Notre Dame vs. Michigan: Can the Wolverine offense be stopped?

My friend Dan spent some time looking at Saturday's evisceration of the Connecticut defense by Denard Robinson and his fast friends.  When I was watching the highlights it seemed like UConn was confused a lot of the time, but I wasn't sure how much of that confusion was a poor scheme and how much was the fact players as fast as Robinson make most defenses look confused.  Dan's words are after the embedded video.

After looking at how Michigan attacked the Connecticut offense, we take a look at the Connecticut defense and what they did, or attempted to do, to stop a talented Michigan offense (29.5 points per game in 2009, 39^th nationally) that returned 10 starters from last year. The play by play breakdown can be found here and the video of almost every offensive snap is above. Rather than providing drive summaries, I have opted to provide a summary analysis of Michigan's offensive tendencies along with the successes and failures of the Connecticut defense.

Like Connecticut, Michigan spent the majority of the day in shotgun. In fact, excluding the final drive, I counted less than 5 plays that started under center. Michigan had 5 drives in the first half (7.8 plays per drive) and 3 in the second half (15 plays per drive). The drive lengths in the second half were not necessarily due to a change in playcalling or an attempt to run out the clock. Instead, the Connecticut defense was doing a much better job of executing their defense and forcing Michigan into mostly short gains.

Michigan ran an extremely vanilla offense. They only ran out about 9 different formations in the entire game. After Michigan's second series, the plays became very predictable based on the formation, and the Connecticut defense adjusted accordingly. Some common formations and plays include:

• 2 WR on one side and 1 WR on the other, triple option with the QB able to read the DE and make a pass to the single side WR (this was their most common play and formation)
• 4 wide, almost always a pass, if not then a QB lead draw
• Trips to one side, usually a WR screen pass, if not then a QB read to the trips side
• 2 WRs to one side with an off the line TE (this was Michigan's most effective formation), if TE pulled it was a triple option with the TE kicking out the LB, if he stayed, then it was a dropback pass or QB draw*
• Double tight ends, is either a QB lead dive or a straight hand off to the RB (this was the most common 3^rd and short formation)

*In the second half especially, I was able to key on the TE to predict the play out of that formation. It was correct every time. I am going to look out for some twists out of that formation this weekend.

The most common play Michigan ran was their version of the triple option. It included the classic QB read of the unblocked end, but also provided a throwing option to the WR if the QB kept it.  [Ed. note: Urban Meyer ran something similar to this with Tebow, where he had the pitch to the running back or a quick shovel pass to the tight end, generally Aaron Hernandez.]

Michigan's two best plays in this game came out of the trips formation and the off the line TE formation. In the trips, Michigan faked the WR screen they had been throwing all game and hit a deep post for nearly 50 yards. On the snap, the safeties sprinted towards the line and took themselves completely out of the play (a common theme for the safeties on the day). In the offset TE formation, the TE pulled and trap blocked the DE. This created a whole twice the size of Chris Stewart that Denard Robinson was able to leisurely jog through for 30 yards. This was the first time Connecticut saw this formation in the game and had much more success against it on future Michigan attempts.

Michigan did not throw the ball much in the first half, and they didn't need to. They were able to march down the field with their basic offensive run plays without much effort. In the first half they threw the ball 4-5 times with all but one (an incompletion) going to what appeared to be the first read, often in a screen situation. In the second half, Connecticut did a much better job of stopping the run and Michigan was forced into many third and long situations that they executed with short crossing routes right at the first down line. This can mainly be seen in the first offensive drive of the second half that lasted 19 plays.

One of the things that surprised me was the lack of push generated by the Michigan offensive line. That is not to say that they didn't do a great job of blocking the entire game, but their technique was less physically dominating than I expected. Instead, the line seemed to take advantage of its large splits and the lack of blitzing by the Connecticut defense. The line often stood up the defensive man lined up directly over him and then uncovered lineman worked to the second level to pick up a linebacker. This led to numerous cutbacks and some pretty easy reads for Robinson in the backfield. Rarely did I see a Connecticut lineman or linebacker blown off the play. Instead, it looked more like Robinson was juking around blocking dummies. Most of the tackles were linebackers or safeties coming from the sides.

Overall, I think Connecticut had a strong gameplan that they simply failed to execute. They rarely blitzed and stuck to their same base offense for nearly every play (goal line was the rare exception). Their base defense was a 4-3-4. They kept the safeties at about 10 yards off the line. They did move the OLBs around a bit to cover WRs, which is one of a few concerns I have about the way that they attacked this spread offense.

First, it did not seem like they treated the QB reads as option plays. Rarely did it seem that there was a man assigned to attack the QB. Rather, the DE crashed down on the RB and the QB was free to bounce it to the outside. At that point, he became the responsibility of the CB on that side or the outside LB. However, the LB almost always got sucked into the line and was out of position for the play. The CBs were often blocked, or at least held up, by that side's WR. This play then had to be made by one of the safeties who were consistently starting 10 yards off the line of scrimmage.* That meant that Robinson was easily getting 5+ yards when he kept the ball. Sadly, I think Robinson missed at least 4 reads that would have gained big yardage. ND is going to have to have a guy assigned to the QB on these plays.

*I think this is one of the biggest areas that ND can learn from this week and one that they almost naturally will due to their 3-4 formation. With the OLBs being outside or at the offensive tackles, it will be difficult for them to be sucked into the play or get blocked. Additionally, ND could consider rolling up the safeties to about 7 yards and help contain the QB on the outside.

Second, the safeties were way too far off the line. It seemed like Connecticut went into the game with a mentality to stop the pass. In general, the defensive line was getting decent pressure. However, most of the run plays were QB leads or options. This almost always took the playside LB out of the play and there was rarely a safety anywhere near the line to make the play.

Third, Connecticut over-pursued and failed to keep their lanes. The entire team seemed to be crashing hard on ball fakes to the RB. Surprisingly, this turned out to be fairly effective. On numerous occasions the play was stuffed in the backfield. However, the DL and backside LB failed to keep backside containment. The cutback lanes were wide open for Robinson and the running backs all day. Many of their big plays in the first half came on cutbacks where the LB over-pursued.

Fourth, the linebacker continued to push out. Michigan often brought out a 3 WR formation. Rather than rolling up a safety to cover this man, Connecticut pushed out a LB, leaving only 2 LBs in the center of the field. Michigan then ran away from the 2 WR side. This created wide open cutback lanes (since the LB was out of the play already) and enough blockers to block the 2 remaining LBs. Again, if ND rolls a safety up to cover this third WR (or brings in a nickel back), then I think it will provide more opportunity to keep our LBs in position to make plays.

Fifth, Connecticut never showed a different look. As predictable as the Michigan playcalling was, I knew the Connecticut defensive formation for every formation that Michigan presented. This has to be a great advantage for the Michigan playcaller. I would like see Notre Dame show some different formations and different personnel groupings. Even if the formation is not entirely effective, it will force the offense to think and could cause an extra second of delay.

The second half was a completely different game for the Connecticut defense. I wouldn't be surprised if a coach came down from the booth and told the defense, "If they come out in X formation then they are running Y play."  In the second half, the defense acted like it knew the play every time it was a rush, especially on the first drive. But Connecticut couldn't stop Michigan on third down (14-19 on the day). Many of those third downs came on that 19-play drive to start the second half. I can't say that the Connecticut defense did anything terribly wrong. The Michigan wide receivers were running 5-10 in routes and crossing patterns that the cornerbacks were just not keeping up with. The linebackers were often bailing on play action, following the QB on a rollout, or taking a TE or RB into coverage. These pass plays are the biggest concern of mine going into the game and ND is going to need to figure out a way to stop them without giving up their option defense.

Looking towards Notre Dame's attempts at stopping this offense, I think we are going to see a lot of different formations from Michigan to adjust to the 3-4 look the Irish will throw out there. Hopefully Michigan's answer to this is for their offensive line to double team and chip off to the second level. I don't think that the Michigan line has the power to push the ND line around. It is going to be more important than ever for the ND defensive line to hold their position and allow the LBs to make plays. Hopefully this will lead to some double teams in the front and prevent blockers from getting to the second level.

Conditioning is going to be important. On two of the long Michigan drives in particular, the Connecticut defense was gassed by the time they reached the red zone. Michigan was able to run the ball right up the middle with little resistance. Notre Dame needs to continue their strong rotation of defensive players, but it will be challenging with the fast tempo of the Michigan offense.

I advocate getting the safeties more involved in the action at the line of scrimmage. Although this compromises the pass defense, I was not impressed with any of the pass plays Michigan ran over the top against Connecticut. In fact, the only play that took advantage of the safeties in pass coverage was the fake screen for 50 yards. If ND is comfortable with putting their CBs on an island in pass coverage (I am not sure that I am), then I think bringing up at least one safety will help.

This offense has some scary elements. They have great athletes carrying the ball and an offensive line that can stand up defensive players with ease. The key for Notre Dame will be finding a way to match up with each player in the option (RB/QB/WR) and also containing the passing offense (especially in the middle of the field). I think it will be tough to keep Michigan under 28 points, so the offense is going to need to show up.