Statistics make two things abundantly clear, and I'm going to get them out of the way immediately: 1) Nebraska's defense last year was horrifically horrific, 2) Joe "Small Sample Size" Ganz put up some pretty obscenely good numbers at the end of the season. Because of that, there's really no reason to plow through my typical gamut of stat categories...the defensive numbers all point out the same thing, and the offensive numbers don't really mean anything thanks to the QB change. Plus...with the '+' concept that I introduced a few days ago, we can kind of take these BTBS posts in different, more team-specific directions.
Now that that's out of the way, let's use this BTBS post to look into 1) why was Nebraska's defense so bad last year, and how much improvement is possible in just one year, and 2) how much can be expected from Ganz over a larger sample size? But first...
2007 In Review
Nebraska had had some embarrassing moments before last season--long winning streaks against the likes of Kansas and Missouri getting broken in blowout fashion, giving up 70 to Texas Tech, etc.--but last season was one car crash after another. The unraveling began with the USC game. NU fans thought they had a legitimate chance in the game, and it was over pretty much immediately. Stafon Johnson ran through holes through which I would have run for about 125 yards, the smoke and mirrors experiment known as Sam Keller evaporated, and Bill Callahan showed his true colors by trying desperately to tack on some late points, leaving his starters in against USC's 8th-string defense.
From then on, it was one step forward, eight steps back for NU. Sneak by Ball State and Iowa State, get demolished by Missouri, get humiliated by Oklahoma State at home. Hold your own (for a while) against ATM and Texas before crumbling late, then give up eleventy billion points and yards to Kansas. Crush K-State (becoming what almost has to be the first team to give up 70 one week then score 70 the next), then put up another embarrassing defensive performance against Colorado to end the most pitiful Nebraska season since my parents were in elementary school. It wasn't that they finished 5-7--that's honestly not bad for an all-out implosion. It's how they lost that made such an impact. Let's see how that impact translated to my BTBS numbers.
EqPts Scores by Game
Here's a game-by-game look at how Nebraska fared, offense vs defense (i.e. no turnovers, special teams, luck, etc.)
9/1/2007 - Nebraska 52.6, Nevada -0.2
9/8/2007 - Nebraska 17.0, Wake Forest 16.4
9/15/2007 - USC 39.4, Nebraska 24.7
9/22/2007 - Ball State 38.5, Nebraska 37.7
9/29/2007 - Nebraska 24.1, Iowa State 18.6
10/6/2007 - Missouri 44.0, Nebraska 9.1
10/13/2007 - Oklahoma State 47.3, Nebraska 15.6
10/20/2007 - Texas A&M 36.8, Nebraska 19.2
10/27/2007 - Texas 28.9, Nebraska 21.5
11/3/2007 - Kansas 65.7, Nebraska 31.1
11/10/2007 - Nebraska 62.6, Kansas State 29.8
11/17/2007 - Colorado 46.6, Nebraska 43.9
The point of this was to see how much outside factors were involved in Nebraska's demise. Was it just that they couldn't make the key plays when they needed to? Not so much. While the CU game was actually a lot closer and lower-'scoring' than the final score (65-51) would indicate, so was Iowa State (real score: NU 35-17)...and really, the only result that should have changed here was that they should have lost to Ball State...which may have kick-started the implosion a bit earlier.
What about the '+' measures? Nebraska played quite a few stout offenses in '07--USC (#17), Missouri (#7), Oklahoma State (#13), Texas (#15), and Kansas (#19) were all in my EqPts+ (Offense) Top 20--so at least a smidge of their defensive ineffectiveness can be attributed to that. But you don't reach the depths they reached by only getting lit up by those teams (though giving up 76 points to Kansas certainly helps). Let's look at two specific '+' measures, game-by-game.
EqPts+ and S&P+ Numbers and Trends (click to enlarge)
Let's look at the offense first. Three times in their first four games, NU's performance was well above what their opponents usually allowed. That includes a 239.4 EqPts+ against Nevada, and a 236.0 EqPts+ against USC (and yes, this includes 'points' gained when the game was well out of reach). A couple cracks appeared--USC sure seemed to dominate them when it counted, and they sure didn't set the world on fire against Wake Forest or Iowa State--but nobody could have guessed what would happen in the middle third of the season.
That's right--over the next four games, Nebraska averaged a 66.7 EqPts+...which is HORRENDOUS. How horrendous? A 66.7 EqPts+ for the season would have gotten them ranked the #113 offense in the country, between Iowa and "FCS Tier 4", the fourth tier of 1-AA teams that I tied together to use for averages. That's bad. Luckily, a) Texas' defense wasn't very good (the pass defense, anyway), and b) Sam "Jesus" Keller was about to get hurt. Nebraska averaged a 190.8 EqPts+ over the last third of the season, helped in no small part by a 203.5 against Kansas (once again while playing from behind) and an insane 247.1 against a Kansas State team that was coming apart at the seams.
So for the season, Nebraska's offense averaged a 146.3 EqPts+, good for #11 in the country. Which seems good...especially considering that middle third. But...well, two of their biggest '+' games of the year came while playing from way behind against USC and Kansas. How'd they perform while games were close? They were #44 in CloseRushingS&P+ and #48 in ClosePassingS&P+. Still above average, but not nearly as good as their final numbers would suggest.
(And you'll notice the goose-egg they put up against OSU in the 'CloseS&P+' category? How'd they do that? They attempted all of 9 plays on offense while the game was 'close', none of which were 'successful'. That's how quickly the game was out of reach. Very impressive.)
Now, the defense. As you see from the graph, they started on a high note and then established a level of steady mediocrity, putting up an 83.6 DefEqPts+ against USC, 82.4 against Ball State, 93.8 against Iowa State, 87.3 against Missouri, 81.0 against OSU, and 83.1 against ATM. That's nothing if not consistent. But after a relatively strong effort against Texas (104.0...their last stand), the bottom dropped out. Let's put it this way: "FCS Tier 6", the imaginary team that consisted of the worst of the worst in 1-AA, averaged 82.6 DefEqPts+ in 2007. Nebraska's DefEqPts+ over the last three games was 74.5. That's really all that needs to be said. They quit on Callahan and Cosgrove, and I guess you don't need my own fancy numbers to come to that conclusion. However, the consistent below-average play before the significant dropoff leads me to believe that the talent/athleticism and execution weren't there either.
So for the season, Nebraska's 95.2 DefEqPts+ average was good for #94 in the country. Take out the last three games (when they "quit"), and the average only rises to 102.0...good for the level of a Temple or Oklahoma State. Outside of those last three games, the Black Shirts established a level of consistency that was too, well, consistent to blame the whole mess on something as simple as "quitting". So with that in mind...
How Much Defensive Improvement is Impossible for 2008?
Let's start attempting to answer this question by trying to pinpoint exactly what was NU's biggest weakness. So let's plow through a bunch of S&P numbers:
- Overall S&P: 0.9095
- Rushing S&P: 0.9849
- Passing S&P: 0.8169
- Close S&P: 0.8695
- Close Rushing S&P: 0.9225
- Close Passing S&P: 0.8046
- Passing Downs S&P: 0.7387
- Non-Passing Downs S&P: 0.9831
- Redzone S&P: 1.0955
- Q1 S&P: 0.8641 (Rushing: 0.7919, Passing: 0.9407)
- Q2 S&P: 0.8730 (Rushing: 1.0495, Passing: 0.6848)
- Q3 S&P: 0.9112 (Rushing: 0.9837, Passing: 0.8227)
- Q4 S&P: 0.9993 (Rushing: 1.0946, Passing: 0.8341)
- 1st Down S&P: 0.8630 (Rushing: 0.8558, Passing: 0.8755)
- 2nd Down S&P: 0.9609 (Rushing: 0.9959, Passing: 0.9184)
- 3rd Down S&P: 0.8391 (Rushing: 1.3839, Passing: 0.5188)
- Pressure (4th quarter, score within 2 possessions) S&P: 1.0328 (Rushing: 1.2471, Passing: 0.6723)
So...1) NU's defense got worse as the game wore on (suggesting they bent over to either conditioning problems, momentum, bad coaching adjustments, or all of the above), 2) the more important the play, the more pliable NU's rush defense was, and 3) the pass defense wasn't amazing, but it really wasn't that bad. Oh, and 4) they sucked on second downs. I have absolutely no explanation for that.
Two things seem to be at work here--the talent/athleticism obviously wasn't where it needed to be, and this squad simply did not rise to the occasion. Bo Pelini can do nothing to address the first part of that, but he can have a big impact on the second. Let's check out where Pelini's LSU D had its strengths and weaknesses.
- LSU Overall S&P: 0.5549 (keep that in mind...their defense was much better, much more talented, and much more athletic. We're looking specifically at where they were better or worse than their overall numbers.)
- LSU Run S&P: 0.6326
- LSU Pass S&P: 0.4862
- LSU Close S&P: 0.6051
- LSU Q1 S&P: 0.5198
- LSU Q2 S&P: 0.5769
- LSU Q3 S&P: 0.6468
- LSU Q4 S&P: 0.5079
- LSU 1st Down S&P: 0.5647
- LSU 2nd Down S&P: 0.5690
- LSU 3rd Down S&P: 0.4415
- LSU 'Pressure' S&P: 0.5485
Some of the same trends at work here, to a degree. Better defending the pass than the run, trending worse from Q1 to Q3...however, Q4 was their best quarter. They rose to the occasion. The question is, how much of this was due to Pelini, and how much was due to the fact that the LSU defense was made up almost entirely of top-shelf athletes and players who aren't going to be suiting up for the Huskers in '08? The 2003 Nebraska defense (coached by DC Pelini) was notable for intensity and its ability to generate momentum...but I seem to recall the wheels coming off when the momentum turned the other way. He's an 'intensity' guy, and that alone will make NU's D better than 2007, but I just don't think the talent available to him is high enough to make a major surge in the rankings. Pelini's D's thrive off of individual play-making ability, and the Front Seven is completely devoid of proven play-makers (outside of possibly Ndamukong Suh). Maybe somebody steps up, but...maybe not.
So that's the defense. On to the offense.
How Much Can We Expect from Joe "Small Sample Size" Ganz in 2008?
Joe Ganz started three games in 2007. In those games, Nebraska scored 163 points. And gave up 172. The planet had obviously fallen off its axis in Nebraska last November, so what can be expected from a Ganz-led offense under more normal planetary alignment? First things first: let's compare SSS Ganz and JC Keller, and let's only look at what happened when game margins were within two possessions. No credit for putting up major numbers in a 30-point game.
Sam Keller's Offense (close games only)
Rushing: 42.7% success rate / 0.20 PPP / 0.627
Passing: 48.0% / 0.45 / 0.935
TOTAL: 45.6% / 0.31 / 0.770
Joe Ganz' Offense (close games only)
Rushing: 60.9% / 0.57 / 1.176
Passing: 46.2% / 0.59 / 1.048
TOTAL: 51.6% / 0.58 / 1.096
Now...Ganz' passing numbers were certainly better, but not a ton better, especially when you realize that nobody's going to average 0.59 PPP passing for an entire season.
(Actually I take that back--while neither Hawaii nor Texas Tech averaged more than 0.53 PPP, Florida did indeed hit the 0.59 PPP mark. So if you believe that Joe Ganz is as good as Tim Tebow, then Nebraska's got a chance to be the best offense in the country. And I have some land in New Zealand I'd like to sell you.)
So anyway, the passing game was better with Ganz than with Keller, but a lot of the improvement probably wasn't sustainable--not only would it be hard to replicate those numbers over an entire season, but Maurice Purify (24 catches, 402 yards, 6 TDs in Ganz' three starts) was the biggest reason for the surge, and he's exhausted his eligibility.
But that's not what sticks out most in those numbers--the running game was literally almost twice as good under Ganz as under Keller. I know Ganz is much more mobile than Keller (hell, my 16-pound cat is more mobile than Keller), but he had little to do with that improvement--Ganz had all of 66 rushing yards in his three starts. What was most responsible? Not Marlon Lucky. He averaged 62 rushing yards a game in that span. It seems the biggest single reason for those impressive rushing totals was that Nebraska was involved in shootouts and was passing two-thirds of the time (151 passes, 80 rushes in Ganz' three starts). So the rushing success may be a bit of a mirage too--at least, NU fans better hope it was a mirage, as the only way they'd be able to duplicate those rushing numbers is if they're playing from way behind two-thirds of the time again.
Summary
So what can we expect from Nebraska in 2008? The easy answer is, who the hell knows? They're the biggest unknown quantity in the conference outside of Manhattan, KS. But we won't settle for that easy answer.
Nebraska's offense averaged 0.27 more PPP with Joe Ganz behind center than with Sam Keller. That's not sustainable. But 0.10 may be within the realm of possibility. Over the course of 60-80 plays, that's a TD more per game. That may be a bit much, but we'll go with it.
On defense, NU just doesn't have enough talent in the front seven to improve too considerably. They have a series of DE's--Barry Turner, Zach Potter, and especially Pierre Allen--who have shown varying amounts of potential, but who haven't proven themselves to be every-down playmakers. Maybe a 'fire under the ass' coach like Pelini will be what they need...and maybe not. Ty Steinkuhler and Ndamukong Suh could be solid at DT, but Steinkuhler is limited and Suh completely disappeared after a decent start to '07. At LB...their leading playmaker is Tyler Wortman...who made all of 3.5 'successful' tackles in '07. Ouch.
(That's the other thing--even though their defense was wretched, players who made 70% of their 'successful' tackles departed after '07.)
Despite the lack of potential, let's say that the defense also improves by 0.10 PPP, which would be hefty improvement. That would mean the offense and defense are both 7 points better. Well...five of NU's seven losses were by 18 points or more. A 14-point improvement doesn't take you that far. That would, however, take them back to the Oklahoma State-level in regard to overall performance. That's a decent step in the right direction, and it--along with their massive homefield advantage--is why I said in yesterday's Roundtable that 8-4 is quite possible.