as above -- which cuts off the tenure of Iowa's Kirk Ferentz but includes everything else -- here's how the current crop of FBS head coaches has performed at the FBS level. We'll break this into two samples: the guys who have coached for at least four years in this sample and the guys who have coached between one and three years.

Our May top 10 list featured eight guys who have been head coaches for at least four years; all eight are represented on this list, including four of the top five. (Sarkisian has averaged a 13.8 rating over the past two seasons, which is a top-five level, but his overall run as head coach at Washington, USC and Texas has featured a number of ups and downs.)

Maybe the name that jumps out the most above is Josh Heupel. I think anyone would consider him a very good coach (he's 37-15 overall), but he doesn't exactly draw any "best in the game?" hype. He benefited from a positive situation at UCF, where he inherited a rising program from Scott Frost in 2019 and produced big ratings in his first couple of years on the job. But his average rating at Tennessee has been a solid 14.0 as well; the Volunteers had been up and down for years, but he has produced four top-20 SP+ ratings in a row and two top-10s in the past three years. He might not be getting the credit he deserves for that.

All in all, I enjoy this list. We've got mostly predictable names at the top, we've got some oldies but (mostly) goodies spread throughout, and we've got room for up-and-comers like Jeff Traylor too. This 60-40 approach probably doesn't give enough respect to the Chris Creightons of the world -- the Eastern Michigan coach has overachieved against EMU's baseline by 7.2 points per season, which is a fantastic average, but at such a hard job, his Eagles have still averaged only a minus-14.4 SP+ rating during his tenure. Still, this is a mostly solid approach.

Now let's talk about some small-sample all-stars.

Four of the top six of this list coached in the College Football Playoff last season, and while the guys ranked fifth and sixth made our May top 10 list, the guys who won big at SMU and Indiana, not Oregon and Notre Dame, take priority here. I was honestly floored that Curt Cignetti didn't make our top 10 list; he led James Madison to one of the best FBS debuts ever, going 19-4 in 2022-23, then he moved to Bloomington and led Indiana -- INDIANA! -- to 11 wins in his first season there.

On this list, however, Rhett Lashlee tops even Cignetti. I'm not sure we've talked enough about the job he has done at SMU. He, too, inherited a rising program, as Sonny Dykes had done some of the nitty-gritty work in getting the Mustangs back on their feet (with help from an offensive coordinator named Rhett Lashlee). SMU hadn't produced a top-50 ranking since 1985 before Dykes did so for three straight seasons (2019-21). But after holding steady in his first year replacing Dykes, Lashlee's program has ignited: 12-2 and 24th in SP+ in 2023, then 11-3 and 12th in 2024. Looking specifically at the 2021-24 range, as the game has undergone so much change, Lashlee's 16.8 average rating ranks second overall, behind only Smart (18.0) and ahead of Kiffin (15.1), Cignetti (15.0), Odom (15.0), Heupel (14.0) and Day (13.9).

Along with quite a few others here, Lashlee made my 2024 list of 30 coaches who would define the next decade; he'd definitely still be on the list -- along with new additions like GJ Kinne and perhaps Fran Brown -- if I remade that list today.