Advanced Stats Breakdown: ND Defense

With the conclusion of the regular season, taking a look back at how the advanced stats view Notre Dame's offensive performance. Most stats come from the great work of Bill Connelly and the Football Study Hall team and Brian Freameau's Efficiency Index.

Defending the Run

Notre Dame Run Defense	Avg.	Rk	Nat'l Avg.
Rushing S&P+	105.1	51	100.0
Rushing Success Rate	40.5%	54	42.3%
Rushing IsoPPP	1.14	91	1.08
Adj. Line Yards	115.4	19	100.0
Opportunity Rate	34.8%	30	39.1%
Power Success Rate	66.7%	73	66.4%
Stuff Rate	23.1%	28	19.5%

The Irish improved from 70^th to 51^st this year in Rushing S&P+ defense, which is a small step in the right direction but far from adequate for a defense with high aspirations. Opponents rushed for 4.52 yards per carry despite the solid work by the defensive line with fairly high stuff rates and low opponent opportunity rates. The long runs were disastrous - when containing them, like against Stanford, opponents frequently were bottled up (3.56 yards per carry by the Cardinal is extremely impressive). In other games long runs (often by quarterbacks) ruined otherwise solid performance, like Boston College (5.94 YPC) and UMass (5.10). The power success rate is also disconcerting - for a team very talented up front, the Irish were beat too often in power and goal-line situations when both teams had a good idea a run was coming.

Individual Defensive Stats

Name	Pos	Ht, Wt	Year	Tackles	% of Team	TFL	Sacks	Int	PBU	FF	FR
Jaylon Smith	LB	6'2, 240	JR	90.5	14.4%	9.0	1.0	0	5	1	1
KeiVarae Russell	CB	5'11, 196	SR	54.0	8.6%	3.5	1.0	2	4	2	0
Max Redfield	S	6'1, 205	JR	51.5	8.2%	2.0	1.0	1	2	0	0
Elijah Shumate	S	6'0, 224	SR	50.0	8.0%	6.5	0.0	1	2	0	0
Joe Schmidt	LB	6'1, 235	SR	48.5	7.7%	3.0	1.0	0	3	0	0
Isaac Rochell	DL	6'3, 287	JR	43.0	6.8%	7.5	1.0	0	2	0	0
Romeo Okwara	DL	6'4, 270	SR	36.5	5.8%	13.5	9.0	0	0	1	0
Sheldon Day	DL	6'2, 285	SR	35.0	5.6%	14.5	4.0	0	2	1	0
Cole Luke	CB	5'11, 193	JR	29.0	4.6%	1.0	0.0	2	4	0	0
James Onwualu	LB	6'1, 232	JR	27.5	4.4%	5.0	2.0	0	2	1	0
Greer Martini	LB	6'2, 245	SO	27.0	4.3%	2.5	1.0	0	0	0	0
Matthias Farley	CB	5'11, 210	SR	26.5	4.2%	0.0	0.0	1	1	1	0
Nyles Morgan	LB	6'1, 240	SO	13.0	2.1%	0.0	0.0	0	0	1	0
Daniel Cage	DL	6'1, 315	SO	12.0	1.9%	3.0	0.0	0	0	0	0
Jarrett Grace	LB	6'2, 253	SR	12.0	1.9%	0.5	0.0	0	0	0	0
Jerry Tillery	DL	6'6, 305	FR	10.5	1.7%	2.0	1.0	0	0	0	0
Andrew Trumbetti	DL	6'3, 260	SO	9.5	1.5%	0.5	0.0	1	0	0	0
Te'von Coney	LB	6'1, 235	FR	8.5	1.4%	0.5	0.0	0	0	0	0
Devin Butler	CB	6'1, 200	JR	8.5	1.4%	0.0	0.0	0	2	0	0
Drue Tranquill	S	6'1, 225	SO	7.5	1.2%	2.5	0.0	0	2	0	0
Nicky Baratti	S	6'1, 210	SR	5.0	0.8%	0.0	0.0	0	0	0	0
Nick Watkins	CB	6'1, 200	SO	3.5	0.6%	0.0	0.0	0	0	0	0
Doug Randolph	DL	6'2, 255	JR	3.5	0.6%	0.0	0.0	0	0	0	0
Nick Coleman	DB	5'11, 185	FR	3.5	0.6%	0.0	0.0	0	2	0	0
Tyler Newsome	P	6'2, 205	SO	2.5	0.4%	0.0	0.0	0	0	0	0

Looking at the performance over the course of the season, you can see strong early performance and consistency early in 2015 go out the window with wild fluctuations down the stretch, much in part due to quality of opponent rushing offense. While my chart-making skills are a work in progress (in keeping with the theme of this post), what's nice is that an average YPC (4.0) and success rate (40%) are very close to matching up. You can clearly see the impact of big runs driving up YPC in some performances where opponent efficiency was limited, especially against Boston College and Navy.

Past, Present, & Future:

• Assuming Jaylon Smith and KeiVarae Russell enter the draft, Notre Dame's defense will lose 56% of its tackling output, 66% of TFL, and 73% of its sack production. Even with a lot of talent on the roster, that's a ton to replace and ask for improvement.
• The loss of Jarron Jones left a gigantic hole in the Irish run defense, but will benefit futures defenses as a result. Jerry Tillery and Daniel Cage both showed flashes of brilliance and have multiple years of eligibility in front of them. Isaac Rochell has shown steady improvement year over year and should anchor a defensive line that shouldn't allow much push in the run game.
• You can make an optimistic case that with Jones returning next year, the front seven won't drop off too much. A defensive line rotation of Rochell, Jones, Tillery, Cage, and Bonner/Trumbetti should be formidable. On paper, the linebacking corps could feature a junior, five-star MLB (Morgan) and returning OLBs with a nice mix of talent and experience (Onwualu, Martini, and Coney). But a jump from another average run defense into the top 30 would require a few players making a leap to fill the playmaking void left by Day and Smith.
• The great Joe Schmidt debates will finally have some more evidence of just how valuable he was to this Notre Dame defense - without him directing players on the field, will Notre Dame's defense struggle early in 2016 like they did at the end of 2014?

Defending the Pass

Notre Dame Pass Defense	Avg.	Rk	Nat'l Avg.
Passing S&P+	109.1	36	100.0
Passing Success Rate	35.6%	26	40.3%
Passing IsoPPP	1.56	96	1.48
Adj. Sack Rate	90.7	83	100

Here's where the defensive scheme starts to become even harder to justify - frequent blitzing and aggression absolutely must translate into results. We've already covered that bringing a lot of heat hasn't resulted in a suffocating run defense, and while the pass numbers here are slightly better, there are still troubling issues. Notre Dame's adjusted sack rate was 78^th in 2014, and 83^rd this year despite the improvement of Romeo Okwara and the return of most starters along the defensive line. While you could argue there seem to be some good qualitative signs of development in Keith Gilmore's first year, and the Irish haven't had a true edge pass rushing threat, there's simply no way with the level of talent available (and an aggressive scheme) Notre Dame should be that bad at bringing down the quarterback.

The rest of the story here is a familiar one - the Irish overall were solid limiting opponents passing efficiency, terrible at giving up long bombs when they couldn't get the stop. The opponent passing success rate ranking (which is non-adjusted) may have been buoyed a little by facing several offenses that have serious issues throwing the ball like Texas, Georgia Tech, and Boston College.

Defensive Footprint

	Team	Rk	Nat'l Average
Std. Downs Run Rate	65.3%	19	60.1%
Pass. Downs Run Rate	36.3%	40	33.8%
Overall Havoc Rate	16.1%	64	16.5%
DL Havoc Rate	6.1%	45	5.2%
LB Havoc Rate	4.3%	68	4.6%
DB Havoc Rate	5.8%	85	6.5%
PD to INC	26.5%	115	32.8%

If bringing pressure isn't producing sacks, the hope would be that Van Gorder's frequent pressures and attempts at confusion would then lead not just to pass inefficiency, but also forced throws into tight windows. Unfortunately that hasn't been the case - Notre Dame's havoc rate (which includes sacks, TFL, and deflections) from the secondary was extremely low. The defense just didn't get their hands on the ball often in pass coverage, which is disconcerting. The Irish tied for 95^th nationally with just eight interceptions, and even some reasonable bounce-back in luck and regression to the mean places this defense around average at forcing turnovers.

This chart is very much skewed by the Stanford game, where the Cardinal did whatever they wanted through the air (66.7% success rate, 12.8 yards per attempt). The Passing S&P+ ranks in parentheses bring to life how many bad passing offenses the Irish defense faced - USC, Clemson, and Stanford were formidable, but half of these opponents ranked 72^nd or worse, including the worst passing offense in the land in Boston. Triple option teams also make this tougher to track week to week (0% passing success rate for Navy!), but show at offense that mostly limited bad opponents to inefficiency. Games with coverage issues (and trick plays!) like Virginia and USC also jump out.

Past/Present/Future:

• There are likely more questions than answers for the passing defense - other than Cole Luke, what should Irish fans feel good about? The replacement options for Elijah Shumate are either recovering from injury (Tranquill) or relatively unknown (Sebastian/Mykelti Williams/freshman?). Russell could return but Brian Kelly indicated NCAA clearance would be needed - does anyone finally step into that final corner position? It seems like 2015 was a constant rotation in public praise between Watkins, Coleman, and Butler, without any clear winner at the end of the year. Betting on at least one of that trio to solidify the 2^nd CB spot seems like a decent (but somewhat frightening) solution.
• While Redfield and to a greater extent Shumate were good against the run, blown coverages and lack of discipline were an extreme issue for the secondary. Communication was consistently stressed last spring - maybe new personnel will actually help? There has to be some potential for a rebound year even with less experienced bodies if they can elevate the communication and mental ability of the defense to make smart adjustments.

Big Picture:

As with Notre Dame's offense, FEI and S&P+ agree on the overall state of Notre Dame's defense- it was just ok. A high rate of three and outs (20^th nationally by FEI) and relative strength limiting opponents efficiency were offset by the fatal flaw of explosive plays allowed that usually translated into points. Notre Dame's defense allowed 1.87 points per drive, ranking 45^th nationally. When opponents were able to get momentum, they were rarely stopped - on opponent drives starting in their own territory but crossing Notre Dame's 30-yard line, opponent's scored 5.45 points per "value drive" (104^th), but these drives thankfully didn't happen often (or the defense would have been well below average).

Top 8 Teams According to Playoff Committee and Offensive Ranks:

Team	CFP Rank	Defense S&P+	Def. Success Rate+	Def. IsoPPP+	Defensive FEI Rank
Clemson	1	6	3	9	7
Alabama	2	1	1	1	2
Michigan State	3	13	21	11	15
Oklahoma	4	12	8	3	3
Iowa	5	26	46	26	24
Stanford	6	54	51	89	60
Ohio State	7	8	5	4	10
Notre Dame	8	33	25	55	52

"Defense wins championships" - not untrue this season, as all four playoff teams feature top-15 defenses according to both S&P+ and FEI. The outliers here are the Irish and the Cardinal, who made it into this top group by virtue of extremely efficient and balanced offenses. Even if a few plays go the other way and Notre Dame had made the playoff, looking at this table shows you why Notre Dame probably would have had the lowest chance of winning it all according to advanced stats (and Vegas). The Irish defense would have been by the worst single unit in the playoff, and that would have placed such an extremely high bar for the offense to hit (against excellent defenses) to have a chance to win consecutive games.

Overall both FEI and S&P+ paint a picture of a defense that's not all that far away from being a top 15-25 defense. Notre Dame's opponent-adjusted success rate (25^th) is awfully close to some of the country's top defenses - Florida State (27^th), Michigan State (21^st), and Washington (23^rd, and yes the Huskies defense flew under the radar but was extremely good this year). The difference between the Irish and those teams is the explosive plays given up when they did allow success - those teams rank 5^th, 11^th, and 14^th respectively in limiting explosive plays, while the Irish rank 52^nd.

That really shapes the critical question for Brian Van Gorder's defense going forward - can Notre Dame maintain the same ability to limit opponents' efficiency while getting better at stopping big plays? Is it a problem of scheme or execution, or simply not adjusting scheme if you don't have the players to execute what you want to do?

Past, Present & Future:

• Notre Dame overall improved from 43^rd to 33^rd in Defensive S&P+ from last year, but that still feels like a loss given the injury woes that were a viable explanation for 2014's struggles and the amount of talent on this defense. Opponent adjustments play a big role here, but explosive plays allowed actually improved as well from 80^th nationally to 55^th.
• With the most talented player at each level of the defense likely headed to the NFL Draft, where do the Irish go in 2016? If we're realistic I don't think a dramatic change in Van Gorder's philosophy is coming, so there are really two options.
  • First, you can stay the course - hope that things start to click in Year 3 of this scheme, that even with the anticipated losses, talented players fill in and Notre Dame continues to force a lot of 3-and-outs and has enough improvement on the back end to limit explosive plays. This places a huge burden on player development to make that happen, in particular for the linebackers and secondary. Even if explosive plays only improve slightly, there's a chance the defense can improve as a whole by forcing more sacks, TFL, and turnovers.
  • Alternatively, you could see a shift not towards a full bend but don't break defense, but a measured move in that direction. This requires a lot of trust in Kelly as well as Van Gorder to self-identify issues defensively, potentially reducing the complexity not necessarily in scheme but in keeping players focused on doing what they do best and slightly less aggressiveness with blitzes. The result is like a uptick in opponents' efficiency but better limitations of explosiveness, and it's a calculated risk that you can limit the big play enough that the likely tradeoff in success rate is a net positive.