Inside the Numbers: Michigan 51, UCF 14

We break down the numbers — and maybe a little bit of film! — that are encouraging, concerning, and fascinating from Michigan’s 51-14 victory over UCF this past weekend.

1992

Before Saturday, this was the year when Michigan last scored 50-plus points in back-to-back games. The Wolverines posted 61 points against Houston on September 26, 1992, and 52 points against Iowa the following Saturday. In 23.5 seasons thereafter, they scored at least 50 points 17 times but never did it in consecutive games. And not once in Jim Harbaugh’s first season as head coach in Ann Arbor did Michigan crack 50 points.

However, that is no longer the case because Michigan surpassed 50 points in each of its first two games in 2016, totaling 63 against Hawaii and 51 against UCF. As a result, Michigan is averaging 57 points per game, which is fifth in the nation. Unlike last season, there has been no adjustment period for this offense. Of their 21 drives that did not end with the Wolverines purposely running out the clock, they have scored on 15, with 12 resulting in touchdowns. That is quite the scoring rate, and it’s because they know Harbaugh’s offense, they understand their roles, and Wilton Speight looks very comfortable as Jake Rudock’s successor. The caveat is that neither Hawaii nor UCF presents a defensive challenge. But Michigan did what great offenses do and hung 50+.

Two times in a row.

37

This was the number of pass attempts by Wilton Speight against UCF -- a number that no one would have predicted prior to kickoff. In the opener, Speight tossed only 13 passes as Michigan kept the ball on the ground and ran through Hawaii. It was believed that Jim Harbaugh wanted to replicate this formula against UCF and continue to keep the pressure off of his new quarterback, if it wasn’t necessary. However, it became necessary because, for reasons that will be discussed in detail later, Michigan could not jump-start its rushing attack. So, in response, Michigan decided to air the ball out, and:

Speight shined.

In his second career start, Speight completed 25-of-37 passes (67.6%) for 312 yards (8.4 YPA), four touchdowns, and no interceptions, and was named the Big Ten Offensive Player of the Week. He was spectacular in the first half. He released 22 passes in the first 30 minutes, and, by my charting, he threw one more dime (5) than so-so ball (4). He dropped in a beauty to a diving Jehu Chesson on a post route for 35 yards on 3rd & 8. Staring down a third-down blitz, he took two steps to his right and hit Jake Butt on a drag route — in a spot where only Butt could catch it — for a touchdown. He found Amara Darboh in stride on a post route for a 45-yard score, as seen above. He fired his best ball of the game when, on 3rd & 12 on the UCF 14-yard line, he anticipated Butt’s corner route to the boundary and, before the UCF defender could break on it, rifled the football through a tiny, outside window into Butt’s hands for a touchdown. His only so-so first-half passes were when he just overthrew Chesson on a bomb, didn’t see an open Devin Asiasi in the end zone with a blitzer racing at him, lobbed a third-down screen too high and outside to Chris Evans that didn’t give Evans enough time to redirect upfield for a touchdown, and underthrew Grant Perry on a deep pass that should have been flagged for defensive pass interference. That was it. Everything else he threw was the right decision, and he displayed excellent accuracy within 20 yards.

Speight was not as sharp after halftime. There were two throws when he opted to aim for a Wolverine that was more covered than his other targets. He also had a few questionable tosses. However, at that point, he did not need to be as sharp. It was behind his arm that Michigan leapt out to a 31-0 lead in the first half and cruised.

What stood out about Speight’s performance was that he performed like this despite UCF’s defense sending waves of blitzes at him. Look at this from Pro Football Focus:

That is impressive, and, by my charting, Speight was pressured more than nine times (14). Michigan would like its interior linemen to make better adjustments pre-snap to prevent blitzing defenders from getting free runs at the quarterback, but Michigan must be happy with how Speight handled himself. He was cool, calm, and collected, standing strong in the pocket and firing with defenders bearing down on him. Not only must it give his teammates confidence in him that he can remain poised, Speight now knows, if he didn’t before, that he can win games when the Wolverines need him to.

8.6

By my count, this was the average number of defenders UCF had in the box for Michigan’s first 24 carries of the game. I stopped tracking thereafter because the Knights’ plan was obvious and additional evidence no longer was needed. They loaded the box repeatedly, brought their safeties within 10 yards of the line of scrimmage, and forced Michigan to beat them through the air. As a result, Michigan ran the ball 38 times for 142 yards and just 3.7 YPC after sacks and Kenny Allen’s bobbled punt snap are removed. This was not because the offensive line struggled to block up front. In fact, Mason Cole and the right side of the line (Kyle Kalis and Erik Magnuson) graded out pretty well. It was because Michigan had a very hard time finding running room as UCF had extra defenders and had its safeties crashing down into gaps where Michigan was headed. And Jim Harbaugh wasn’t going to reveal any fun wrinkles in this one.

Michigan responded in two different ways. The first was to suck in UCF’s defense even more with play action and let Wilton Speight beat them over the top. The second was to deploy more outside runs and try to beat the UCF defenders in the box to the edge. This had mixed success. Chris Evans had an 18-yard scamper to the outside, while Eddie McDoom took a jet sweep for 16 yards. However, there were more yards to be had. Whereas Michigan’s offensive line executed their blocks fairly well, the tight ends and receivers made more mistakes. For example, Jehu Chesson uncharacteristically missed some crackback blocks on sweeps and pin-and-pulls, which disrupted the timing of the runs. Also, some old bad habits for DeVeon Smith and Ty Isaac flared up. Yet again, they failed to see open cutback lanes when the blocks in front of them had been executed:

So not everything with Michigan’s running game is hunky-dory, but the issues are not nearly as severe as the numbers would lead one to believe. UCF loaded the box and never intended to let Michigan beat them on the ground. Even if Michigan had blocked many of these runs perfectly and the backs hit the correct gap, UCF still would have had a defender unaccounted for and nearby ready to cut down the run for little gain.

293

This was the number of rushing yards that UCF tallied once Michigan’s three sacks are removed, reaching that number on 43 carries for 6.8 YPC. On the surface, that seems horrid and extremely concerning given that the Wolverines will face more talented spread-to-run offenses down the road around, say, Thanksgiving. However, when I drilled into UCF’s rushing stats, I realized Michigan’s run defense fared much better.

Of UCF’s 293 non-sack rushing yards, 212 were recorded on just five snaps. That’s it. This means that, on UCF’s other 38 runs, they gained only 81 yards for 2.1 YPC. The Wolverines’ run defense actually was extremely good on a down-to-down basis. They shut down the running lanes and stuffed the Knights at the line of scrimmage often, which is why Michigan still ranks second in the country in S&P+’s Rushing Success Rate.

The problem is that Michigan surrendered five huge gains on the ground. However, again, this is not necessarily an indictment of the run defense. Three of those five gains were the result of quarterback scrambles, and all three were in clear passing situations when UCF did drop back to throw. So Michigan rushed the passer and dropped back into man coverage with its defensive backs trailing the UCF receivers and turning their back to the quarterback. Therefore, Michigan’s pass rushers needed to be disciplined because they had no support behind them if the UCF quarterback took off. Unfortunately, Michigan was too overeager to get to him on all three of the scrambles.

Here, Mike McCray shoots too far upfield, while Rashan Gary vacates his rush lane and pushes to the right where Ryan Glasgow and Chase Winovich are executing a stunt:

Here, Jabrill Peppers sprints too far vertically and does not attempt to constrict the pocket. Yet I believe Ben Gedeon was more at fault because, from his initial position, he could diagnose the play and see that a gap had opened to Peppers’ left. Instead, Gedeon slammed himself into the wall of bodies and was shoved aside to UCF’s benefit:

Here, ABC cut back to the play after the ball had been snapped, so we don’t see how we got here. However, based on Winovich’s position, it’s clear he was too aggressive:

However, unlike the first two scrambles, UCF’s quarterback slipped out to the left and headed toward the near sideline, and McCray should have shut this down quickly:

Instead, McCray runs away from the quarterback — maybe he doesn’t want to commit for fear that the quarterback will pitch it forward at the last sec — and gets sealed:

So Michigan’s run defense wasn’t the problem so much as it was the Wolverines’ quarterback containment, ceding 91 yards on three scrambles. They need to do a much better job being disciplined when they rush the passer. This is coachable, and I’m sure Don Brown will emphasize it in practice this week. If the ends want to utilize a speed rush, they need to power back towards the middle if they aren’t going to win the edge. Gary will learn to stay in his lane as he gets more reps. And maybe Brown will consider using a spy — Peppers? — on obvious passing downs against dual-threat quarterbacks.

As for the two long non-scramble UCF runs that went for touchdowns, those fell on Dymonte Thomas and Gedeon. Initially, I thought McCray was held on UCF’s 87-yard touchdown sprint, but, upon review, it’s 50/50 at best. McCray needs to do a better job fighting off that block to force the ball carrier back inside sooner. Either way though, the run should not have been more than 15 to 20 yards. Thomas clearly was not aware of Adrian Killins’ track speed and crashed down too far inside, permitting Killins to race around him up the sideline for a score. Thomas is not Jarrod Wilson and will be prone to giving up big plays from time to time, so I guess it’s better for Michigan that he did it when they held a 31-point lead. As for UCF’s 34-yard touchdown run, a gap opened right in the center of Michigan’s defensive line, but, for some reason, Gedeon vacated the middle and shifted to the left side of the line after the snap. UCF’s back took the handoff and burst right through the heart of Michigan’s defense for the touchdown. Gedeon has had a tendency to react too quickly upon the snap and over-commit to a gap or position, leaving Michigan’s defense vulnerable in spots, and it backfired there.

So there are things upon which Michigan’s defense needs to improve, such as containing dual-threat quarterbacks and preventing the home run on the ground. But, for the most part, Michigan’s run defense actually was pretty stingy against UCF.

27.3

This was UCF’s completion percentage versus Michigan. The Knights’ two quarterbacks, Justin Holman and Nick Patti, combined to go 6-for-22 for 56 yards (2.5 YPA), no touchdowns, and no picks. That equates to an abysmal quarterback rating of 48.65. There were maybe two incomplete passes that should have been gains for the Knights. The first was Mike McCray’s pass deflection in the first quarter. If Holman waits a beat longer, the receiver breaks open behind McCray for a nice pickup. The second was a slant over the middle that exploited a Jabrill Peppers blitz. The slant was under Delano Hill, and, if it had been completed, it likely would have gone for a touchdown. However, it was Patti’s first throw after entering the game, and he airmailed it. Otherwise, the Wolverines’ pass defense was impeccable. UCF’s receivers could not get separation whatsoever. The Knights had only one true chance to connect on a bomb, and Jeremy Clark sprinted step for step with the receiver and raked out a perfectly thrown ball at the last second. Accordingly, Michigan is ranked sixth in the nation in quarterback rating allowed (69.75). And they have done it without their All-American cornerback.

4

This was the unofficial number of kicks that Michigan blocked against UCF. Michigan found a weakness in the Knights’ punt formation in the scouting report because Tyree Kinnel flew in from the left side to get a piece of their first two punts. However, these were not registered as official blocks in the stat sheet. Then, without getting too much push, Chris Wormley stood tall and blocked UCF’s low 49- and 50-yard field goal tries:

Essentially, Michigan’s special teams unit transformed into Dikembe Mutombo:

It was the first time that Michigan blocked multiple kicks in a single game since Terrance Taylor knocked down an extra point and Mark Moundros smothered a punt against Utah in the 2008 opener. But, more importantly, these blocks kept points off the scoreboard for UCF and handed Michigan’s offense excellent field position — Michigan began its drives at the UCF 49-yard line on average in the first half. There had been questions whether Michigan’s special teams would experience a dip with John Baxter’s departure. If this was any indication, there is no need to ask such questions.

84.0

This is Kenny Allen’s current career field-goal percentage, which is the highest in Michigan program history. On Saturday, Allen booted all three of his field goals (24, 36, 37) through the uprights to improve to 21-of-25 in his career. Allen doesn’t have the biggest leg, but there are not many other kickers that are more reliable than him from 40 yards or closer. From that distance, Allen is 20-of-21 (95.2%), and the Wolverines have a peace of mind knowing that field goals from that range are automatic with him.