Analytics to Prove Purdue Was Beaten Like a Big Drum
Notre Dame secured by far their most stress free win of the season Saturday, handing Purdue its first loss to secure a 3-0 record. The Irish defense was excellent and is continuing to come into its own after a rocky start in Tallahassee. The offense was good enough to get the win, but may need to improve in some areas to compete against tougher competition. For an explanation of the statistics we will use in this analysis, please follow this link to our analytics primer.
Converting a 4th and 3 into a touchdown resulted in 24% Win Probability Added, placing the Irish back in the driver’s seat after trailing early in the game. Avery Davis’ long touchdown catch was also crucial, and while Purdue was able to claw back a little bit after that it felt like the Irish were firmly in control of the fourth quarter.
This was the worst week of the season so far for the passing game by a pretty wide margin. While Coan’s -0.04 EPA/Play is not awful, the 29% success rate is. This difference is driven by the fact that the long touchdowns Kyren Williams and Avery Davis are outliers and pulling Coan’s average EPA up. Coan struggled a bit with accuracy and Irish receivers dropped five passes, with Braden Lenzy dropping two including a long touchdown and Kevin Austin dropped two balls as well en route to the worst PFF Grade for any Irish player. It was just a bit of an out of sorts week in the passing game, where once Coan had shaken off his poor start his receivers began to struggle.
The other variable in Coan’s inaccuracy was how often he was under duress. Coan was under pressure on 41.7% of his drop backs, but only two of the pressures were assigned to him holding onto the ball too long per PFF. Coan was excellent when he wasn’t under duress on Saturday, throwing for 8.2 yards per attempt, 3 Big Time Throws, and posting a PFF Passing Grade of 91.0, which is elite. Josh Lugg, Cain Madden, and Tosh Baker all posted pass blocking grades below 60 (replacement level). The lack of protection also led to a breakdown in the play action game, as Coan did not have time to fake gives and then make multi-stage reads. In addition, 40% of Irish catches in this game were made by running backs, partially due to a lack of time to allow deep routes to develop.
If you remember from our preview we discussed how the pass protection has actually been good and the high sack total was more of a problem of Coan not getting rid of the football. This was not as much the case this week, and the offensive line will need to be better as the competition ramps up. To be fair to the line, they were facing potential 1st Round Pick George Karlaftis, who racked up 4 pressures in the game. But they also gave up multiple pressures to 4 other Boilermakers, two of them being interior linemen, which can’t happen.
Giving the offensive line some credit though, this was the best week so far in terms of resetting the line of scrimmage with the run game. While it was still not a spectacular week for running the football, it was the first time the Notre Dame rushing success rate exceeded Notre Dame’s passing success rate, and this is partially due to the offensive line beginning to gel as a run blocking unit.
This was a spectacular performance from the Notre Dame defense and they continue to silence critics who were concerned after Florida State. After that game we had talked about that the underlying metrics suggested a good season was forthcoming so long as they could reduce the number of big plays allowed. That’s exactly what’s happened this week. Purdue’s passing game was fine, but look at how rarely Purdue pushed the ball down the field. Just 7 of their 49 targets travelled more than 10 yards down the field and they averaged only 0.41 EPA on throws over 20 Yards, a welcome development after watching Florida State and Toledo have success attacking deep.
The Irish performance against the run was somehow even more impressive than their performance against the pass. No matter how many blockers Purdue chose to use, or what their advantage was, or how many players Notre Dame committed to the box, Marcus Freeman’s unit was able to stop the run all day. They also did so often playing without a box advantage. Normally the defense has at least one more player in the box than the offense can block, but Freeman regularly trusted his front five or six to play the run without this advantage and they rewarded him for it. I will admit I was skeptical of these light boxes after Week 1 when Florida State ran so efficiently against them, but they have worked beautifully the last two weeks. NFL fans may note the similarities in his schematic choices so far with those of Brandon Staley and Vic Fangio. Those coaches choose to run Cover 2 High most of the time and present light boxes, defending the pass first and daring teams to run. So far it has worked but September 25th against Wisconsin will be a new challenge. We’ll save the Badgers preview for Friday, but I’m already intrigued to see if Freeman feels he can continue to play a bit lighter against a heavy running Wisconsin team, or if we’ll see some uncharacteristically high box counts from Freeman next week in Chicago.