I've never been able to think that turnovers are a matter of luck.
2011 Preview Review: Defense
Vastly underrated; properly rated
Previously: The Offense
My look back at Brian's epic 2011 football preview continues with the defense. This one got a lot more interesting than the offense, because despite all the warm fuzzies we felt from the GERG-to-Greg transition*, expecting a jump from the #110 total defense to #17 would have been outrageous. As in get-this-man-a-straitjacket outrageous.
Thankfully, the performance of the defense exceeded all reasonable expectations, and even most of the unreasonable ones. Let's peep last year's predictions, shall we?
*Not to mention the Tony-Gibson-to-Anyone-But-Tony-Gibson transition.
The move to three-tech won't be an issue [for Ryan Van Bergen]. He played it two years ago and when Michigan went to a four man front last year they stuck him back inside. He's now 290, a three year starter, and a senior. He's a good bet to crack double-digit TFLs and get some All Big Ten mention.
RVB actually ended up at strongside DE, which probably helped him lead the team with 12.5 TFLs. He ended up earning All-Big Ten honorable mention from both the coaches and media and graduating as one of the most beloved Wolverines in recent memory.
Demens will benefit from the move to back to the 4-3 under more than anyone save Craig Roh. With RVB and Martin shielding him from linemen he won't be in nearly as many hopeless situations where he's one-on-one with a guard He should be the team's leading tackler by a healthy margin and see his TFLs skyrocket from the measly 1.5 he managed a year ago.
Michigan's defense will probably be too bad to warrant much All Big Ten consideration, but honorable mention seems reasonable.
A year after inexplicably having to move past not just Obi Ezeh, but converted fullback Mark Moundros, on the depth chart at middle linebacker despite subsequently making it painfully obvious that he should've been the starter all along, Demens had his breakout season. He led the team with 94 tackles—second was Jordan Kovacs at 75—and saw his TFLs jump to a respectable five. Like Van Bergen, Demens was an all-conference honorable mention.
Even so, [Kovacs's] season was a step forward from obvious liability to "certainly not a liability." Even if he's a walk-on and even if he's obviously small and slow, he should continue improving. He'll be a little less small and slow with another year of conditioning. Being in a coherent defensive system should help put him in positions to make plays. His redshirt year was not spent on the team so he's not as close to his ceiling as your average redshirt junior.
He's not going to be Reggie Nelson. That won't keep him from becoming the first Michigan safety you only hate a little tiny bit since Jamar Adams.
This may still be underselling Kovacs, who took to competent coaching even better than expected and became the team's rock in the secondary, covering for his athletic limitations with usually-impeccable positioning. No, he's not Reggie Nelson, but I don't think you can find a remotely rational Michigan fan who harbors even the tiniest bit of ill will towards Kovacs. Michigan's shocking lack of big plays allowed—both against the pass and the run—can largely be attributed to his play; despite missing a game, Kovacs led the team with 51 solo tackles. He also notched 8 TFLs. All hail Kovacs.
I have the same optimism about this Johnson/Gordon combo that I had last year. This, of course, terrifies me. It seems unnatural to think an unproven Michigan safety could be competent. I like Gordon's agility and tackling, though, and while there will be rough spots early by midseason he should settle into that midlevel safety range like Englemon or Barringer.
This time around, the optimism regarding the free safety position was justified. Thomas Gordon had his share of struggles, especially late in the season, but for the most part he was quite competent. Around here, safety competence is a luxury on par with consistent placekicking.
Sacks almost double from 1.4 per game to 2.4. That would be a move from 98th to around 30th.
Michigan finished with 2.3 sacks per game. That put them at... 29th. Tip o' the cap.
Turnovers forced go from 19 to 27.
Brian's continued insistence that turnover luck would someday go Michigan's way finally paid off; the Wolverines forced 29 turnovers. It also helped that this defense actually tackled people.
EVERYTHING SEEMS WONDERFUL
YOU HAVE NO IDEA HOW RIGHT THIS WOULD BE.
Morgan was the MGoBlog Sleeper of the Year based on a wide array of scouting reports that praise his instincts, lateral mobility, and toughnosed hard gritty gritness. I thought he'd have to cool his heels behind Demens for a couple years, but he may get on the field quicker than anyone expected.
No full credit simply because Mike Jones was projected as the starter at WLB, a fact I had completely forgotten about until I looked back at the preview. Morgan ended up playing in 12 games, starting seven (the first being in week two against ND), and finished fifth on the team in tackles.
If [J.T. Floyd] gets a lot better this year it's time to take the Gibson chatter seriously.
This wasn't really a prediction, but... yeah. Tony Gibson minus all of the points.
Beyond Talbott it's true freshmen, but at least there's a horde of them. Maryland's Blake Countess arrives with the most hype and should be the biggest threat to play. (Caveat: last year Cullen Christian arrived with the most hype.)
Points for mentioning Countess as the most likely freshman to see the field. No points for giving him one sentence when he took over the starting job by midseason, especially considering the Christian caveat. As you'll see, the hype that should've surrounded Countess went—justifiably, in the preseason—to Courtney Avery.
Not So Much
Healthy again and less abandoned in the middle of the defense, Martin's numbers should soar. Before the sprain Martin was on pace for 11 TFLs and 4 sacks; after it he got just a half TFL the rest of the year. While the front of the schedule is a bit easier, Martin had 8.5 TFLs and 51 tackles a year ago. Reasonable progression should have gotten him to 11. Add in further progression plus three DL coaches plus a bit more help on the line plus a free-roaming QB attack role and 15 to 18 TFLs plus a little more QB terror should be within reach. He should be All Big Ten. He might be better.
I hate that I have to put this prediction in this category, but here it is. While Martin was the best player on the defense, his numbers were hampered by having to play the nose; he finished with six TFLs and 3.5 sacks. Despite the lack of statistical production, Martin's efforts were recognized with second-team All-Big Ten honors. He also forced a pitch on a speed option. See you on Sundays, MM.
"Experience" was why [Will Heininger] got the nod; that experience consists of backing Brandon Graham up. In is time on the field he rarely did anything wrong; he rarely did anything right, either. He was a non-factor. As a guy spotting Graham from time to time that's cool, but as a starter or a guy rotating with another equally obscure walk-on that's a recipe for zero production out of a spot that should see its fair share of plays. If this spot averages out as a zero next year that's probably good—and that's not good.
The biggest swing-and-a-miss on the list. Heininger swapped spots with RVB and started all 12 regular-season games at five-tech DT before missing the Sugar Bowl with a foot injury. He exceeded all expectations of a walk-on raised in the shadow of the Big House, proving he could hold his own against Big Ten competition and be a positive force on the interior. After the season, Brian ranked him as the third most siginificant departure on the defense, behind only Martin and Van Bergen. While part of that is due to the remaining depth along the defensive line, I don't think anyone thought Heininger's absence would be felt in such a way.
Brink will play. After mentioning Heininger's experience he said Brink has "practiced very well, played well, been productive" and promised to rotate six guys on the line. Six is a weird number because it means one of Black, Campbell, or Brink is on the fringe. Given the lineups Campbell seems the most likely even though that seems unlikely.
If you're saying "who?" you're probably not alone (though you read this blog, so you probably aren't saying "who?"). Walk-on Nathan Brink was penciled in as the starting SDE at one point in the fall, earning much preseason praise for his unlikely rise up the depth chart. After garnering all that hype, however, he made almost no impact, recording just one tackle while barely seeing the field. He's a prime example of why you must take all offseason practice hype with a grain of salt, especially when said hype involves previously-unknown walk-ons.
We've yet to see the much of the pass-rushing skill that made Roh a top 50 recruit. He's displayed hints of the ability to zip past tackles before they know what hits them when suffered to rush the passer—there's a chance that when he puts hand to ground and is told to let it rip that he goes bonkers. Roh is the biggest X factor on the team. He could end up with anywhere from a half-dozen to twelve sacks.
Playing his third position in three seasons, Roh didn't quite go bonkers, tallying four sacks and eight TFLs. Roh's play still markedly improved from his previous two seasons, but he still hasn't lived up to the sky-high recruiting hype. Much of the blame for that can fall upon the shoulders of Greg Robinson and Co., and we'll see if one last position switch, this time to SDE, finally results in Roh producing double-digit sacks.
In high school, Ryan was an outside linebacker in an actual 3-3-5. As such, he spent a lot of time screaming at the quarterback from angles designed to make life hard for offensive linemen. That's not far off his job in the 4-3 under but it comes with a lot more run responsibility—the SLB has to take on blockers in just the right spot so that he neither lets the play escape contain nor gives him a lane inside too big to shut down. Expect to see him on passing downs but only passing downs this fall.
Ryan became a pleasant early-season surprise when he started against Western Michigan and made his presence felt by batting an Alex Carder pass that Brandon Herron would intercept and return 94 yards to the house. While certainly more of an asset against the pass than the run—his balls-to-the-wall approach was great on blitzes, but not always sound when keeping contain—Ryan proved that he was by far the best option on the strong side. Just one year later, all-conference honors are very much in play.
Assuming he's healthy, another year to learn the position and get bigger should see him improve on his previous form. There is a nonzero chance his earlier performances were not representative of his ability, but the smart money is on Woolfolk being at least average. It wouldn't be a surprise to see him go at the tail end of next year's NFL draft.
Troy Woolfolk's return from the exploding ankle of doom wasn't as triumphant as we all hoped. While he started ten games—six at corner and four at safety—Woolfolk never looked fully comfortable on the field and was supplanted at each position by a younger player (Countess at corner, Gordon at safety). It would be quite a surprise to see him taken in this week's NFL draft.
Courtney Avery busts out. Going into next year people are talking about him as an All Big Ten performer.
After showing much promise as a true freshman, Avery was the obvious candidate to grow into a big-time role as the team's top corner of the present and future. Instead, he started the first two games, then ceded that role to J.T. Floyd, Woolfolk, and eventually Countess. Avery was a solid nickel corner, and should reprise that role in 2012, but his progression wasn't as great as expected.
Craig Roh leads the team in sacks with eight.
Nein. Despite Michigan's impressive rise in team sacks, they were spread pretty evenly across both the D-line and the back seven thanks to Mattison's blitz-happy approach. Ryan Van Bergen paced the team with 5.5, with Jordan Kovacs actually tying Roh for second with four.
Michigan noses just above average in yardage allowed. Advanced metrics have them about 50th.
I know Brian has no complaints about being so hilariously wrong on this one. As noted above, the Wolverines finished 17th in yardage allowed, and they also shot up to sixth (faints) in points allowed. Football Outsiders's FEI metric ranked them as the #16 defense in the country. Despite watching every second of the 2011 season (usually twice), I still have a hard time not believing I'm the victim of an elaborate hoax or a drug experiment gone horribly awry. If you see me waking up in a gutter and GERG is still the defensive coordinator, please do me a favor and run me over with an SUV. Make sure to double-tap, please.
I don't think interceptinos tend to be luck-based(inexperienced QBs tend to throw more of them, more experienced CBs cause more of them), but I feel like there is a certain element of good fortune needed to recover fumbles on both sides that might cause one to state that "TOs are random." because half the number can vary somewhat wildly from year to year.
In the NFL, at least, I believe Football Outsiders has shown year-to-year interceptions to be somewhat correlated to a team's pass rush, which makes sense. They've also shown that forcing fumbles is not luck.
What does seem to be luck, however, is recovering fumbles, and that's where the "luck" comes in.
Fumbles aren't a matter of luck(unless you are ND), as defenses can be coached to force them. It's the recovery that is somewhat a matter of luck. I know Mattison is coaching the defense to swarm to the ball and they were quick to get after the ball when it was on the ground, but due to the shape of the ball and other factors, there is definitely luck involved in who recovers the fumble.
The way Ace phrased it was somewhat unclear. There seem to be three issues in this debate.
- In a given year, many turnovers are not "luck" in the sense we usually mean, like deus ex machina luck where, say, Tommy Rees throws a ball behind himself in the red zone. That shit never happens. Hmm... Well, it's more that most turnover opportunities are forced--good positioning on D, good pass rush, hats to the football, etc.
- As someone below said, fumble recoveries do have a stochastic component to them because the ball bounces so dang funny. Still, you can't recover 'em if you don't cause 'em, in which case see #1.
- The largest issue that Brian and others have harped on is the year-to-year correlation in turnover margin. It is low, for two reasons. One, a margin by definition has two components to it, which can work in opposite directions. Two, given that college players are variable and shuffle in and out due to graduation and matriculation, that means that if there are really good turnover-forcers on your team, they won't be around anymore in a couple of years (see: Woodson, C, hallowed be His name).
They are luck when our ratio is negative and effort/coaching when our ratio is positive.
dammit, who spilled the beans and told Ace about our drug experiment?
for competent coaching. Hip hip...
The improved defense was the most exciting part of a great year in UM sports for me.
The MBB team split big ten title was great! And getting BCS was great! But after watching our defense the past couple of years seeing the improvement this year was by far the best.
Having to watch games where you feel like if we don't get a stop if could mean trouble was rough... Thought the switch to competence with the defense would take much longer than it did- and that dramatic shift was great to watch! Here's hoping it can continue adn we can become a top defense nationally once more!
Yeah... last year helped me realize ho wmuch I took good defense for granted. I almost got bored with it.... (never again!)
Even more than taking good defense for granted it was suffering hoorible defense for me. I can never get bored of three and outs and teams averaging 2 yds/carry. I suppose the last couple of years prior to this last year will help me appreciate good defense more- but more than anything I want to never again loathe watching our defense play- and worrying every time we don't score the game may be a loss...
Will only put up 4 TFL and 1 sack. Morgan equaled those numbers and Hawthorne fell just one TFL short (though he had a pick as well). Brian basically said, "If he can be a PUSH, then that is a positive outcome." Heininger didn't make many plays but he didn't do anything to give up plays either. A nice push, but one that also allowed the defense to be as successful as it was.
The only reason he was listed as the third most important departure is that after Martin and RVB, he is the only starter who is leaving, and Woolfolk is the only other regular contributor who won't be back next year. Next year we should see a lot more production out of that spot, whether it be from Roh, Black, or one of the younger guys.
An amazing season.
The biggest differences I noticed about your D last year were definitely tackling and positioning. Fundamentals. Wrapping up opposing players and driving through them more. Being gap sound and knowing where to be when the offense lines up in a certain formation. I know all of this has been repeated many times, but imagine what RichRod could have done if he had just hired competent Defensive coaches.