Charting The 2011 Texas Longhorns Defense

by tearaway20 on Aug 31, 2025 2:02 AM CDT in Texas Longhorns and Other Football

LonghornScott's recent post about last year's Kansas State game got me thinking about how much the defense improved over the course of the season. Enticing to forecast 2012 based on that, given that last year was Manny's first and that damn near the entire 2-deep is back. Probably dangerous, but good fun.

Here is a quick take that is simple yet I think pretty solid. I use yards per play relative to the offensive team's season average. This is a whole lot better than raw stats, but not as good as more carefully normed data or some fancier stat. But simple to understand and probably >75% of what any other single number will do for you. Using this stat, last year's top defenses were: Alabama (-2.06), FSU (-1.37), South Carolina (-1.31), Texas (-1.25) and LSU (-1.20). Feel pretty good about that top 5, and even the separation between Bama and the others, which I've seen in much fancier stats.

So here are/were your Horns:

Texas_defense_2011_medium

For context, -2 and lower is top 5% - only 80 of ~1500 games last year where that low (Alabama had 7 of them, btw). -3 and lower is top 1%, with Texas ringing up 2 of those 15 games.

A few quick observations:

Impressively consistent.
Baylor was even worse than it felt, and OU probably not as bad.
Validates LonghornScott's impression of the Kansas State game, though also putting KU on the table.
Lovely trend. Good fodder for optimism.

Tweet Comment 22 comments | Add comment | 1 recs |

Do you like this story?

Comments

To further emphasize your already very valid point:

These results above are corrected for opponents’ averages as well. That Baylor game. Yikes..

by pleaseplaykindle on Aug 31, 2025 3:26 AM CDT reply actions

RGIII

So freaking good.

Ubbenator

Last season, Baylor set or tied 101 school records for offense.
The Bears were the nation’s only team to finish in the top 10 in rushing and passing offense.

Our O, so freaking bad.

Follow @BarkingCarnival

by Sailor Ripley on Aug 31, 2025 9:45 AM CDT up reply actions

Love the graphs, but...

Our D ranked 64th nationally in Pts Allowed vs. Ranked Opponents (6th in Big 12). While OSU, Baylor, and Sooners had wicked good O’s last season, it’s hard to claim we rocked the house.

Not sure how OSU came it at an adjusted -3 yards/pass and OU adjusted to -1 rushing and 1 yd/pass. Sure didn’t feel that well “adjusted” watching those games.

by BevoBytheHorns on Aug 31, 2025 1:06 PM CDT up reply actions

Primary fallacy: that points allowed is strictly a measure of defense.

by Scipio Tex on Aug 31, 2025 2:26 PM CDT up reply actions

When your offense doesn’t work, your field position suffers. When your offense turns the ball over a lot, your field position suffers. When BOTH happen, your team gives up a lot of points, and that makes your defense look bad—weather it is or not.

by Stiendam Hall on Sep 1, 2025 5:12 PM CDT up reply actions

Wow, I understand both of those graphs...

The overarching difference between last year’s defense and this year’s is that the Texas defense won’t have to learn Manny’s concepts on-the-fly and try to execute them as designed so much as just let root knowledge take over and play. Plus, even without Robinson and Acho, I like the uptick in athletic ability of this year’s squad over last year’s, particularly at safety and NG/3-tech. It may take until mid-season for the LB’s experience and knowledge to match their athleticism…I’m hoping by OU Hicks, Edmond, Cobbs, Jackson, Thompson, and Santos are close to playing downhill with their hair on fire.

by dkrandmack on Aug 31, 2025 8:05 AM CDT reply actions

The difference needs to be the offense.

Follow @BarkingCarnival

by Sailor Ripley on Aug 31, 2025 9:46 AM CDT up reply actions

+1. About 1,111 times +1

Any sign of actually competent offensive play — even a reduction of 1 turnover per game by the offense — puts last year’s defense ahead of all but ’Bama’s. And makes this year’s scary good.

Both charts get me riled for another reason: The damned giveaway of the K-State game.

Mizzou, we didn’t have any running backs. Or Shipley. Baylor was RGIII. The Oklahoma schools were just better than us, though the OU score was inflated because of offensive and special teams turovers.

But Kansas State was ours to win. May I live long enough to see us begin balancing the scales against the Wildcats!!!!!!

by edsp on Aug 31, 2025 2:16 PM CDT up reply actions

good stuff

another factor to consider would be standard deviations in addition to comparing our performance to the mean. I would bet Huck already has that data. It would also give you an even better ability to compare defenses on an apples to apples basis across CFB.

by LonghornScott on Aug 31, 2025 9:25 AM CDT reply actions

you're killing me

The one time I don’t z-score something. ;)
Agreed fully of course. Went simple (and quick).

by tearaway20 on Aug 31, 2025 11:06 AM CDT up reply actions

Great stuff

Supports fans assertions that UT’s defense improved over the year. This is not news for anyone, but I think injuries during the season are a significant factor in progression/regression in performance as depth is a critical issue for most college teams. This isn’t currently being tabulated but should be accessible. I wonder if there is a mean +/- St Error that could be calculated.

My gestalt is that UT’s defense was relatively injury free last year. Tech fans have been vocal about their team’s injuries. Could relative number of injuries explain and/or predict in-season progression/regression?

by Quigley on Aug 31, 2025 10:40 AM CDT reply actions

this is a great point

We probably take last year’s health on D for granted. The cost of injuries, especially broken down by position, is about to be a hot little frontier in analytics.

by tearaway20 on Aug 31, 2025 11:08 AM CDT up reply actions

The “conventional wisdom” is that indicates that losing starters should hurt teams. However, when it was looked at in the past, it was found that this really didn’t impact win/loss. From Phil Steele:

“In a study over the last 8 years, my research shows that if a team had 32 or more starts lost to injury the prior season, they improved or had the same record the next year on 62 out of 76 occasions for an 81.6% success rate. How about teams that are really banged up? Well from 2001-2005 12 teams went through seasons of having 40 or more combined starts lost and ALL 12 had the same or better record the next season. It surprises me that 17 teams met that criteria in 2006-’07 but only 7 had a stronger record the next year and 9 had a weaker record.”

However, as you mention, starters at particular positions may be critical. I think OL starts may be the most important, followed by CB, for different reasons. OL is suppose to be about individuals working as a team. Disruption of the OL grouping may be important. CB, in some schemes, are isolated. Their loss could be crippling to a defense.

The other way to measure it would be to calculate loss of a superlative player, say OU losing Sam Bradford and Jermaine Gresham for a whole season. Losing an “all-American” may be critical while losing an average player could be absorbed.

by Quigley on Aug 31, 2025 12:22 PM CDT up reply actions

by Quigley on Aug 31, 2025 12:23 PM CDT up reply actions

"Conventional Wisdom" isn't much good any more...

It’s a new game. Against fast paced O’s (i.e. spread), starters on both sides of the ball will take a much lower % of snaps. In old school play starters might take every critical snap. These days against a Baylor, OState or OU our D starter might take 50-60% of the snaps (if we want to try and keep up). Net effect is “Starter” is much less meaningful than “2-deep”.

With our RB stable, think maybe H/AW look at teams that are 1-deep on quality DT’s and LB’s? Think Monken/Briles/Huepel are looking for 1-deep quality Corners to exploit on the next sub?

by BevoBytheHorns on Aug 31, 2025 1:14 PM CDT up reply actions

Point taken

Your point would explain why “starters lost” doesn’t predict wins. I’m suggesting the “conventional wisdom” should be tested by factoring some granularity, other variables like specific position groups or exceptional players.

The null hypothesis is that starters lost don’t matter, which is what Steele’s data shows. This finding would negate the whining of some fanbases that claim that injuries prevented a good season and also impact the way betting spreads change base on injuries/unavailable players.

The original poster and several people here like to use data to back up our assertions. I guess what your asserting is testing whether loss of players on the 2-deep predicts win/loss. Is that correct?

by Quigley on Aug 31, 2025 2:27 PM CDT up reply actions

Good stuff.

The rare visual that really is worth a thousand words. Have to take your hat off to Baylor. RGIII was sublime, their OL did a really nice job, and Briles did a great job of picking at our inability to cover ground at safety.

by Scipio Tex on Aug 31, 2025 10:55 AM CDT reply actions

such an outlier

Made me wonder about the extent to which it was a matchup issue with Baylor.

by tearaway20 on Aug 31, 2025 11:11 AM CDT up reply actions

Matchups and an amazing dissection of our personnel by Briles.

And RGIII is as good as it gets at the college level.

by Scipio Tex on Aug 31, 2025 11:31 AM CDT up reply actions

Nice Graph

And good on you for using Stata. Just out of curiousity, what do the lines like for the other Top 5 schools?

by Dave Stoller on Aug 31, 2025 12:19 PM CDT reply actions

Briles is really doing some good things at BU

by Randy Watson on Aug 31, 2025 12:44 PM CDT reply actions

It's not a total shock on Briles

He was a superlative HS coach

He was part of the offensive machine at Texas Tech

He took a UH team that wasn’t far from 0-11 to national prominence, and the turnaround was more than just 1-2 QBs

If anything surprises me about Baylor, it’s that they’ve recruited at so high a level so soon into Briles’ tenure. It’d be scary to see him run the show at OU or Nebraska.

by edsp on Aug 31, 2025 2:21 PM CDT up reply actions