The Texas Defense: Or How To Create A Statistical Outlier
Posted by Scipio Tex – Barking Carnival
I’ve documented the weird disconnect between our defense’s raw performance in yards allowed and yards per play vs. points allowed since 2009, but this year that trend has only worsened and is, in large part, responsible for the fact that we’re sitting here at 4-4 rather than 7-1.
Despite what many believe, limiting opponent scoring output is a function of team play, not just your defense.
Similarly, last year’s defense and special teams proved that offensive scoring output is not just a function of offense, scoring a record 11 non-offensive touchdowns along with many short fields. We were only slightly above average on offense last year – struggling mightily in five games – but we were extraordinary at team scoring. To the unsophisticated observer, team scoring is mistaken for great offense just as scoring allowed is always (dis)credited to the defense.
My hope is to enrich that understanding.
To prove my point, consider the top 10 major FBS defenses ranked in yards per game in the NCAA (I exclude Kent State as a MAC team for what I hope are obvious reasons):
1 TCU 217.3
2 Ohio St. 234.2
3 Boise St. 236.4
4 West Virginia 249.8
5 Texas 267.0
6 Utah 267.7
7 LSU 277.6
8 UCF 279.8
9 Arizona 286.8
10 Alabama 291.3
Excluding Texas, those nine teams have a combined record of 64-9 and a .877 winning percentage.
That’s an average record of 7-1, including three of college football’s four unbeaten teams.
So how is it Texas could be 4-4?
Because our defensive raw output doesn’t translate to scoring defense. Which it does – very neatly – for almost every other team in America.
1 TCU 8.67
2 Alabama 12.5
3 Boise St. 13.4
4 Ohio St. 13.5
5 West Virginia 13.6
6 Utah 14.1
7 Arizona 14.3
8 Iowa 14.50
9 Missouri 15.3
10 LSU 15.6
Do those bolded names look familiar? Look at the first set of rankings. The overlap between yardage allowed and scoring defense is almost total – 8 of 10 teams and 7 of the first 7. BTW, Iowa is ranked #12 in yardage allowed and UCF just missed making the points per game list, ranking #11. So this isn’t correlation, it’s causation. If you want me to run the p-value, I can get Hucklerobot on it, but take my word for it that we’re looking at substantially less than .05.
Missouri is the only outlier, ranking in the low 50s in yardage allowed, but still checking in at a respectable 15.3 ppg. Missouri’s bend-but-don’t-break schemes and an offense and special teams that don’t put points on the board for the opposing team are largely responsible. I expect those numbers to converge in the 20s as the season progresses (and it has begun), but on balance Missouri is putting on a clinic in how to protect your defense in the other phases of the game.
The clear outlier here is Texas. There is no other team in the Top 5 (indeed, the Top 7) in yardage allowed not also found in the scoring defense Top 10. And you won’t find us in the Top 15.
Or the Top 25.
Or the Top 30.
Or Top 40.
There we are. We’re #41 in the country. Allowing 21.4 points per game. Talk about an outlier. Let me assure you that 21.4 ppg and 267 yards allowed per game are not a common pairing. To the point of statistical improbability. You have to work hard to achieve this. One should expect a ppg output more along the lines of 10-14 ppg.
So, what’s up?
Well, assuming you’re not a regular reader, or haven’t synthesized some of my themes since August, here are the broad strokes:
1. Bad offense. However, it’s not simply a function of overly conservative or bad offense. There are somewhat careful bad offenses (LSU, West Virginia) and conservative good offenses (Ohio St, Alabama). And there are explosive offenses that allow your defense to play with a lead (Boise State, Utah).
We have a careless, bad offense. That is a deadly combination. Turnovers + lack of production = bad juju. A high scoring but careless offense (high output, high turnover) or even a conservative mistake-free offense (low output, low turnover) would greatly aid our calculus.
We can do neither. We are low output, high turnover. With a particular knack for offering up a short field. Though our ability to generate 4th quarter yardage is enviable.
2. Special teams. We turn the ball over in the punt return game inside our red zone, allow the field to shift with poor decision making (70th in net punting), and are awful on kick returns (115th in the nation). Special teams hold hidden yards and they are decisive in football games. That’s why I write a special teams post-mortem and why I babble about them so much, much to your irritation.
Like your mother making you eat your brussel sprouts, I’m doing it because it’s good for you.
Mack Brown believes we won the Nebraska game because of some mystical focused energy and because we “played hard.” In fact, all of his explanations for game outcomes are almost entirely emotional. Which is incredibly worrying. Emotion is just as often a result of good play as it is the cause.
In fact, the Nebraska game was very simple: our offense didn’t turn it over – playing offense in a way that would make SweaterTress smile – and we dominated field position with special teams.
Between those two things, we accumulated 150 hidden yards. Nebraska had to attack long fields. We attacked short fields. That’s how you win at war, in business, and at football.
Against UCLA, Iowa State, and Baylor we pissed away the same amount of hidden yardage the other way. That’s how they could score 30.7 ppg on average against us while averaging 318 yards per game. 30.7 ppg and 318 yards per game don’t match up. You have to look to special teams and offense to find those hidden yards of output. Yet the defense still gets pinned with “allowing X number of points.”
3. Team play incompatibility with our defense. This may be a subset of #1 and #2, but it deserves its own heading. From a macro-standpoint, we’re very poorly constructed to support a defense that would thrive off of playing with a lead and 70-80 yards of field to defend. The rules, style of play, and structure of college football today mean that your offense and special teams should reasonably expect to put up points. Lots of them. If you’re of the belief that 17-21 points should be enough to win a college football game with the way we expose our defense, you’re living in another era.
Scoring trends are up. Way up.
It’s not 1977. WRs aren’t lining up in a three point stance. Offensive linemen can use their hands and can legally hold if they get their fit on a defender. QBs nowadays are selected for actual passing ability and athleticism instead of gutty leadership. The playmaking QB on offense is now someone to be loosed rather than tamed. Offense has been the beneficiary of every rule change in college football over the last 30 years, with the only exception perhaps being possession rules on a catch.
Play accordingly. Recruit accordingly. Scheme accordingly. We do not. And did not. And will not.
4. The defense itself. For all of its positive metrics, we’re not great on 3rd down (36%, 31st in the country, which is very acceptable, but not great), largely because of inadequacies in safety play and an amazing ability to collect 3rd down penalties. Additionally, the turnover fairy has been unkind, particularly as it relates to fumble recoveries. We’re 6 of 23 on the year in recovery and it has been proven pretty conclusively that recovering fumbles (not causing) is a coin flip. We’ve had a run of tails.
A lack of interceptions is as simple understanding the difference between Earl Thomas at safety and not having Earl Thomas at safety. Thomas was responsible not only for the interceptions he grabbed, but he also protected Blake Gideon, allowing him to operate essentially as the end of a turnover funnel with no responsibilities and the ability to stand in center field catching pop ups. No Earl, no easy picks. And with two safeties – neither of which can cover – we’re forced to play predictable coverages that don’t allow the secondary games that enfeeble QB minds and create easy picks.
This is not a push for dumb equivalence – I’m merely pointing out that our offense and special teams are operating at 20% maximization while defense operates at 85%. The defense isn’t perfect, but it’s playing at an A- level while the offense and special teams are a D- and C- respectively.
Statistics are useful when you know which ones matter. When you don’t, the discourse looks a lot like this:
Right now, I suspect we have some younger guys on this staff who know which ones matter and, more crucially, what they mean. Equally, we have some other guys on this staff who will create their own statistical categories when it suits them or cherry pick favorable ones while holding some units to impossible standards. The tension between them and how it all plays out will tell the story of Longhorn football and how Mack Brown defines his ultimate legacy.