Analytics, Basic Stats, and Recent Historic Context: The ABC’s of Penn State Football – Nebraska Recap

Where there is no hope, it is incumbent on us to invent it. – Albert Camus

Introduction

Welcome back to The ABC’s of Penn State Football which is our weekly look at analytics, basic stats, and historic context of our favorite college football team. Maybe, just maybe, the worst of the season is behind us and there are glimmers of hope. Enjoy!

Correction from Maryland Recap

In last week’s column I discussed the explosiveness stat against Maryland in the context of the last 10 years of college football and saying that this Penn State team is on pace to have the worst season ever. I reviewed the stat over the past several days and the this was incorrect. While the performance against Maryland was very bad, it was not quite as bad as stated. Full season and single game performances cannot be directly compared based on the way the statistic is calculated. Instead, I should’ve normalized the comparative statistics to a per game basis rather than the total of the season. When this is corrected, the performance against Maryland would rank 1,250th out of the total 1,383 total FBS team seasons since 2009. Through Nebraska, Penn State’s 2020 average explosivity of 1.37 ranks 863rd and is on pace to be Penn State’s worst performance since 2015. It’s still bad, but not as poor as stated after Maryland. 

Basic Statistics – Summary

In the aftermath of the Maryland game, there was very little positivity to be taken away. Yes, Penn State gained more yards than the Terps but this was based on 32 more offensive plays. The Lions turned the ball over, couldn’t run, couldn’t complete passes – pretty much everything was bad. Well – against Nebraska, Penn State still lost the turnover battle (2-to-1) but outgained the Cornhuskers by over 200 yards with a better yards-per-play average, appeared to run the ball with purpose, and just generally had more energy (especially in the 2nd half) than they showed a week ago. If you didn’t look at the score row in the table below and you only knew that Penn State won nearly every basic statistical category, you’d think they won handily. Unfortunately, points are all that matter and in the battle to get the ball over the goal line or through the uprights, Penn State still fell short. 

Basic Stats Offense – Context

The graph below perfectly summarizes the frustration of the offensive production of the 2020 Nittany Lions for me. It is a plot of yards gained in a game (x-axis) versus points scored in that game (y-axis). Each gray circle represents one Penn State game since 2014 and the black line is the expected point total for a given yardage output. The further a point below the line, the worse the scoring relative to yardage gained. If a point is above the line, it indicates that the offense is scoring at a higher than expected rate. The 2020 Nebraska game, with 501 yards generated, should’ve yielded about 40 points. You implicitly knew that this relationship existed if you watched the game. After Will Levis came in, the team moved the ball fairly well until it got inside the redzone. (To note, I’m not suggested that outside of the redzone they were LSU 2019. Even when they were moving the ball it was clunky and disjointed but they were chewing up yards.) But when they field got compressed at our near the 10 yard line, they ran out of answers. On the last two drives they ran Penn State ran 8 plays at or inside the Nebraska 11 yard line went 0-for-7 passing with 1 sack allowed and gained -4 yards. That’s… not great Bob. Not great at all.

In addition to yesterday’s performance, the performances against Indiana and Maryland are below the line. Only the Ohio State game, where the team gained a paltry 323 yards of offense and scored 25 points, did they manage to reach the expected level of point output for the yardage gained.

There may be a glimmer of hope however and we need look no further than the comparison of the Maryland and Nebraska games where there was one very big difference. Statistically speaking Maryland and Nebraska’s defenses are comparable – they are similar levels of bad. Additionally, the Lion offense ran virtually the same number of plays in each game (91 against Nebraska and 90 against Maryland). In the graph below we see how these 91 plays manifested performance in each game. Bigger circles mean more plays, darker blue represents a higher per play average for the given measure, and the y-axis indicated the total performance in total yards, rush yards, and passing yards. The offense was just better against Nebraska than it was against Maryland. Yes, there were fewer passing yards against the Huskers but PSU averaged (a still poor) 6.6 yards-per-attempt versus 6.0 against the Terps. Additionally, Penn State relied much more heavily on the run against Nebraska and rushed 52 times with 39 pass attempts (57% run play ratio). This is a complete flip versus the Maryland game where the team threw on 63% of their plays (57/90). 

Finally, and not to pile on Sean Clifford, but it must be said that the QB switch yesterday generated results. For the whole Maryland game plus the first three possessions against Nebraska, Penn State averaged 34 yards per possession, had 5 turnovers, and scored 22 points. Will Levis had 8 possessions against Nebraska and generated 391 yards of offense, gave up zero turnovers (not counting failed 4th downs), and scored 20 points. This is the invention of hope…

Basic Stats Defense – Context

I don’t know what to make of the defense. On one hand, they’ve done a decent job of limiting yardage in two games holding Indiana to 211 and Nebraska to 298. On the other hand, teams are scoring 34 points per game against Penn State. And I know that against Nebraska, the defensive only allowed 23 points but this year is still an outlier. Look at the graph below. Each point represents a game and each line is the season trendline for points allowed for a given number of yards since 2014. From 2014 – 2019, each line has a roughly similar slope (indicating more yards allowed would give more points). There is one clear outlier though and it is the purple line. That’s the 2020 season through 4 games. Based on the current trend, no matter how many yards the opposing offense gains, they’ll score between 30 – 40 points. I write that statement somewhat jokingly but it is an interesting trend so far. Indiana gained 211 yards and scored 36 points. Ohio State gained 526 yards and scored 38 points. Maryland – 405 yards and 35 points. And finally, Nebraska gained 298 yards and scored 30 total points (granted 23 against the defense). Through four games, it seems that the total Penn State team is bought in to giving up at least 30 points per game. That seems a poor strategy. But again, considering that this is a clear outlier from the previous six seasons under James Franklin, we should see regression to the mean sooner or later and the purple line should collapse towards the others. 

And again, and now maybe I’m piling on a bit, but there was absolutely a difference in performance of the defense before and after the quarterback change. 

Before Will Levis took over, the Penn State defense: 

• Yielded 150 yards on 3 possessions (50 yards/possession)
• generated zero turnovers, one tackle-for-loss
• gave up 17 points (plus the fumble touchdown)

With Levis in the game, the defense:

• gave up 148 yards on 7 possessions (21.1 yards/possession)
• forced 4 punts and generated an interception
• had one sack and 4 TFL’s
• gave up 6 points

As we hear all the time, correlation does not equal causation, and the infusion of Levis’ energy into the offense probably didn’t cause the defense to play better. But you’d better believe there was correlation to the defense’s performance before and after the switch at QB. Assuming Levis starts against Iowa this week, will this trend continue? Again, we’re inventing our new hope here…

Advanced Stats – Defense

Last week we focused the Advanced Stats section on the offense and since it’s still not great, let’s turn our attention this week to the defense. A couple of stats we’ll look at are my version of Havoc rate, allowed efficiency, points per minute, and overall defensive effectiveness.

We start with Havoc Rate. Havoc Rate is the % of plays where the defense does something demonstrably good – sack, tackle for loss, interception, or recover a fumble. Since 2009, in FBS, the highest Havoc Rate that I’ve calculated is 24% by six different teams including 2017 Michigan, 2014 Clemson, 2016 Michigan, and 2019 Ohio State. The 2015 Penn State defense was the best PSU team of the last decade and had a Hacov Rate of 22%. In the graph below, you can see the full year performances of 2014 – 2019 Penn State and then the individual games this year. An average defense will generate Havoc on 15% of their plays and PSU’s lowest season of the Franklin era was 17.9% in 2014. I am certain that there have been some low performing games during this stretch, but through four games this season they are just average (14.7%) with one good game (Maryland), and two poor games (Ohio State and Nebraska). This team just doesn’t seem to be at the level of the previous Pry lead defenses at consistently generating negative plays for the opposing offenses and it’s allowing drives to continue and the points allowed to go up.

The next sign that this team is either below average, or if we’re being generous, inconsistent, is the defensive efficiency. I calculate it as: sqrt(yards/possession*minutes of time on field). It’s a bit complicated but it’s an estimate of how well an opposing offense moves the ball for each possession controls the game. The lower the number, the better the defense is and the quicker the defense is getting off the field – average is 30; best since 2009 is 21 (2011 Alabama). By this metric, on a full season basis, the Penn State defenses under James Franklin have always been better than average with the high-water mark being 2014 at 25.6 (which ranks 38th overall since 2009) and the “worst” being 2019 (30.9, still basically average). This year, the overall performance is 28.5 which is technically better than average but that includes a GREAT showing against Indiana (again… 211 yards allowed on 15 possessions where the defense was on the field for less than 20 minutes) and a MISERABLE showing against Ohio State (526 yards on 11 possessions). The last two games have been somewhat more consistent but they need to continue to dial it in and get off the field more quickly before I’ll feel confident.

In terms of overall effectiveness, I calculate a combination of the efficiency above with points allowed per minute the defense is on the field. And here is where we see real weakness in the defense. In Franklin’s tenure, they have not been great at minimizing yards allowed but they’ve consistently been among the national leaders in points allowed where, except for 2016 where they allowed 25.4 points per game, they were always 21.8 or fewer. And when we combine the efficiency with the point allowance in 2020, we have real reason for concern. First off, in points-per-minute, the defense is allowing an average of 1.3 points per minute they’re on the field. Compare this to the 2009 – 2019 national average of 0.9 and the PSU average of 0.66 over the same time period (0.64 under Franklin). In context, if the opposing offense has the ball for 20 minutes, they’d expect 26 points this year versus 13 points in years past. This is a huge difference and points to this year’s squad allowing LOTS of big plays and/or the offense turning the ball over in the opponent’s territory leading to short, quick scoring drives. These are both things that happened far less frequently than before and is a danger sign and gives an overall defensive efficiency of 37.3 (2009 – 2019 global average: 29.4 – lower is better; PSU 2014 – 2019 average: 18.6). Very not good.

So here we are, halfway through a season with a defense that we expected to be above average that isn’t as active and aggressive as we’d hoped (Havoc) and giving up far too many points. It’s a combination that was quite unpredictable going into the year and for this team to be competitive in the back half of their schedule, they must change something. 

Summary

I want to end this week’s column with a story. It goes back to the Penn State-Michigan game from 2013 and it’s one that’s probably memorable for many fans. For my family and me it is one that will live with us forever. It was exactly one week before my 35th birthday and my parents in-law were visiting us in Texas from Pennsylvania and my brother in-laws family was coming from California a couple of days later. I was working on building some Penn State cornhole boards all weekend and had lots of work to do in the garage. We all remember that it was a back and forth game that Penn State led 21 – 10 at halftime. Then, following a 17 – 3 third quarter, the lions trailed 27 – 24 going into the 4th. Michigan scored again to go up by ten. We hadn’t been very good that year and I went to work in the garage because I thought it was over. Some time later, my mother-in-law called me back into the house and we were getting the ball back down by seven. You all remember the – “Robinson comes down with it!” – pass from Hackenberg to send it to overtime. Penn State missed their first field goal attempt in overtime and probably should’ve lost on Michigan’s first possession (also a missed field goal). After trading FG’s in 2OT, Penn State fumbled on their first play in 3OT and were lucky to survive a 2nd missed Wolverine field goal. It was only in the 4th overtime that Bill Belton broke left to score the game winning 2-yard touchdown. Following the game and seeing how many times I thought it was over and we were going to lose, my mother-in-law said – Never, ever give up on your team. You just don’t know what can or will happen. Why is this memory so poignant to me? It’s because that was the last game she and we would ever watch together. She passed away the following week on October 19 but that memory and the message she left us with, will never leave us and is very impactful not only in our Penn State fandom but in everything we do. There is almost always a glimmer of hope and reason to keep pushing through the dark times towards what can be a bright future. We Are!