As David Hale noted on Saturday night, the coaching class of 2018 hasn’t distinguished itself in its second season as much as was assumed, at least not in a good way. That point was driven home when Florida State’s Willie Taggart, one of said second-year coaches, got fired.
FSU had improved this fall after bottoming out in Taggart’s 2018 debut, but only marginally so. They had rebounded from 71st in SP+ to only 51st, and after going 5-7 last year, they still had, per SP+, a 55% chance of finishing with six or more wins. This was “improvement” in name only — 51st might be better than 71st, but it’s still awfully bad for a team with top-15 recruiting talent.
It was becoming pretty difficult to see Taggart thriving in Tallahassee, but dumping a guy this quickly when he isn’t cratering is awfully rare. It’s not unheard of for a coach to get fired after just two years (or slightly less), but it’s usually a situation like that of Tyson Summers or Turner Gill. Summers started 3-0 at Georgia Southern but lost 13 of his next 15 games and was fired at 0-6 in year two. At Kansas, Gill started 2-2 in 2010, then lost 18 of 20 against FBS competition.
There are 14 programs that have either won a national title in the last 20 years or played for one in the last 10, a reasonably quick definition for what constitutes a potentially elite program: Alabama, Auburn, Clemson, Florida, Florida State, Georgia, LSU, Miami, Notre Dame, Ohio State, Oklahoma, Oregon, Texas, and USC. Here are the guys who were fired a) within four seasons of their hiring, and b) during/after a merely mediocre season (we’ll say five or more wins):
• Lane Kiffin (USC): 3-2 in his fourth season
• Randy Shannon (Miami): 7-5 in his fourth season
• Mike Shula (Alabama): 6-6 in his fourth season
• Charlie Strong (Texas): 5-7 in his third season
• Tyrone Willingham (Notre Dame): 6-5 in his third season
• Ron Zook (Florida): 7-4 in his third season
These are the programs with the highest expectations in the country; Taggart is the only one in this time frame to not be granted a third year on the job.
I’m very curious, by the way, what this move does for the patience levels of other schools with struggling second-year coaches, like Arkansas, Arizona, or Mississippi State. It could be a very strange, and panic-stricken, coaching carousel.
Whither the second-year leap?
Taggart’s relative struggles, along with that of big-name second-year guys like Nebraska’s Scott Frost and UCLA’s Chip Kelly, have been noteworthy, not only because they all had previous success on their records, but also because it’s their second season on the job. This is prime leap time.
If a head coach is going to see a leap, it’s probably going to happen pretty quickly after his hire. Over the past 10 seasons, 23 teams have seen their SP+ rating improve by 18 points per game in a single season. Eighteen of those teams were led by a head coach in either his first, second or third season at the helm. […]
Over those 10 years, 70% of second-year coaches enjoyed a win total at or above that of their first year and 63% saw their SP+ rating improve (46% by at least three adjusted points per game, 17% by at least 10).
If that trend were to continue for this year’s 20 second-year coaches, approximately 13 of them will improve on paper, including three by a large amount.
Believe it or not, 2019’s crop of second-years is almost perfectly replicating those trends. Using updated SP+ projections, 11 of the 20 are expected to exceed their 2018 win totals (at least with decimals involved).
• SMU (Sonny Dykes): 9.8 projected wins, up 4.8 from 2018
• Oregon State (Jonathan Smith): 5.0 projected wins, up 3.0
• UL-Lafayette (Billy Napier): 9.5 projected wins, up 2.5
• Kent State (Sean Lewis): 4.4 projected wins, up 2.4
• UCLA (Chip Kelly): 5.0 projected wins, up 2.0
• Oregon (Mario Cristobal): 10.5 projected wins, up 1.5
• UTEP (Dana Dimel): 2.0 projected wins, up 1.0
• Nebraska (Scott Frost): 5.0 projected wins, up 1.0
• Tennessee (Jeremy Pruitt): 5.7 projected wins, up 0.7
• Florida State (Willie Taggart): 5.6 projected wins, up 0.6
• Arkansas (Chad Morris): 2.6 projected wins, up 0.6
• Arizona State (Herm Edwards): 7.0 projected wins, even
• Arizona (Kevin Sumlin): 4.5 projected wins, down 0.5
• Florida (Dan Mullen): 9.4 projected wins, down 0.6
• Rice (Mike Bloomgren): 0.9 projected wins, down 1.1
• South Alabama (Steve Campbell): 1.6 projected wins, down 1.4
• Texas A&M (Jimbo Fisher): 7.2 projected wins, down 1.8
• Mississippi State (Joe Moorhead): 5.7 projected wins, down 2.3
• UCF (Josh Heupel): 9.5 projected wins, down 2.5
• Georgia Southern (Chad Lunsford): projected wins, down 2.7
If Arizona State were to win its projected bowl game, that would likely push the Sun Devils over last year’s win total and make it 12 of 20.
Similarly, 11 of 20 have seen their SP+ rating improve (four others are close enough to still pull it off), and four of the 20 (awfully close to the 17% average) have currently improved their SP+ rating by more than 10 adjusted points per game: Oregon State (up 17.7), UL (up 14.8), SMU (up 12.5), and Oregon (up 12.2).
This year has felt disappointing for second-year coaches, but that’s primarily because the biggest names of the bunch have in no way been the best performers. Of the seven schools I declared most likely to leap in 2019 — Arkansas, FSU, Nebraska, Oregon, Oregon State, Tennessee, and UCLA — only two have seen definitive steps forward (UO and OSU). And one’s now looking for a first-year coach again.
QB security: almost as tenuous as your coach’s job security
One of Taggart’s biggest issues in 2019 was one that a lot of coaches — second-years (Moorhead, Mullen, Sumlin, Morris, Frost, Pruitt, Bloomgren) and otherwise — have shared: an inability to either keep your starting QB healthy or figure out who your starter should be at all.
Due to a combination of injuries and a number of schools struggling to settle on a starter, we’ve seen an absurd number of QBs taking the field for FBS teams this fall.
• In a given season, about 45% of teams end up with one quarterback passing for over 90% of his team’s passing yardage — a nice sign of stability at the position. That’s been no different in 2019; right now, 58 of 130 teams (45%) are over that 90% mark.
• However, the teams with uncertainty have had loads of uncertainty. From 2005-18, only about 15% of teams had saw their “leading” passer throw for less than 60% of their team’s total passing yards (which means about half, or maybe more, of the yards came from a guy you didn’t really expect). This year, however, 31 teams (24% of FBS) qualify.
• Meanwhile, 29% of teams have seen their respective No. 2 passers throw for at least 25% of their yardage (as compared to the 2005-18 average of 25%), and 15 have seen their No. 3 leading passer throw for at least 10% of their yardage (normal: 8%).
Translation: While the same number of teams as normal has enjoyed reasonable stability at the QB position, those suffering instability have suffered spectacular instability.
The number of starting QBs in the power conferences alone has crossed 100 so far this year, and it’s having an obvious effect on production.
• Teams in the “one guy, 90+% of yards” club are averaging an Off. SP+ rating of 30.9 adjusted points per game and a ranking of 54.8.
• Teams with a leader between 60% and 90%: 28.1 adjusted PPG and a 65.8 ranking.
• Teams below 60%: 24.3 adjusted PPG and an 81.5 ranking.
It’s almost like stability at the most important position in sports is beneficial…
Why all the cool kids are going for two
2:49
Antonio Gibson and Damonte Coxie score five touchdowns as Memphis sends SMU to its first loss 54-48.
Down 14 points late in the fourth quarter of Saturday night’s delightful SMU-Memphis game, one of the aforementioned second-year leapers did something pretty trendy. Sonny Dykes’ Mustangs scored a touchdown and, without hesitation, elected to go for the 2-point conversion.
This has become steadily more common this season, but it hadn’t quite reached everyone’s radar yet. Announcers Chris Fowler and Kirk Herbstreit — as experienced and plugged-in a duo as you’ll find — were befuddled by the move until a coach evidently texted Herbstreit in the booth.
Rufus Peabody laid out the logic as concisely as I’ve seen on Twitter on Sunday morning:
Let’s assume you’re down 14 and score 2 TDs (and opp doesn’t score). For simplicity, PAT=100%, 2pt=50%.
S1: 2pt✅, PAT✅, WIN (50%)
S2: 2pt❌, 2pt❌, LOSE by 2 (25%)
S3: 2pt❌, 2pt✅, Force OT (25%)If WP(OT)=50%, WP=62.5% going for 2 vs. 50% if you kick 2 PATs to force OT. https://t.co/uDSLXRsUv2
— Rufus Peabody (@RufusPeabody) November 3, 2019
Basically, if you make the 2-point conversion, you’re in position to take the lead if you score another touchdown (at least, unless you’re snakebitten in the place-kicking department — earlier this year against Virginia Tech, Miami went for two, got it, scored again, and missed the PAT). If you miss the 2-pointer, you can still go for two to tie if/when you score again.
The logic is basically unassailable with the assumptions he lays out above (100% chance of making a PAT, 50% chance of converting the 2-pointer). This quickly becomes a blurrier and more fascinating situation, though, when you take into account the fact that 2-point conversion rates are only about 45% thus far in 2019 and have typically been even lower than that in recent years.
As Peabody points out, the math still works with lower 2-point conversion rates, but it very much becomes a quality vs. quality debate. Is your offense better than average? Is their defense? The numbers can pretty quickly flip from “do it!” to “not so fast, my friend,” depending on the answers. (Needless to say, in a game that featured 102 points and 1,067 yards, the math all but screamed “DO IT” in Memphis on Saturday night.) As a fan of endless gray area, then, this has quickly become one of my favorite topics. I hope the opportunities continue to arise in big games down the stretch.
Thoughts from inside CFP headquarters
The first College Football Playoff rankings of the season will be released on Tuesday night. ESPN’s Heather Dinich laid out all of the major questions we’ve got — How does the committee view a Clemson or a Penn State? How much did OU’s loss to K-State hurt? Where exactly do you put Baylor and Minnesota? — but I wanted to add a few extra thoughts as preparation.
In October, with a lot of ESPN coworkers, I got to take part in the annual College Football Playoff mock selection process. It’s a pretty fun exercise — you meet the CFP higher-ups, spend a day in the committee’s debating room, use the information they use, and walk through the repeated steps of the voting process.
We used the 2014 selection process, step-by-step, for the exercise. Here are some thoughts about what I learned about the process and what it might mean for this year’s selections.
1. First things first: The committee usually gets it right. I should start there before I get to any critiques. I haven’t had a problem with the selection of any of the 20 teams that have yet made the playoff. Sometimes more than four teams are worthy of a spot, which results in dissatisfaction from the spurned teams and fans, but the selections have all been defensible.
For what it is, this process probably is about as good as it can be. There is constant discussion, there are constant comparisons and there are opportunities for revotes if committee members want to relitigate a past discussion. The process is clean, and anonymous, enough so that there’s really no room for “they ranked this team here because they wanted this CFP matchup or this bowl selection” conspiracy theories.
2. It’s still basically a poll, though. The people involved with this process take their jobs very seriously, but it’s still a collection of pretty similar individuals — mostly with extensive involvement in one aspect of football or another (former coaches, athletic directors, media members, etc.) — discussing and then anonymously ranking a selection of teams. This isn’t “Are we all in favor of Team A being No. 1?” “Aye!” “Well let’s move on to No. 2.” It’s chunk by chunk, and there’s opportunity for feedback … but it’s basically a poll. The wheel hasn’t been entirely reinvented.
3. The stats are … OK. SportSource Analytics was contracted to provide data for the committee to consult while comparing teams. The tool they created is impressive — you can compare up to four teams side by side, looking at their schedule, results and averages/rankings in key statistical categories that correlate well to wins and losses.
I had a few qualms, though.
First, one of the stats they use for comparison is yards per point. YPP tells us very little about quality, only fortune. It’s similar to one’s turnover margin in that regard — if it’s too good or too bad, it’s almost certainly regressing to the mean (and taking your results with it) soon.
Phil Steele has long referred to YPP as a way to gauge which teams are going to regress or improve the next year. The current top five in offensive YPP includes Fresno State and Illinois. The top five in YPP margin includes Wyoming.
This is a problem because of how these measures are used. For each category, the team with the superior ranking in the comparison group gets a green dot, and the teams with the worst ranking get a red dot. When surveying this list of stats, you inevitably default to counting up the green and red dots for each team.
Second, comparisons of results almost immediately break down into “ain’t played nobody” territory. You see each opponent’s record and ranking at the top of the screen, and you instinctively absorb “they’ve played this many ranked teams and this many .500 teams” and ignore almost everything else.
That creates a couple of problems.
4. A Group of 5 team is almost definitely never going to make it into the top four. I’ve always assumed this (I’m obviously not alone), and walking through the process pretty much affirmed it.
5. It’s “most deserving,” not “best.” CFP executive director Bill Hancock, committee chairman Rob Mullens and anyone else associated with the CFP insist they choose the best four teams for the playoff. “Not the four most deserving, not the fan favorites, not the Cinderellas.”
But their committee picks the teams with the best combination of wins (most of all), ranked opponents and green dots. The 2015 Iowa Hawkeyes finished within about 30 seconds of a 13-0 record and a Big Ten title despite probably not even being one of the 25 best teams in college football. They would have likely scored a playoff bid.
This is fine! Results matter! Choosing the most deserving teams is perfectly acceptable! But the CFP continues to fight an unnecessary semantic battle on this one.
What does this mean for 2019?
There will be no UCF Scenario this year.
So point No. 4 above does not apply. UCF went unbeaten in each of the last two regular seasons but couldn’t top 12th in the CFP rankings in 2017 or eighth last year. It was a major point of contention, but this won’t be an issue because Week 10 losses by Appalachian State and SMU assured that there won’t be an unbeaten G5 team this year.
Minnesota and Baylor are fine, no matter where they rank this week.
Both the Golden Gophers and Bears get much friendlier grades from the advanced stats than Iowa did in 2015, but they still get the lowest of the unbeatens — Minnesota’s 10th in SP+ and 20th in FPI, while Baylor’s 15th and 21st, respectively. They both played no one of importance in nonconference, and Minnesota’s back-loaded conference schedule means the Gophers haven’t played anyone better than Nebraska (50th in SP+) so far.
That will change, as both teams have top-10 opponents on the horizon. But no matter where they start out in Tuesday’s rankings, their path to the CFP is pretty clear. And whether they’re one of the four “best” teams or not, they’ll get in at 13-0.
Ohio State’s schedule probably won’t get quite enough respect.
I create strength-of-schedule ratings based on one simple question: How would the average top-five team (per SP+) fare against your schedule? It projects this top-five team’s win percentage if it played each of your opponents infinity times each.
Among the remaining P5 unbeatens, Ohio State’s schedule grades out as the most difficult. A top-five team’s expected win percentage against the Buckeyes’ schedule to date would enjoy about a 0.895 win percentage. That’s not exactly Texas A&M-level difficulty, but it ranks 36th overall, a hair ahead of LSU’s (47th) and Penn State’s (53rd), and well ahead of Baylor’s (70th), Alabama’s (79th), Clemson’s (91st), and Minnesota’s (98th).
Depending on how the committee views Indiana, though, the Buckeyes might only boast one win over a CFP top-25 team. LSU will have at least two (Auburn and Florida, with a slight chance of Texas slipping in as well). Despite Ohio State’s incredible dominance to date, then, it wouldn’t be a surprise if LSU’s “résumé” (despite an SOS that isn’t actually any more difficult) powers the Tigers ahead of the Buckeyes.