When a coach changes an offensive system: It's all coming back to me now

The Head Coaching Carousel in college football never stops. Programs are always trying to become more successful and because players can’t stay in college forever, it’s common sense to invest a lot of time and money in to the right head coach. Yet, the coaching profession is mired in nepotism, favoritism and a strong aversion to proof of any kind as to how good a coach is beyond the Win column.

I wanted to take a data-driven look at the most common unanswered questions around coaching changes, specifically looking at what can happen when a new head coach brings a new offensive system. Does the offense actually get better?

I started with the head coaching changes in 2010, so that I would have at least 3 seasons of data to compare. Next, I only looked at programs where the offensive scheme fundamentally changed. For instance, Notre Dame under Charlie Weis ran basically a "pro-style" offense, and when Brian Kelly came in, the offense became more of a spread, with an emphasis on spreading to run, similar to what Kelly did at Central Michigan and Cincinnati. Now, I wasn’t there installing plays and haven't watched all the offensive film to discern this, so if there’s a glaring weakness in my analysis it’s this: My "system changes" coding is crude at best.

I ignored coaching changes like Butch Jones’ move to Cincinnati (Jones worked under Kelly at Central Michigan), where the basic offensive scheme stayed the same despite the coaching change. This would rule out the Pete Carroll-to-Lane Kiffin transition at USC too, with Kiffin tinkering rather than overhauling the Trojans' offense. If I only wanted to look at how, for instance, Kiffin’s coaching abilities stacked up against Carroll’s, I might do this kind of analysis for programs where some offensive continuity was maintained.

I also ignored places like Kansas, Memphis and Akron, where the coaches brought in in 2010 didn’t even last until this offseason. It’s safe to assume that they did not perform well enough to keep their jobs. Ditto Derek Dooley at Tennessee, who lost his job late in the 2012 season after an 0-7 record in the SEC. What I was left with were five cases that provide a cross section of what can happen when the new coach brings a new ball system.

I used Offensive Fremeau Effeciency Index (OFEI) and F/+ rankings from Football Outsiders to compile this data.

Spread to Pro

When Mike MacIntrye was hired to head up San Jose State in 2010, he had never been a head coach before and had spent nearly his entire coaching career working with defensive backs and on the defensive side of the ball. Add to that a stated preference for going back to a pro-style offense instead of the spread he found there, and San Jose State has to be low on the list of expected turnarounds, right?

First, let’s establish just how bad a team MacIntrye inherited. From 2007 to 2009, SJSt was ranked 117th on OFEI, 110th in Offensive S&P+; it’s not as if they had a dominant defense either -- overall they ranked 113th. Not one offensive metric was in the top 80 of NCAA FBS programs.

Macintrye improved the team right away, SJST jumped up 19 spots to 98 in OFEI in 2010. 2011 was a sophomore slump of sorts, but not a wholesale disaster as 2011 was still a better year according to OFEI and F/+ than 2007, 2008 or 2009. And last year, not only did SJST find itself 32nd in OFEI, it was second in Explosive Plays. In three years, the Spartans went from 118th in that same category to second, earning MacIntrye the prize of getting to overhaul a moribund Colorado Buffaloes program.

David Fales, junior QB, put up 4,193 yards and 33 touchdowns to only nine interceptions, and De’Leon Eskridge, their 1,000 yard running back, had never topped 700 yards before. Hats off to MacIntrye for coaching up the talent he had available.

Pro to Spread

Any conversation about Notre Dame’s return to relevancy will feature Heisman finalist Manti Teo and a defense that, except against Alabama, bent and didn't break. However, that would only be half of the story. In 2009, Notre Dame was led by All-Americans Jimmy Clausen (quarterback), Golden Tate (receiver) and Eric Olsen (center), and all three left for the NFL’s greener pastures after an ignominious 6-6 season. Upon Brian Kelly’s arrival, the top of the offensive depth chart had Dayne Crist as the starting QB, a player so uniquely gifted at getting benched that he managed that feat at perennial big 12 doormat Kansas after losing his job at N.D.

To say that the cupboard was bare might be an overstatement, but Kelly certainly had his work cut out for him in 2010. Not surprisingly, the Irish regressed that fall, from third in OFEI to 36th. As Kelly was able to recruit more players appropriate to his scheme, Notre Dame improved each year, especially in methodical drives of 10 or more offensive plays. Last year, ND finished 12th in OFEI and 14th in Methodical Drives, an offensive juggernaut that wore teams down and kept that vaunted defense fresh.

We don’t know what next year will hold for the Fighting Irish; what we do know is that 2009 was an outlier - in 2007, 2008 and 2009, Weis’ teams averaged 60th in OFEI, while Kelly’s have averaged 25th since.

Spread to AirRaid

Between 2010 and 2012, Sonny Dykes, brought with him the explosive Air Raid offense he gleaned at Hal Mumme's and Mike Leach’s tutelage to Louisiana Tech, which had been previously coached by Derek Dooley.

Louisiana Tech improved steadily under Dooley, from 113th in OFEI in 2007 to 85th in 2009, but the Bulldogs took off under Dykes. In 2010, the program had to adjust to the Air Raid and only managed to scrape 92nd in OFEI but improved 20 spots in Methodical Drives and 29 spots in S&P+ and value drives (drives that start on the offense’s side of the field and move inside the opponent’s 30). That momentum picked up in 2011 and culminated in 2012, with Louisiana Tech finishing at 11th in OFEI and 20th in Explosive drives, with a net plus-23.7% improvement from 2007 in Off F/+.

Coach Dykes now takes his offensive philosophy to Cal. It will be very interesting to see how Cal’s performance improves over three seasons as it deals with better talent in the Pac-12.

Pro to Spread Part 2 and Pro to AirRaid

Charlie Strong is rightly noted as a defensive genius, pioneering wider use of the 3-3-5 to beat more pass happy offenses with inferior size. Ruffin McNeal of East Carolina brought the AirRaid to the Pirates to simplify and overhaul a broken pro style system. I’ve lumped these two cases together because neither Louisville nor East Carolina is, on average, all that much better than before. They are indeed better, yes, but Louisville averages only eight spots better in Offensive Efficiency and Available Yards, 13 spots worse on Explosive Drives and nearly the same on Methodical Drives. It’s clear that Strong’s defensive coaching abilities and the outstanding play of quarterback Teddy Bridgewater lift this middling program from an average of 55th OFEI over 2007-2009 to 30th, not the system itself.

In East Carolina’s case, too, the only consistent result over McNeal’s tenure has been to make drives more methodical, going from an average rank of 90th to an average rank of 39th in the last three seasons. In fact, most of this improvement is from 2010, when East Carolina jumped to 23rd overall in OFEI. If anything, ECU has regressed as a program.

Lessons to be learned

In the best case, changing an offense means immediate impact -- three of the five teams I looked at in depth improved in the first year of a new scheme. But not all of those teams actually got better in the following years, with only Louisville and San Jose St. improving each year compared to the previous three. Where it gets really interesting is looking at Notre Dame (which has improved each year under Brian Kelly) and Louisville. With both of those programs, it could be argued that recruiting has played a bigger role in those regimes than offensive scheme, and that, from recruiting flowed offensive improvement. A lesson from old school coaches everywhere seems to ring true: When it’s time to play the game, it’s not about X’s and O’s, but Jimmies and Joes.