As I mentioned in my Wisconsin recap, Nick Toon ran free through the middle of Michigan's secondary on Saturday and was largely untouched before getting 20- and 30-yard gains. Michigan never changed their defensive scheme even though Wisconsin had clearly figured it out, and it finally hit me: Michigan has been vulnerable in the secondary because they're unable to play anything that's not a cover-2--or zero blitz, when they bring the house--due to the lack of talent at safety and depth elsewhere. Take a look at the pre-snap alignments from a few of Michigan's big disaster passing plays:
The first one is from MOEAKI!!! and the birth of our inability to stop tight ends, and the second is Penn State and one of Andrew Quarless' scoring plays. Against Quarless, Michigan is sitting in a base cover-2 set, and the opposing tight end split the middle of the field. I think a lot of this falls on the linebackers who need to drop back into coverage and make sure that the pass to the tight end doesn't get there, but we all know how our linebackers play in coverage. Against MOEAKI, Michigan played in this really weird quasi cover-2, quasi man, quasi cover-1... I don't now. It's definitely one of the formations we had heard about preseason. As you know, it was a disaster. So as Michigan experiemented with different formations to try and get everything right, the need to go to a cover-2 arose as Michigan continually gave up huge, crushing plays.
Wisconsin saw Michigan's trend to play cover-2 and started to exploit it. (With MGoVideo down and, apparently, no where else to find the game, I can't really bring you any images. If I find some video in the near future I'll post some pictures.) You may have noticed that in the second half, Wisconsin started a lot of plays by showing a two-wide, max protect set. Before the snap, they would motion Nick Toon from one side of the field to the other to, presumably, see how Michigan's defense would react. And what did they do? The corner that was initially on Toon would slide over to the end of the LOS on the side of the field he was on--presumably, he was then in charge of covering the weakside flat, but that move completely tipped Michigan's defensive alignment. This was a glaring signal to Wisconsin that Michigan was playing zone. Tolzien only had to do a pre-snap read on the safeties to see if they were playing a cover-2 (they were), and Toon simply ran skinny posts through the middle of the zone. This play happened exactly this way at least 4 times during the Wisconsin game.
As I mentioned before, this is the biggest condemnation of Greg Robinson this season. He's been working with a relative lack of talent and the poor performances are at least moderately acceptable given the depth and physical limitations of his players. But when you're outwardly outschemed and unable to react to it, that's a problem. I've been fairly forgiving of Robinson this season because of the personnel trouble. This, however, was ugly.
But maybe this scheming (or lack thereof) can't all be blamed on Robinson. For starters, it's not as if Michigan had been succeeding with the cover-2 they'd played all year. To give the secondary a completely new look would probably be disastrous (think 3-3-5 stack against Purdue last year). But what needs to break first: Being bad at playing cover-2 and changing to something else, or trying to get good at playing cover-2? At this point, you almost have to say the former.
But that's where the problems really begin. With Michigan's lack of any depth in the secondary, it's difficult to play anything but cover-2. And the lack of talent that the current set of safeties has makes playing cover-1 a worse gamble than they're current formation. The only reasonable compromise would be to play Williams, Kovacs, and Brandon Smith at the same time in a cover-3 and sacrifice one of Michigan's linebackers--but given their trouble in run stopping with the personnel they have currently on the field, one less linebacker would probably spell doom in the run game.
If all of this sounds like I'm talking in circles, well, I am. The fact of the matter is, the lack of any respectable depth or talent in the Michigan secondary makes it exceedingly difficult to counteract opponents attacks. I'm not implying that Robinson shouldn't try. He has, to relatively poor results. In the end though, the inability of the linebackers to stop the run, the lack of depth at safety, and the fear of the big play if Michigan strands Kovacs, Williams, or Smith in the secondary alone, all contribute to a defense that's behind the 8 ball. This isn't supposed to come off as an excuse for this defense. A plea for understanding and an explanation as to why Michigan is unable to stop, well, anyone.
Showing posts with label mike williams. Show all posts
Showing posts with label mike williams. Show all posts
Wednesday, November 18, 2009
Tuesday, October 27, 2009
The New Defense (Part 2)
Ed. Read The New Defense (Part 1) before reading this.
This isn't much different than the play I broke down in The New Defense (Part 1), in fact, it's the exact same defense and the exact same offensive play call. But the result is a little different. And that's encouraging.
Here's the same alignment by Penn State, only they flipped the field. The tight end is now on the bottom of the screen. Kovacs is shading to the bottom and Woolfolk has man coverage on his wide receiver. Mike Williams is the outside linebacker on the top of the play and Donovan Warren is the corner playing 10 yards off the line. What to watch here is that Williams reacts faster and there's a better result. If Michigan can continue to improve, this defense might actually work.
Here's the play right after the snap. You can see that Williams is still watching Clark, like he did the first time but is already shading toward the receiver who is sitting down and waiting for the screen pass. The next shot is particularly telling.
The ball is in the air here (you can see the spec just above the 0 of the 40 yard line). Most importantly, though, Williams is already starting to run at the receiver. Warren does exactly as well as he did on the last play and is crashing down on the pass.
Williams is still a step or two late on this pass, but you can see that he's close and able to make the tackle near the line of scrimmage. Warren looks to be playing cautiously so that if Williams misses the tackle, he hasn't overrun the play.
Three yards past the line of scrimmage, Williams has tripped up the wide receiver who rolls for another yard. Warren has successfully kept contain on the play.
You can see on film at the end of The New Defense (Part 1) that Greg Robinson grabs Williams after the play and gives him a very specific directive. I assume that this is exactly what he told him. Ezeh and Mouton should be able to give chase if it ends up as a run play, and Williams crashing on that short zone should limit or take away these short passes. Again, empirical improvement in game. This is encouraging.
The one problem with this defense, as you might guess, is that it takes arguably our best defensive player (Warren) and mostly neutralizes him. I think the positives and negatives balance out in the end, though. This alignment makes it difficult for Michigan to get beaten on the long ball while, if everyone plays their respective zones correctly, also takes away the short passes. The 4-4 alignment also helps with run defense and puts a lot of players in the box--there's actually a really good breakdown of a similar defensive play call over at Trojan Football Analysis. There are obviously flaws here and Michigan is going to need to work on a few things, but this is mostly encouraging.
The final part of the series will be coming around Noon.
This isn't much different than the play I broke down in The New Defense (Part 1), in fact, it's the exact same defense and the exact same offensive play call. But the result is a little different. And that's encouraging.
Here's the same alignment by Penn State, only they flipped the field. The tight end is now on the bottom of the screen. Kovacs is shading to the bottom and Woolfolk has man coverage on his wide receiver. Mike Williams is the outside linebacker on the top of the play and Donovan Warren is the corner playing 10 yards off the line. What to watch here is that Williams reacts faster and there's a better result. If Michigan can continue to improve, this defense might actually work.
Here's the play right after the snap. You can see that Williams is still watching Clark, like he did the first time but is already shading toward the receiver who is sitting down and waiting for the screen pass. The next shot is particularly telling.
The ball is in the air here (you can see the spec just above the 0 of the 40 yard line). Most importantly, though, Williams is already starting to run at the receiver. Warren does exactly as well as he did on the last play and is crashing down on the pass.
Williams is still a step or two late on this pass, but you can see that he's close and able to make the tackle near the line of scrimmage. Warren looks to be playing cautiously so that if Williams misses the tackle, he hasn't overrun the play.
Three yards past the line of scrimmage, Williams has tripped up the wide receiver who rolls for another yard. Warren has successfully kept contain on the play.
You can see on film at the end of The New Defense (Part 1) that Greg Robinson grabs Williams after the play and gives him a very specific directive. I assume that this is exactly what he told him. Ezeh and Mouton should be able to give chase if it ends up as a run play, and Williams crashing on that short zone should limit or take away these short passes. Again, empirical improvement in game. This is encouraging.
The one problem with this defense, as you might guess, is that it takes arguably our best defensive player (Warren) and mostly neutralizes him. I think the positives and negatives balance out in the end, though. This alignment makes it difficult for Michigan to get beaten on the long ball while, if everyone plays their respective zones correctly, also takes away the short passes. The 4-4 alignment also helps with run defense and puts a lot of players in the box--there's actually a really good breakdown of a similar defensive play call over at Trojan Football Analysis. There are obviously flaws here and Michigan is going to need to work on a few things, but this is mostly encouraging.
The final part of the series will be coming around Noon.
The New Defense (Part 1)
Watching the Penn State game, you may have been thinking to yourself, "Why in Christ's good name is Donovan Warren playing with a 10-yard cushion on the line?" You probably thought, "I'm sure there's a reason for this. I trust my coaches. This seems reasonable." And then three bubble screens later and blown coverages galore, you were searching in your napsack for the lighter fluid to spark your torch.
You may remember last year's Purdue game in which Rich Rodriguez basically got fed up with Scott Schafer's defensive schemes and wanted to see something he used to run in West Virginia, the 3-3-5 stack. Against a nobody quarterback in his first collegiate start, Michigan struggled with their assignments and got absolutely torched on passing plays. That was the public birth of the rift between Rodriguez and Shafer, and what eventually led to his firing and the introduction of Greg Robinson and his wacky schemes.
This Penn State game was almost exactly the same, only this time, Michigan decided to play essentially a 4-4 zone on every down--it's kind of a small shift from the basic 4-3 under Michigan is supposedly running, but I think there's a little more it than that. Let's take a look at the presnap alignment and then follow it throughout the game:
This is the basic presnap alignment. Troy Woolfolk is lined up in man coverage on the top of the screen. Jordan Kovacs will give him help over the top and is lined up as the safety in the middle of the field. Donovan Warren is the deep corner at the bottom of the screen. What's interesting about this formation is that Michigan left the same personnel on the field that they usually run with their 4-3 under. The defensive line is the same. Mouton and Ezeh are the middle linebackers. The changes come with Mike Williams, who is lined up as the outside linebacker on the bottom of the screen, and Stevie Brown, who is the outside linebacker on the top of the screen.
This is sort of a half zone, with Warren's side playing zone. But it can be played either way. The way Michigan is defending this play is with Williams covering the underneath route and the bubble screen, and Warren covering the deep route. On the other side, Woolfolk has man coverage with help over the top from Kovacs. I think the reason there is this man coverage aspect to the play is because of the tight end on the strongside of the play. Stevie Brown is supposed to cover him when/if he releases into the secondary--we'll see how Penn State counteracted this later in the game and ended up with a touchdown because Obi Ezeh is pulled into coverage on a seam route.
Here's the snap. Warren stays in his deep zone and Williams is spying Daryll Clark (incorrectly so). Williams here is supposed to hover over toward the wide receiver on the bottom of the screen to cut off the bubble screen that will eventually be thrown. It's sort of a robber play. If Williams plays and times this perfectly, it's a pick 6. He doesn't.
Clark takes the presnap read when Williams is too slow to get over to his zone. You can see that Clark is winding up to throw and Williams has barely started to move. Warren is sitting back in his deep zone and already sees the screen being thrown. He's breaking toward the play.
The receiver has the ball now and a clear 5 yards to run. Even if Williams takes a perfect angle and makes a perfect tackle, he still gives up four yards.
Williams and Warren are now closing in but have already given up at least four yards. Williams will miss on the tackle and the wide receiver will end up getting seven yards:
So, this is how Penn State brutalized us on Saturday: A completely new defensive scheme that our team had no idea how to run.
The thing is, as poor as the results were, I am actually a little encourage by this. As the game went on, the team got better at running it. And moreso, this isn't the same Shafer/Rodriguez dynamic that eventually ruined the defense last year. I honestly believe that this is a concession by Robinson to the talent he has to work with. What the team had been trying was clearly not working. This is a step in the right direction. And I think with more time with this system, they can get better. Plus, this is only a small shift from the 4-3 under that is Michigan's base defense. Instead of having two deep safeties, two linebackers, and five men on the line, Michigan is putting six men on the line, by bringing the free safety down to cover the short zone. Robinson trusts Warren enough not to give him help over the top, and Kovacs will probably cover the strong side of the play going forward.
Essentially what it does is take advantage of the athletes Michigan has without putting them in one-on-one coverage. It also allows Michigan to send a variety of looks. Williams and Brown are good enough athletes to play a hybrid linebacker role, and both were/are safeties and should, presumably, be good in coverage. Once they understand what they're supposed to do, the only weakness is seeing whether or not Kovacs makes the right plays to cover over the top--he didn't in this game, which I'll talk about tomorrow, but hopefully he can grow.
Take a look at how this progressed through the game: Part 2; Part 3
You may remember last year's Purdue game in which Rich Rodriguez basically got fed up with Scott Schafer's defensive schemes and wanted to see something he used to run in West Virginia, the 3-3-5 stack. Against a nobody quarterback in his first collegiate start, Michigan struggled with their assignments and got absolutely torched on passing plays. That was the public birth of the rift between Rodriguez and Shafer, and what eventually led to his firing and the introduction of Greg Robinson and his wacky schemes.
This Penn State game was almost exactly the same, only this time, Michigan decided to play essentially a 4-4 zone on every down--it's kind of a small shift from the basic 4-3 under Michigan is supposedly running, but I think there's a little more it than that. Let's take a look at the presnap alignment and then follow it throughout the game:
This is the basic presnap alignment. Troy Woolfolk is lined up in man coverage on the top of the screen. Jordan Kovacs will give him help over the top and is lined up as the safety in the middle of the field. Donovan Warren is the deep corner at the bottom of the screen. What's interesting about this formation is that Michigan left the same personnel on the field that they usually run with their 4-3 under. The defensive line is the same. Mouton and Ezeh are the middle linebackers. The changes come with Mike Williams, who is lined up as the outside linebacker on the bottom of the screen, and Stevie Brown, who is the outside linebacker on the top of the screen.
This is sort of a half zone, with Warren's side playing zone. But it can be played either way. The way Michigan is defending this play is with Williams covering the underneath route and the bubble screen, and Warren covering the deep route. On the other side, Woolfolk has man coverage with help over the top from Kovacs. I think the reason there is this man coverage aspect to the play is because of the tight end on the strongside of the play. Stevie Brown is supposed to cover him when/if he releases into the secondary--we'll see how Penn State counteracted this later in the game and ended up with a touchdown because Obi Ezeh is pulled into coverage on a seam route.
Here's the snap. Warren stays in his deep zone and Williams is spying Daryll Clark (incorrectly so). Williams here is supposed to hover over toward the wide receiver on the bottom of the screen to cut off the bubble screen that will eventually be thrown. It's sort of a robber play. If Williams plays and times this perfectly, it's a pick 6. He doesn't.
Clark takes the presnap read when Williams is too slow to get over to his zone. You can see that Clark is winding up to throw and Williams has barely started to move. Warren is sitting back in his deep zone and already sees the screen being thrown. He's breaking toward the play.
The receiver has the ball now and a clear 5 yards to run. Even if Williams takes a perfect angle and makes a perfect tackle, he still gives up four yards.
Williams and Warren are now closing in but have already given up at least four yards. Williams will miss on the tackle and the wide receiver will end up getting seven yards:
So, this is how Penn State brutalized us on Saturday: A completely new defensive scheme that our team had no idea how to run.
The thing is, as poor as the results were, I am actually a little encourage by this. As the game went on, the team got better at running it. And moreso, this isn't the same Shafer/Rodriguez dynamic that eventually ruined the defense last year. I honestly believe that this is a concession by Robinson to the talent he has to work with. What the team had been trying was clearly not working. This is a step in the right direction. And I think with more time with this system, they can get better. Plus, this is only a small shift from the 4-3 under that is Michigan's base defense. Instead of having two deep safeties, two linebackers, and five men on the line, Michigan is putting six men on the line, by bringing the free safety down to cover the short zone. Robinson trusts Warren enough not to give him help over the top, and Kovacs will probably cover the strong side of the play going forward.
Essentially what it does is take advantage of the athletes Michigan has without putting them in one-on-one coverage. It also allows Michigan to send a variety of looks. Williams and Brown are good enough athletes to play a hybrid linebacker role, and both were/are safeties and should, presumably, be good in coverage. Once they understand what they're supposed to do, the only weakness is seeing whether or not Kovacs makes the right plays to cover over the top--he didn't in this game, which I'll talk about tomorrow, but hopefully he can grow.
Take a look at how this progressed through the game: Part 2; Part 3
Monday, October 26, 2009
Penn State: Where the similarities are uncany
When the Michigan men's basketball team had an improbable run against Top 5 teams last season, beating both UCLA and Duke early in the year, the team went from promising upstart to potential Big Ten contender, rightly so or not. As the season waned, so too did the team's momentum, and their youth and flaws began to show. Close matches between Savannah State and Indiana, coupled with back-to-back-to-back losses to conference foes (Illinois, Ohio State, and Penn State) raised alarms for fans and made an NCAA berth go from a certainty to a realistic, best-case scenario.
So too seems to be the arc of the football team. After a nonconference win against then #18 Notre Dame, the preseason projections of 7-5 suddenly looked more like 8-4 (or 9-3 at best). But in the midst of the Big Ten slate, Michigan has stumbled, falling to Michigan State and Iowa (which included a benching of the team star, Forcier, much like the similar, controversial benching of Manny Harris against Iowa in an overtime game), a close win against lowly Indiana, and now a crushing defeat to Penn State; 8-4 is a pipe dream now, and the team is now in line to play the worst two Big Ten teams in what feel like must wins.
What can we can learn from this Penn State game is a little more concrete: David Molk really is as valuable as we all thought; the shine is beginning to come off of freshman savior Tate Forcier; the turnover problems from 2008 are far from resolved; and the defense is worse than we all thought.
I'll get into this later in the week, but the secondary is in far more dire a situation than was widely presumed at the beginning of the season. Stevie Brown and Mike Williams are mostly incapable of covering anyone in space. Jordan Kovacs, who had been anointed as Michigan's Rudy two weeks ago, has been exposed time and again, allowing a number of touchdown passes that have crushed the team. The defense appears incapable of playing man coverage, while the defensive line is flailing without much success. Simply put, this team is in a lot of trouble as the offense continues to have difficultly scoring behind an injured Forcier, fumbly running backs, and receivers whose recent case of the dropsies is nothing short of a massive disappointment after their respective performance early in the season.
On the other side of the field, it would appear that I greatly underestimated how good Penn State is. Not that any team that beats Michigan so resoundingly is suddenly much better than we thought, but Penn State is clearly a legit contender in the Big Ten and will still finish second or third in-conference; their stats, though against mostly cupcakes, are no joke. Evan Royster trounced what was otherwise a particularly stout defensive front, putting up 5 YPC on 20 carries. Meanwhile, Michigan's rushing attack looked exceedingly mediocre. Basically, everything that could've gone wrong did.
In other words, Michigan is reeling. I don't know what they need to do to get better. The secondary doesn't look like it has the talent to perform against Big Ten offenses. The offense has gone from fast and high-powered to a sputtering collection of players that look like they are trying to win by themselves (see also: Michigan 2008). Meanwhile, I've got questions about the defensive play calling and the offensive personnel that Rodriguez has been playing. When once we hoped that the Illinois/Purdue stretch would be gravy, it now becomes a crucial stretch of the season. Michigan is in trouble and they don't look like they have any answers.














