DaggaStats 2015 College Football Coaching Year in Review, Part II: The P5 Conferences

If you read DaggaStats Part I, you might've anticipated that Tennessee could have a tough time with Appalachian State, my pick to win the Sun Belt Conference.

This year App State should be playing like a scrappy-but-mature P5 conference team. According to their Q metric, they would've finished 4th in the SEC East last year, only two spots behind Tennessee. They have an outstanding Bill-Snyder-type small school system coach whose teams are never sloppy, and they're returning three-fourths of their total production. There's a reason they have a history of being a dangerous team to play in Week One and those reasons are more pronounced than usual in 2016.

And as we'll see below, while Butch Jones has accumulated talent at Tennessee at a frenetic pace and that talent is starting to mature, his coaching value-added has been middling so far. There's little reason to expect them to suddenly be a team that throttles quality opponents like Jay Mariotti in a round of speed dating.

Don't get me wrong, Tennessee should've been a clear favorite, but if Twitter is any indication it seems like most of the sports media expected App State to get rolled. I'd understand how casual fans (like um, most of the sports media?) would automatically expect a recently promoted DII school to get rolled by any SEC school. But just a simple look at last year's results would portend a fairly close game most of the way with Tennessee pulling away late. Which is more-or-less what happened.

Moral: Don't dial down your estimates of Tennessee's quality too much. Raise your esteem of App State instead.

ONWARD THRU THE FOG (h/t Max Redfield):

#6 - FBS Independents (mean CxT: 66.81)

FBS Independents

Team	2015 Record	2015 S&P+ Rk	Proj. 2016 S&P+ Rk	Proj. 2016 Wins	5-Yr Rec. Rk	Ret. Experience (Off. / Def.)
Notre Dame	10-3	8	11	8.6	8	53% (60% / 46%)
BYU	9-4	38	35	7.0	70	70% (67% / 73%)
UMass	3-9	97	127	3.4	116	24% (15% / 34%)
Army	2-10	123	124	4.6	126	78% (71% / 85%)

** Link to Bill C.'s Independents capsule **

Since everyone's going to focus on Notre Dame here, let's use this as a reminder how to interpret DaggaStats. If you don't care, just skip to the last paragraph.

I look at Notre Dame's line and immediately think: Brian Kelly is a damn fine coach.

But why? After all, Kelly's C rating last year was a measly 68. C stands for Coaching, right? 68 is below average, right? Not exactly.

C represents the relationship between T and Q. Specifically, it's the part of Q for which T doesn't account. You can think of the relationship as Q = C + T.

But C is then scaled between 40 and 100 and centered at 70, so that a 70 C score means the team has precisely lived up to its estimated recruited talent; teams over 70 are overachieving, teams under 70 are underachieving.

The core idea of CxT is this: we can think of these simple numbers as standing in for bigger concepts. Q can represent current team quality, C can represent current coaching value-added, and T can represent the coach's full résumé to date.

To explain that last point: talent levels don't typically follow the coach around; they tend to apply to the school. Only the most elite recruiting coaches can budge a school's T level by more than five points or so. The talent level of a coach's current school testifies less to their recruiting skills and more to their ability to land a good job.

So we can use school talent to place a rough numerical value on the coach's résumé. If Brian Kelly has a T score of 90, that means he's been a good enough coach in the recent past to get a job at a 90 T school.

Ultimately coaches should be judged by the quality of their teams. However, when comparing coaches whose teams perform at the same level of quality, those who have a substantially larger gap between T and C should be discounted. Why? Because the data we have for their current performance (C) and past performance (T) don't obviously tell the same story. There's more uncertainty about how good they actually are, and that extra uncertainty risk should dent their perceived value.

For example: Tom Herman will almost certainly have a better C score than Urban Meyer in 2016. But Meyer's resume is MUCH more impressive and that's reflected mostly in the fact that he's coaching at a school MUCH richer in talent, which in turn makes it harder for his team to overachieve. So if this year both Houston and Ohio State turn in performances of 90 Q, Meyer should still be considered the more valuable coach even though Herman's team is playing wa-a-a-ay over their heads while Meyer's is just meeting expectations. And if both coaches turn in performances of 70 Q, Herman's going to have a better CxT because under those circumstances Meyer would be deep-sixing his own program with massive underachievement.

So CxT is basically an adjustment to Q, taking its main components and recombining them to reward coaches who guide teams to a high quality performance and have built a quality résumé.

Ahem. Back to Brian Kelly.

T is a "dumb" number. It doesn't account for attrition or injury or failure to develop. Which means it doesn't account for Brian Kelly losing his starting QB, RB, NT, S, and CB over the course of the first few games of the season, plus his best defensive player in the middle of their bowl game.

So Kelly's C of 68 tells us that despite a rash of critical injuries, Notre Dame still almost played up to its (very high!) expected talent level.

Kelly's a very good coach, he's got a lot of talent, and he knows how to cover his weaknesses without losing more than a step. The game on Sunday should be very tough for Texas to win.

My Pick to Win: EVERYONE GETS A TROPHY

Most Likely Art Briles Landing Spot: BYU

#5 - Big XII Conference (mean CxT: 73.72)

Big XII

Team	2015 Record (Conf.)	2015 S&P+ Rk	Proj. 2016 S&P+ Rk	Proj. 2016 Wins (Conf.)	5-Yr Rec. Rk	Ret. Experience (Off. / Def.)
OU	11-2 (8-1)	4	4	9.6 (7.1)	16	69% (72% / 66%)
TCU	11-2 (7-2)	26	31	7.1 (5.0)	35	48% (29% / 68%)
OSU	10-3 (7-2)	33	23	7.9 (5.4)	38	74% (79% / 69%)
BU	10-3 (6-3)	14	13	9.0 (6.1)	30	69% (71% / 66%)
Texas	5-7 (4-5)	72	34	6.7 (4.8)	11	80% (79% / 81%)
WVU	8-5 (4-5)	24	33	6.9 (4.8)	50	62% (86% / 38%)
Texas Tech	7-6 (4-5)	62	43	6.4 (4.2)	42	74% (76% / 72%)
KSU	6-7 (3-6)	84	67	5.2 (3.3)	61	70% (77% / 62%)
ISU	3-9 (2-7)	70	71	5.1 (3.3)	65	68% (65% / 71%)
KU	0-12 (0-9)	127	112	2.6 (1.0)	71	83% (81% / 85%)

** Link to Bill C.'s Big XII capsule **

Yes, DaggaStats is a coach ranking system. Yes, the Big XII has some great coaching. So why is the Big XII dead last among the P5 conferences in mean CxT?

Because the talent level in the Big XII is dead last among the P5 conferences and it's not especially close. Kansas, Iowa State, and KSU are arguably the three worst schools in the entire P5 when it comes to star recruiting. The middle tier of the league - WVU, TTU, OSU, TCU, Baylor - all have T levels between 73 and 76, none of whom would break the top eleven in the SEC. In the Big XII, that scrum is basically tied for third. Even the top tier, OU and Texas, are currently weak-T in comparison to the elite teams in football.

Also, to state the obvious: in 2015 Texas and Kansas were perhaps the two worst underachievers in the P5. We know the situation with Texas - the T level didn't reflect the actual talent on the roster or its distribution at key positions, and the offense was schematically frayed and psychologically Mommie Dearest on an endless loop. With the last two recruiting classes and a new set of coaches on offense, this set of circumstances is due to change rapidly over the next two seasons (crosses fingers) (knocks on wood) (pushes Aggie under a ladder).

But...Kansas. It's really impressive to be the most talent-poor team in the P5 AND the most underachieving. If it wasn't for George O'Leary's failed matter/antimatter experiment in Orlando, they would've also had the coach with the worst CxT in the FBS.

Which means Kansas can't even win the sh*tball triple crown. Sad!

My Pick to Win: Texas (don't believe me, I'm a homer)

Most Likely Art Briles Landing Spot: incredibly, Baylor

#4 - Atlantic Coast Conference (mean CxT: 75.08)

ACC Atlantic

Team	2015 Record (Conf.)	2015 S&P+ Rk	Proj. 2016 S&P+ Rk	Proj. 2016 Wins (Conf.)	5-Yr Rec. Rk	Ret. Experience (Off. / Def.)
Clemson	14-1 (8-0)	2	3	10.0 (6.5)	14	70% (90% / 50%)
FSU	10-3 (6-2)	7	5	8.5 (5.6)	3	63% (77% / 49%)
Louisville	8-5 (5-3)	28	20	8.2 (5.0)	39	87% (98% / 77%)
NC State	7-6 (3-5)	36	40	6.2 (3.4)	47	60% (50% / 70%)
Syracuse	4-8 (2-6)	71	44	5.5 (3.1)	63	90% (100% / 80%)
BC	3-9 (0-8)	67	50	6.4 (2.9)	66	81% (92% / 71%)
Wake	3-9 (1-7)	92	74	5.5 (2.5)	62	84% (89% / 79%)

ACC Coastal

Team	2015 Record (Conf.)	2015 S&P+ Rk	Proj. 2016 S&P+ Rk	Proj. 2016 Wins (Conf.)	5-Yr Rec. Rk	Ret. Experience (Off. / Def.)
Pitt	8-5 (6-2)	40	29	7.0 (4.3)	37	72% (72% / 72%)
UNC	11-3 (8-0)	29	27	7.5 (4.5)	23	64% (50% / 78%)
Miami	8-5 (5-3)	51	30	7.0 (4.3)	17	70% (78% / 62%)
Va. Tech	7-6 (4-4)	59	32	6.8 (4.5)	25	74% (66% / 82%)
Ga. Tech	3-9 (1-7)	68	54	5.6 (3.2)	45	70% (97% / 42%)
Duke	8-5 (4-4)	75	51	5.9 (3.5)	52	76% (80% / 73%)
Virginia	4-8 (3-5)	73	68	5.1 (2.9)	48	72% (80% / 64%)

** Link to Bill C.'s ACC capsule **

First things first: look at the returning experience. All 14 schools are returning most of their overall production. That means the conference leaders - Clemson and Florida State - will have a tougher road to hoe, especially considering they're in the same division as each other and Louisville, where Bobby Petrino returns nearly his entire offense. Running the table will be hard.

The Coastal division also looks to be wide open, with Justin Fuente and Mark Richt now heading up the traditional division powers while Larry Fedora and Pat Narducci try to parlay last season's solid play into a multi-year string.

And while he may be a couple years away from getting the roster in shape, I'll be interested to follow former BYU stalwart Bronco Mendenhall as he tries to pull a Bad News Bears with eternally moribund Virginia. In terms of recruiting, they're capable of doing much better. Bronc should post LeBron posters everywhere to help people realize that Cavaliers aren't ALWAYS destined to lose.

Also: see how thoroughly unimpressed this system is with Duke's 8-5 performance last year? Don't expect a repeat.

My Pick to Win: Clemson

Most Likely Art Briles Landing Spot: Boston College

#3 - Pac 12 Conference (mean CxT: 75.10)

Pac-12 North

Team	2015 Record (Conf.)	2015 S&P+ Rk	Proj. 2016 S&P+ Rk	Proj. 2016 Wins (Conf.)	5-Yr Rec. Rk	Ret. Experience (Off. / Def.)
Stanford	12-2 (8-1)	10	16	7.9 (5.7)	18	36% (20% / 52%)
Oregon	9-4 (7-2)	25	18	8.2 (5.8)	19	64% (49% / 79%)
UW	7-6 (4-5)	12	10	9.0 (6.1)	29	76% (72% / 79%)
Wazzu	9-4 (6-3)	63	48	6.2 (4.0)	57	79% (87% / 71%)
Cal	8-5 (4-5)	34	49	5.2 (3.5)	36	39% (15% / 62%)
OSU	2-10 (0-9)	102	86	4.0 (2.5)	54	75% (72% / 78%)

Pac-12 South

Team	2015 Record (Conf.)	2015 S&P+ Rk	Proj. 2016 S&P+ Rk	Proj. 2016 Wins (Conf.)	5-Yr Rec. Rk	Ret. Experience (Off. / Def.)
USC	8-6 (6-3)	17	8	7.7 (5.9)	2	66% (54% / 77%)
UCLA	8-5 (5-4)	23	12	8.6 (6.6)	13	75% (61% / 88%)
Utah	10-3 (6-3)	27	39	6.6 (4.4)	46	55% (35% / 76%)
ASU	6-7 (4-5)	57	57	5.6 (3.4)	28	35% (27% / 43%)
Arizona	7-6 (3-6)	77	64	5.7 (3.5)	44	63% (61% / 65%)
CU	4-9 (1-8)	98	82	4.3 (2.6)	56	80% (81% / 78%)

** Link to Bill C.'s PAC 12 capsule **

The Pac 12 was just a murderer's row of bad luck last year, with lots of mediocre teams winning games against better opponents in conference and regressing pretty much everyone but Stanford aggressively to the mean. One significant sufferer was Washington's Chris Petersen, who with a normal run of luck may be poised to challenge Stanford for conference superiority this year. Utah and UCLA are primed to duke it out in the South while USC tries to survive a brutal schedule (non-division games: Bama, ND, Stanford, Oregon, Washington, Utah State).

There's plenty of other interesting storylines across conference. Wazzu is finally hitting Full Pirate Mode, Oregon's Helfrich needs to reassert some hint of Oregonian dominance in the North while Stanford reloads or he'll be on the hot seat, both Gary Andersen and Mike McIntyre return a bunch of players and desperately need a turnaround in fortunes this year.

Too bad they all play their games in the West Micronesian Time Zone. Hopefully they'll hire NBC to manage the tape delays, and along the way we'll get that soft-focus Sarkisian confessional/redemption interview for which we've all been waiting. It's Hoda's time to shine y'all.

My Pick to Win: Washington

Most Likely Art Briles Landing Spot: Colorado

#2 - Big Ten Conference (mean CxT: 75.11)

Big Ten East

Team	2015 Record (Conf.)	2015 S&P+ Rk	Proj. 2016 S&P+ Rk	Proj. 2016 Wins (Conf.)	5-Yr Rec. Rk	Ret. Experience (Off. / Def.)
Michigan	10-3 (6-2)	6	6	9.6 (6.8)	9	42% (25% / 59%)
Mich. St.	12-2 (7-1)	13	22	7.8 (5.8)	21	45% (26% / 64%)
Ohio St.	12-1 (7-1)	3	14	8.1 (6.1)	4	29% (22% / 36%)
Penn St.	7-6 (4-4)	30	28	7.5 (5.4)	26	58% (53% / 63%)
Indiana	6-7 (2-6)	60	56	6.0 (3.8)	55	66% (47% / 84%)
Maryland	3-9 (1-7)	65	62	5.9 (3.6)	41	70% (93% / 46%)
Rutgers	4-8 (1-7)	105	87	4.1 (2.3)	53	85% (81% / 90%)

Big Ten West

Team	2015 Record (Conf.)	2015 S&P+ Rk	Proj. 2016 S&P+ Rk	Proj. 2016 Wins (Conf.)	5-Yr Rec. Rk	Ret. Experience (Off. / Def.)
Nebraska	6-7 (3-5)	48	26	7.9 (5.6)	24	78% (94% / 63%)
Wisconsin	10-3 (6-2)	31	37	6.3 (4.4)	34	44% (34% / 54%)
Minnesota	6-7 (2-6)	37	42	7.5 (4.9)	58	66% (84% / 48%)
Iowa	12-2 (8-0)	47	38	7.4 (4.8)	51	72% (72% / 72%)
NW'ern	10-3 (6-2)	56	46	6.2 (4.0)	43	63% (68% / 58%)
Illinois	5-7 (2-6)	61	76	5.0 (3.2)	60	56% (74% / 38%)
Purdue	2-10 (1-7)	86	88	4.5 (2.6)	67	66% (68% / 65%)

** Link to Bill C.'s B1G capsule **

One more time: check out the returning experience. Four of the six ten-win teams from last year are returning less than half their production...and the other two are Iowa and Northwestern.

Michigan and Ohio State will still be the favorites and Michigan State and Wisconsin will still be strong, but expect some mayhem in the standings if those teams don't find ways to answer all their key question marks in the early part of the season. Furman says holla.

And someone gotta say it: for what's supposed to be a quality league, the new coaches are as thrilling as a dry slab of turkey breast on a rice cake. Except (and this is wholly subjective) Lovie Smith. Mostly because he's a ornery old school coach named "Lovie" and I root for dudes like that 100% of the time even when they're walking humanitarian nightmares. But also because with his name recognition in Chicagoland I could see him recruiting quite well at Illinois if he can get that program resuscitated. Big "if" though.

My Pick to Win: Michigan

Most Likely Art Briles Landing Spot: Purdue

#1 - Southeastern Conference (mean CxT: 76.73)

SEC West

Team	2015 Record (Conf.)	2015 S&P+ Rk	Proj. 2016 S&P+ Rk	Proj. 2016 Wins (Conf.)	5-Yr Rec. Rk	Ret. Experience (Off. / Def.)
Alabama	14-1 (7-1)	1	1	9.5 (5.9)	1	55% (45% / 66%)
LSU	9-3 (5-3)	9	2	9.3 (5.6)	7	92% (96% / 88%)
Ole Miss	10-3 (6-2)	5	7	7.8 (4.6)	20	64% (74% / 54%)
Arkansas	8-5 (5-3)	15	17	7.3 (3.9)	31	63% (37% / 89%)
MSU	9-4 (4-4)	16	21	7.4 (4.0)	27	56% (44% / 69%)
A&M	8-5 (4-4)	42	25	6.7 (3.3)	10	65% (59% / 72%)
Auburn	7-6 (2-6)	35	24	6.5 (3.4)	5	62% (64% / 61%)

SEC East

Team	2015 Record (Conf.)	2015 S&P+ Rk	Proj. 2016 S&P+ Rk	Proj. 2016 Wins (Conf.)	5-Yr Rec. Rk	Ret. Experience (Off. / Def.)
Tennessee	9-4 (5-3)	22	9	8.6 (5.2)	15	81% (84% / 78%)
Georgia	10-3 (5-3)	32	15	8.5 (5.2)	6	76% (75% / 76%)
Florida	10-4 (7-1)	18	19	8.0 (4.8)	12	61% (63% / 59%)
Missouri	5-7 (1-7)	85	47	6.1 (3.1)	33	75% (82% / 69%)
Vandy	4-8 (2-6)	83	69	4.9 (2.6)	40	71% (59% / 83%)
Kentucky	5-7 (2-6)	95	83	4.3 (1.9)	32	62% (62% / 61%)
S.C.	3-9 (1-7)	88	63	5.4 (2.8)	22	62% (61% / 63%)

** Link to Bill C.'s SEC capsule **

You may not be aware, but Nick Saban is pretty good.

Seriously - what he's done at Alabama is ridiculous. The DaggaStats only go back to 2005 but he's the best coach in the dataset by a country mile. I can't think of a coach in the modern era whose teams have been so consistently elite. Bobby Bowden is the only one who even comes close, I think. The SEC has superb talent all over the place and Bama has to replace a lot of talent. But picking anyone else to win the conference is ill-advised until further notice.

For the rest of the league: LSU returns errrrbody. Les Miles will be under pressure to challenge Bama this year or he might get Mark Richt'd.

Speaking of Richt: it's hard to see why he got fired when you look at Georgia's 2015 record. It gets a little easier when you look at their actual quality of performance. Per adjusted stats they finished 8th in the conference in overall quality. As a late-era Mack Brown fan, I get it. I really do.

Jim McElwain fans might want to take a deep breath because despite their 10-win year in 2015 they were only 7th in conference in Q rating and return the least amount of experience in the SEC East. They might impress again but the odds are somewhat lower this year than last.

Really, this is Butch Jones' best chance to re-establish Tennessee as the team to beat in the East. Patent those bounce pass touchdowns and they might run the table, methinks.

A&M's projected S&P+ rank is 25th, because Bill C.'s system assumes the teams regress to the mean with regards to recruiting talent. I'm not sure if Bill C.' s system takes into account the exodus of five star Aggie QBs. Nor does it consider the effect the Kyle/Kyler flashbacks are having on the Borg-Collective-slash-Napoleonic-Complex known as the 12th Man™.

And here's the kicker: if they live up to that preseason S&P+ ranking, they'd still be likely to finish fifth in their division.

I anticipate we may get a blissful redux of late-stage Sherman: disheartened play, folding into untimely on-field catastrophe, folding into open fan dismay, folding back into disheartened play, repeating ad infinitum like a particularly glum and goat-themed MC Escher painting.

(crosses fingers) (knocks on wood) (pushes Aggie down a well)

My Pick to Win: Texas A&M

Just Kidding, My Actual Pick to Win: Alabama

Most Likely Art Briles Landing Spot: Texas A&M