The topic of how much teams can improve in the second year of a head coach’s tenure has been a fascination of mine for a long time. Back in college in the summer of 2006, I started a Florida football blog titled “Year 2” in anticipation of an improvement in Urban Meyer’s second season in Gainesville after his second-year improvements at Bowling Green and Utah.
The 2000s were a golden age for sophomore surges for head coaches. Bob Stoops (2000), Jim Tressel (2002), Meyer (2006), and Gene Chizik (2010) all won national championships in their second years on the job. Chizik won his over another coach in his second year, Chip Kelly. Mark Richt almost got there with a 13-1 campaign in 2002, getting unlucky that there were two undefeated teams that year, and Nick Saban might’ve gotten the job done too had he won the 2008 SEC Championship Game.
To try to quantify things a bit, I went through every head coaching hire in the SEC this century and looked at how their wins and point differential (points for minus points allowed) changed from their first years to their second. There were 33 such hires; the new 2018 coaches and Lane Kiffin at Tennessee aren’t in there since I have no second-year data for them. I also didn’t count interim tenures, so Ole Miss’s Matt Luke isn’t included here and Ed Orgeron’s LSU tenure started in 2017 for my purposes.
Before I get too far, I will note the caveat that you should only take away some rough approximations here. The hires of 2001, 2003, and 2005 all played different numbers of regular season games in their second years than their first, for instance. Some coaches also had different numbers of games in their first two years depending on whether they played in the SEC Championship Game or a postseason bowl or not.
So with those caveats in mind, the average SEC hire since 2001 improved by 1.75 wins and 71.2 points from Year 1 to Year 2.
Chizik had the greatest win jump, going from eight to 14 for a gain of six. Kirby Smart, Guy Morriss (Kentucky, hired in 2001), Saban, and Richt tied for the next-most at a gain of five. Kevin Sumlin and Rich Brooks (Kentucky, 2003) saw their win counts fall by two; the only one worse than that was Gus Malzahn going from 12 wins to eight in his second season.
The largest gain in point differential went to Bret Bielema. His first team’s opponents outscored his Razorbacks by a combined 121 points, but his second team outscored its opponents by a combined 165. That amounts to an improvement of 286 points. Smart is second at a gain of 278, followed by Bobby Petrino (253), Morriss (192), and Chizik (165).
Here is how things went for Florida’s four entries on my list. Ron Zook won eight games in each of his first two seasons for a change of zero, but he did improve the point differential by 62. Meyer jumped four wins from nine to 13, and his point differential improvement was 110. Will Muschamp tied his predecessor with an improvement of four wins going from seven to 11, but his point improvement was just 89. Jim McElwain dropped a win from ten to nine, though a hurricane cancelation robbed him of a gimme win. His point differential improvement, also impaired by losing an FCS opponent, was a modest 24.
Dan Mullen had one of the better second-year improvements at Mississippi State. He bumped up the record by four wins from five to nine, and his point differential change was eighth-highest at 133.
Here’s the real question: how much could Mullen improve Florida in his second season in Gainesville?
By winning ten games in 2018, Mullen cannot tie Chizik at plus-six wins because he can play a maximum of 15 games. Unless Hawaii gets a cancelation and UF burns a bye week to play in the islands — NCAA rules allow teams that play road games at Hawaii to have 13 regular season games — he can’t possibly improve by six wins.
Setting aside record chasing, the more you win, the harder it is to improve win count. Seven of the 33 coaches I studied won at least nine games in their first seasons. Malzahn (-4), Sumlin (-2), and McElwain (-1, with the storm asterisk) won fewer games the next year. Les Miles and Houston Nutt (Ole Miss, 2007) won the same number in their second seasons as their first. Orgeron at LSU improved by one from nine to ten. Only Meyer, going from nine to 13, had a substantial improvement.
Mullen’s 2018 Gators had a point differential of 195: 455 for, 260 allowed. That will also be hard to improve upon looking at this group of SEC hires. Only Sumlin (295), Malzahn (207), and Miles (198) had a larger Year 1 point differential than Mullen did at UF, and Nutt (170) is the only other guy who’s close.
Sumlin (-139), Malzahn (-93), and Nutt (-16) saw declines in their point differentials in their second seasons. Miles improved his by 76 points. However, the Mad Hatter’s 2005 record of 11-2 included a 20-point loss in the SEC Championship Game, whereas his 2006 record of 11-2 dropped the conference title game for an extra cupcake with the addition of the 12th regular season game. The smallest margin of victory in any of those body bag games was 32 points. That swing alone almost accounts for the entire improvement, and swapping a decent 2005 Arizona State for a subpar 2006 Arizona more than takes care of the rest.
I used this data with one exception to create a regression model that uses Year 1 wins and point differential to project Mullen’s change in win count and point differential in his second season. The exception is that I tossed out UK’s Brooks, who took over a program just starting to feel NCAA sanctions for Hal Mumme’s misdeeds. I will note before continuing that the wins model is stronger than the points differential model, but neither explains even half the variance in Year 2 changes.
Long story short, the model sees a point differential of 196 and 10.85 wins for the 2019 Gators. I used Pythagorean expectation (something you’ve seen before) and an assumption of 13 games played to get 396 points scored and 200 allowed.
Pythagorean expectation puts UF at 10.27 wins using 2018’s points scored and allowed, so the Gators were about right on. What I’ve projected here sees the team’s expectation improving by about half a win.
The win part feels about right. The schedule isn’t any easier in 2019 than it was in 2018, so with a bowl win it’s projecting a 9-3 or 10-2 regular season. The point differential seems wrong since I wouldn’t expect this fall’s offense to score 59 fewer points than last season’s did, though trading 2018 Colorado State for 2019 Miami will put downward pressure on this category. Like I said, though, the point differential is the weaker of the two projections.
Second year improvements are common and expected. Only six of 33 coaches — less than one in five — saw win count declines, some of which had to do with a storm cancelation for McElwain and sanctions for Brooks. Only six saw point differential worsen, five of whom were in the win decline group. The sixth was Nutt, who basically broke even at 1.2 fewer points per game than in his first.
It’s very hard to say how much Mullen can improve the team this year. By winning ten games with a +195 point differential, he’s already in a rare spot for recent SEC head coaches. Put it this way: he was one win and three points’ worth off from Miles, who inherited Saban’s LSU that had won a national championship two years prior. Mullen inherited McElwain’s program that had recruited at a much lower level and was two years removed from merely an SEC East title.
If Mullen can simply tread water from last year, that’ll put him ahead of what Sumlin and Malzahn did in their second years. If he can noticeably improve the point differential, that’d put him in league with a coach who won the national title in his third year in Miles.
The most likely thing is probably somewhere from treading water to a modest improvement given that he set the bar so high in his first year. Then again, Mullen was an assistant on those Bowling Green, Utah, and Florida teams that had stellar second years, and he had his own Year 2 leap at his last stop. Given the rather obvious flaws in execution (Kentucky) and motivation (Missouri) in losses last year that in theory shouldn’t be around this year, there would appear to be room to grow. Mullen is capable; we’ll just have to see if the team can pull it off.