We can all agree that the team Alabama put on the field this football season was an impressive one. They mauled everyone they faced (hooray for second chances!) and did it mostly on the strength of a scary good defense. Quantifying such a defense is one of the joys of the off season, when there are no distracting Tulsa-Toledo games on Tuesday night you can really dive into the stats.
People are doing that, and the results in numbers are every bit as impressive as the Tide were on the field. The defense was such a force that even our frenemies at Roll Bama Roll magnanimously gave a nod to one of Auburn’s best defenses ever: “To find a team that gave up fewer points per game than this Alabama squad you have to go back to the 1988 Auburn squad that allowed just 7.67 points per game.”
Kleph studied how Bama stacked up against the competition, digging through dusty old databases and figured up points per game. He did yeoman’s work studying the scoring defense numbers. (It is a fine post and I commend it in full to those who appreciate defense.) but there are, of course, always more layers in the numbers.
Because defenses do not live in a vacuum we wanted to take this one step further.
Lies, Damned Lies, Statistics
Give someone long enough and they can make any numbers look good. Run enough regressions, ANOVAs, Chi-squares, Pearson product-moment correlation coefficients, Spearman’s rank correlation coefficient and you can convince your data to sing you a song. We’re just going to play with a few simple averages, though, and apologize for it below. But the thinking is this: comparing defenses is a great off-season diversion, but it is a fruitless exercise. Simply put, things change all the time in the game. The numbers are a constant source of comfort to those that fear change, but they’re moving around, too.
Offensive styles change. Rules change. Athletes, schemes, jersey technologies, they all have been in flux. Steve Spurrier has so much as admitted that the reason he was successful at Florida in those early years was because he was willing to throw the ball all over Gainesville when seemingly everyone else had forgotten how.
So with that in mind, let’s compare three great defenses: ’86 Oklahoma, ’88 Auburn and ’11 Alabama. First you’ll see their schedules, the outcomes and a few notes on the opposing offenses. Below all that we’ll have a few crunched numbers and, finally, the limitations of this brief bit of research.
1986 Oklahoma, 11-1 under Barry Switzer:
|
W/L |
PF |
Opponent |
PA |
Opp PPG |
Opp Nat’l Rank PPG |
|---|---|---|---|---|---|
| W | 38 | UCLA | 3 | 32.1 | 8th |
| W | 63 | Minnesota | 0 | 21.8 | 61st |
| L | 16 | Miami (FL) | 28 | 35.8 | 3rd |
| W | 56 | Kansas St. | 10 | 12.2 | 102nd |
| W | 47 | Texas | 12 | 20.8 | 66th |
| W | 19 | Oklahoma St. | 0 | 16.5 | 93rd |
| W | 38 | Iowa St. | 0 | 23.4 | 51st |
| W | 64 | Kansas | 3 | 10.2 | 104th |
| W | 77 | Missouri | 0 | 17.8 | 83rd |
| W | 28 | Colorado | 0 | 20.2 | 70th |
| W | 20 | Nebraska | 17 | 37.2 | 2nd |
| W | 42 | Arkansas | 8 | 25.9 | 29th |
1988 Auburn, 10-2 under Pat Dye:
|
W/L |
PF |
Opponent |
PA |
Opp PPG |
Opp Nat’l Rank PPG |
|---|---|---|---|---|---|
| W | 20 | Kentucky | 10 | 19.7 | 72nd |
| W | 56 | Kansas | 7 | 17.2 | 90th |
| W | 38 | Tennessee | 6 | 19.3 | 76th |
| W | 47 | North Carolina | 21 | 19.7 | 73rd |
| L | 6 | LSU | 7 | 20.8 | 66th |
| W | 42 | Akron (OH) | 0 | 21.2 | 62nd |
| W | 33 | Mississippi State | 0 | 15.6 | 96th |
| W | 16 | Florida | 3 | 22.3 | 56th |
| W | 38 | So. Miss | 8 | 29.4 | 24th |
| W | 20 | Georgia | 10 | 29.8 | 20th |
| W | 15 | Alabama | 10 | 26.4 | 34th |
| L | 7 | Florida State | 13 | 37.9 | 5th |
(Some guy named Mack Brown put up the most points on Auburn that year. North Carolina was 1-10 on the season.)
2011 Alabama, 12-1 under Nick Saban:
|
W/L |
PF |
Opponent |
PA |
Opp PPG |
Opp Nat’l Rank PPG |
|---|---|---|---|---|---|
| W | 48 | Kent State | 7 | 17.1 | 114th |
| W | 27 | Penn State | 11 | 19.3 | 110th |
| W | 41 | North Texas | 0 | 24.8 | 77th |
| W | 38 | Arkansas | 14 | 36.8 | 15th |
| W | 38 | Florida | 10 | 25.5 | 71st |
| W | 34 | Vanderbilt | 0 | 26.7 | 61st |
| W | 52 | Ole Miss | 7 | 16.1 | 116th |
| W | 37 | Tennessee | 6 | 20.3 | 106th |
| L | 6 | LSU | 9 | 35.7 | 17th |
| W | 27 | Mississippi State | 7 | 25.3 | 25.3th |
| W | 42 | Auburn | 14 | 25.7 | 70th |
| W | 21 | LSU | 0 | 35.7 | 17th |
On one hand, even without doing any math, you can see how the modern game puts Alabama’s defense in something of a bind when it comes to comparing statistics.
In 1986 35.8 ppg was good for 3rd in the nation. In 2011 35.7 ppg was merely 17th. Tennessee posted 19.3 ppg in their 1988 campaign, ranking 76th. Penn State had the same average in 2011, and their final rank was a disappoint 110th. Nebraska’s 37.2 ppg was one of the nation’s best in 1986. This year they’d merely find themselves in a shootout with Arkansas.
On the other hand, Oklahoma played two teams impotent enough to rank in the triple digits. Auburn had no such opponent in 1988. Alabama faced four even in a more offensively offensive era.
So, to the math then:
Simply taking the opponents ppg and dividing across the season, and then doing the same with those subsequent ppg ranks we see the best defense is … ’86 Oklahoma. But only barely.
As a collective, Oklahoma’s opponents in 1986 averaged 22.82 points per game. The average opponent PPG rank was 56. Auburn’s 1988 foes averaged 23.27 points per game that year, also lining up neatly with an average opponent PPG rank of 56. Alabama stared down teams scoring 25.75 points per game. But the average opponent PPG rank across the year was 70th.
Some notes on limitations: First, I did not include Georgia Southern in calculating Alabama’s numbers this year. Across these three championship-caliber teams that was the only non D-1 opponent. (Akron had just made the jump before visiting Jordan-Hare Stadium.) Other things we can’t know include how many of those points scored against Oklahoma or Auburn were against their defenses. How many special teams points were put up against them? Were there pick-sixes? Fumble returns? How often did these defenses line up on a short field? What about situational statistics like garbage time? This depth of data entry hasn’t made it into the online records, but they would figure into a truly rigorous examination.
Ultimately we’re talking about less than three points per game across a quarter of a century of football which should perhaps include the 2011 Alabama defense in this conversation. Sure, Oklahoma and Auburn line up better in the numbers here, but Alabama’s defense is pretty close. Without trying to control for the productivity of offensive improvement in the intervening years their effort is still subjectively sound. Less than three points. One good field goal mitigates the difference.
It always comes down to field goals.
Source material:
Oklahoma’s CFBDatawarehouse
Auburn’s CFBDatawarehouse
Alabama’s CFBDatawarehouse
1988 SEC summary
Various PPG and ranks across the years, via Sports-Reference
Kenny graduated from Auburn at the turn of the century. He worked in newsrooms across the region and then earned a master’s degree at UAB. He met and married a Yankee, who declared her Auburn allegiance at her first home game. She’s now on the faculty at Auburn. He’s finishing his PhD at Alabama and teaches at Samford University. See him online at www.kennysmith.org and @kennysmith.
…
Keep Reading:
* Auburn vet saves Lady Hoggers
* The 1973 Delta Chi Miss Hot Pants pageant
* Bjork wrote a song about Auburn’s “Hey Day”
* Crazy Auburn mural at Firehouse subs
* Auburn Playboy shoot at Chewacla during power outage
* Glee star Naya Rivera wears Auburn shirt in FHM photo shoot
* The giant card Tampa Bay fans sent to Bo Jackson
Like us on Facebook. Follow us on Twitter. Want to advertise?
One thing I noticed about those numbers… you left out the performance of my Georgia Southern Eagles against Bama. 341 total yards, 302 rushing, 21 points (seven were special teams points). I’m guessing you weren’t using FCS numbers for the reason that they are FCS. But that was still one of the Bama D’s worst performances on the year.
Wait. If Oklahoma faced offenses with a lower PPG average, then they were facing easier offenses than AU or bama faced. How does that make OU better?
And as long as we are making the point that offensive scoring is higher these days, then I don’t think the rankings matter at all. Bama faced better offenses than those OU and AU teams did. Period.
Bottom line:
Bama: 7.08 PPG / Opp 25.75 PPG = Held opponents to 27.5% of their average production
OU: 6.75 / 22.82 = 29.6%
AU: 7.76 / 23.27 = 33.4%
This bama offense is the best ever assembled in college, in my opinion. Hate to say it, but it’s true.
I also agree from withholding GSU from the analysis because 1) their PPG versus a FCS slate is irrelevant and 2) bama clearly decided not to work on the option that week and rely on athleticism alone. We all saw how well that option worked for LSU (twice) when bama took the time to prepare for it.
“This bama DEffense is the best ever assembled in college, in my opinion. Hate to say it, but it’s true.”
NOT offense. Sorry.
Alex,
You make a fine, fair point. There are a handful of ways we could break these numbers down and make all three sides look great — easy since they were all great sides, really. The difference, from a football point of view, is negligible.
Without, hopefully, being too much of a cherry-picker here, the best comparison may be in the relative strength of opponent offenses, Bama’s opponents being a bit more diluted, despite playing in a more offensive-minded era. But you play who shows up and they performed admirably.
Walt — that’s why, you’re right! I would have included them if each of the teams played one, but the schedules didn’t work out that way. The biggest problem to me, even in this most lax of analysis, would have been how do you reconcile where Ga. Southern’s offensive statistics would fit into the D-I pile.
I just cannot swallow the relative rank argument.
By that logic, if all three teams had played the top 12 scoring teams of their respective years, then the one which gave up the lowest point total would be the best defense — regardless of the fact that teams today score significantly more than they did in the 1980s.
Also, if relative rank is the best tool, then why even perform the analysis? Let’s just say that all three defenses are equal because they had the same relative rank (#1) in their respective years.
Either way, as you point out, these are all great defenses. I’d take a repeat of the 1988 defense on The Plains ANY TIME. Maybe BVG will bring us back to that point.
WDE!
thanks for the mention and the link to my post. i do want to point out one thing though, given the incomplete nature of a lot of the old stats, my intention was simply to provide as good a general comparison of the teams historically as i could relying as much as possible on the agreed-upon authority for this information.
would it be better to, say, remove special teams points to get a more clear look at what the defense actually accomplished? of course. but i don’t have a breakdown of that available for every team over the past quarter century.
my real hope was to use what numbers are available to create an apples to apples comparison that might suggest other areas deserving of more detailed examination. which, i have to add, is what this post has accomplished very well.
Kleph,
Yours was great. I was simply trying to think of all the places where we are limited based on what information we had between us. That wasn’t meant as a criticism of your work by any means. I apologize if you took it that way as I was impressed by what you (and I) were able to put together considering the readily available material.
Thanks for the kind words on your part. The apples-to-apples comparison would be great — which is part of what Alex is giving me grief over — but short of a video game function or something even more silly, like serious statistical interrogation, I’m not sure if it is possible. That’s why I am just pleased to be able to say those are three great defenses, notwithstanding their respective era or opponents. People can pick their own variety of apple from that.
(There’s an interesting discussion going on in the NFL right now about what stats mean because everything has been so thoroughly reshuffled the last few years. To a sensibly lesser degree that is probably true among a great deal of college football, too.)
Thanks again!
Those who cannot do, criticize. I excel in this arena.
Great work – both of you!
Alex,
You’re points are good, as I said. We’re limited in what we know/recall/can data mine. And since the differences seem to be marginal on the surface — and knowing we’re likely missing something important — I’m having a hard time firing up SPSS to suss it all out.
Keep criticizing, my friend!
i offer no criticism to your work, kenny, and took none from your comments. i simply wanted to offer a clarification of what i was trying to accomplish with this.
Kleph,
You did it will, sir! Kudos.
Honestly (and you touched on this) I’ve always had a problem with the way offensive and defensive scoring are recorded.
PPG on both sides are never adjusted for special teams and defensive scores.
In my opinion, If my defense gives up one TD and then scores one on a pick six, they should net to zero in the scoring defense calculation.
Instead, both my offense and the opponent’s offense get credit for seven points that they did not earn.
Special teams points (PATs on non-special teams TDs excluded) shouldn’t hit either offensive or defensive scoring stats.
By all rights, these stats should be kept descretely as Net Offensive PPG, Net Defensive PPG, and Net ST PPG.
But there I go criticizing someone again.
In any discussion of greatest defenses of all time, one must consider the four major defensive categories (total yards surrendered per game, rushing yards surrendered per game, passing yards surrendered per game, and points surrendered per game) and one must compare those statistics to the contemporaries of the teams in question.
2011 Alabama, 1992 Alabama, and 1986 Oklahoma are the ONLY teams to lead all 4 major categories since 1980.
But most significantly, how did 2011 Alabama fare against its contemporaries in total yards? 2011 Alabama’s #1 defense surrendered 78 fewer yards than did 2011 LSU’s #2 defense. In the entire history of NCAA statistics (since 1936), there has never been that large of a difference in total yards surrendered between the #1 and #2 defense in ANY year.