Analytics, Basic Stats, and Recent Historic Context: The ABC’s of Penn State Football – The Mike Yurcich Hiring

In a move that was more shocking than The Red Wedding, Penn State has another new Offensive Coordinator. Today, we analyze what to expect from Mike Yurcich who (like yours truly) has been privileged with significant PSAC experience in his days.

Introduction

By now, you’ve probably watched the excellent video breakdown on fortheblogy.com on what to expect schematically from Coach Yurcich’s offense. If not, click back to the main page, scroll down and check it out. We’ll wait…

…OK, welcome back. Coach Nick does an amazing job with X’s and O’s, but when it comes to ABC’s we focus on numbers, statistical trends, and results. Today, while we’re all still in the honeymoon phase of this hire, let’s dig into what Yurcich did as an offensive coordinator at Oklahoma State and Texas – and compare those tenures to what Penn State’s previous OC’s accomplished.

Background

Before we start, let’s set the stage a bit. All data herein is sourced from cfbstats.com, which is a tremendous resource for folks like me. It only includes games against FBS competition between 2009 and 2020, and all graphs were made via Tableau Public. 

The focus years are:

• Mike Yurcich: Oklahoma State (2013 – 2018) and Texas (2020)
• Kirk Ciarrocca: Western Michigan (2013 – 2016); Minnesota (2017 – 2019); Penn State (2020)
• Ricky Rahne: Penn State (2018 – 2019)
• Joe Moorhead: Penn State (2016, 2017), Mississippi State (2018, 2019), Oregon (2020)
• John Donovan: Penn State (2014 – 2015)
• Bill O’Brien: Penn State (2012 – 2013)

I am aware that Joe Moorhead wasn’t the OC at Mississippi State (it was Luke Getsy) but I am choosing to include those Bulldogs teams since Moorhead HAD to have a very significant impact on those offenses.

Finally, I fully understand that the defenses of the Big 12, where Mike Yurcich has been for his whole FBS OC career, are likely not as strong as those the previous Penn State coordinators have faced in the Big 10. But, most of what we’ll present is completely controlled by the offense…meaning the stoutness of opposing defenses doesn’t matter. Things like Run-Pass Splits, Tempo, Havoc Avoidance (to some extent) should all be relative from one conference to the next. Others statistics like Scoring, Total Yards, Explosivity, Overall Offensive Effectiveness need to be taken with a grain of salt since those data points are helped/hurt depending on who you’re playing. We will try to look deeper at the relative conference performance for Yurcich when we’re bored and craving football in the spring and summer but for now, enjoy this monumental change and get excited for the future!

The Basics

We’ll start by looking at how each of these coordinators mix the run and the pass and how balanced they are. Perfect balance would imply a 50/50 run-to-pass split and from the histogram below, we see that most FBS teams are close to this ratio. The height of the bars represents the number of teams that were within a certain percentage of their plays being passes (Pass Share %). A lower Pass Share % implies a more run dominant team and vice versa. The extremely run-heavy teams are mainly the Military Academies (we thank you all for your service!) and other option teams like pre-Geoff Collins Georgia Tech and Georgia Southern. The teams who are upwards of 70+% passing are those run by Mike Leach or others with pass-heavy “Air Raid” schemes. But mostly, teams fall within 35% – 55% Pass Share. 

Now, let’s isolate on the playcaller we’re studying. In this graph, we see (by year, team, and OC) how run or pass dominant each OC had been. Most of the previous Penn State OC’s were somewhat run-dominant. Only in two years – 2014 and 2017 – did Penn State throw the ball at least 50% of the time. Kirk Ciarrocca, for his career and during the 2020 Penn State season, ranges between 36-44% passing. 

Look at Coach Yurcich’s data, though. His lowest pass share year was 44% (2014 Oklahoma State) and in total has averaged 49% passing. Therefore, expect the Nittany Lions’ pass share to go up this year versus what we’ve seen in recent Penn State history – this should be good considering we have some excellent pass catchers in Jahan Dotson, Parker Washington, Theo Johnson, and the running backs. As far as the current roster of pass throwers, time will tell…

Next, let’s explore the yard share for running and passing by OC. Below, we have the average seasonal rush yards per game (x-axis) against the average passing yards per game (y-axis). Here, I think there are several useful observations. 

• First, it is clear that Ciarrocca has always been a pretty run-dominant guy, as we saw this year. There are a lot of Gophers and Broncos in the bottom right-side half of the graph that proves this. 
• Next, Moorhead’s MSU teams really didn’t do much through the air. At Penn State, his teams averaged 261 and 290 passing yards per game, and he averaged 246 PYPG at Oregon in 2020. At MSU, though, the Cow Bellers generated 155 and 177 passing yards per game in 2018 and 2019, respectively.
• From 2012-2020, Penn State has been pretty middle of the road compared to the rest of the FBS with 162 RYPG (FBS average: 167) and 246 PYPG (FBS average: 229). 
• Look at the Oklahoma State and Texas logos, though. Not only have the Yurcich teams been mostly comparable in terms of rushing yards to most of the Penn State teams, but they also generated considerably more passing yards. The two exceptions are 2014 (131 RYPG) and 2015 (124 RYPG). Even with those two clunker seasons, Yurcich teams have averaged 164 RYPG and 311 PYPG with a high-water mark of 389 yards in 2018. So, if we get a comparable rushing attack as previous Penn State squads with an extra 60 or 70 passing yards per game, that would be fantastic.

Let’s also look at yards-per-play (YPP). An average FBS team will generate 5.6 YPP. In the graph below, you’ll see Penn State has employed OC’s who generally exceed the average…sanction-stricken years notwithstanding. The 2020 team, with all of their QB draws and low-percentage fades still managed to generate a respectable 5.5 YPP. Coach Moorhead’s Penn State and Oregon offenses were fantastic in this regard racking up 6.5, 6.6, and 6.8 YPP in 2016, 2017, and 2020 respectively (which rank around or within the Top 10% of all offenses since 2009). 

However, there’s one name on the graph above that’s slightly above Joe Moorhead. That’s Mike Yurcich. His average is 6.4 YPP — ranging from 5.3 (2014) to an incredible 7.3 (2017). The 7.3 number is the 24th best in the FBS since 2009. Three Yurcich offenses finished in the Top 125 out of 1510 schools charted during this time period. 

Okay, Yurcich’s teams generate yards – but yards don’t win games (see weeks 1, 3, 4, and 5 of Penn State’s 2020 season). No, points win games. Let’s go back to another histogram to look at the whole landscape of FBS for the last decade. The global average of per-game scoring is 27.7 points and that’s borne out by the shape of the plot below. Teams tend to average between 21-33 PPG. 

If we isolate our subset, we see that most of our OC’s fall in the 21-33 PPG range for their careers (career average represented by the dashed lines). Even Coach Moorhead falls in this range for his career at 32.4 PPG — although his Penn State teams exceeded that range thanks in part to tremendous personnel.

There is one exception though, and that’s Mike Yurcich, whose teams average 38.1 PPG – a number brought down by 26.6 PGG 2014 Oklahoma State squad. To put a finer point on this, from 2009-2020 there have been 1,510 seasons played by FBS teams and 88 of those teams (5.8%) have scored more than 40 PPG – Mike Yurcich called plays for three of those 88 teams and ranks 27th (45.0, 2017 OSU), 53rd (42.7, 2020 Texas), and 86th (40.2, 2015 OSU). Having one season in the top 5% would be great and could be considered a fluke – having three of them is outstanding.

Advanced Stats

Tempo

I am sure there are different, and likely better, measurements for tempo than this…but a good proxy is Plays-Run-per-Minute-of Possession – so total offensive plays divided by total Time of Possession. Make sense? We’ll abbreviated this stat as Plays/Min. Below we have the histogram for Plays/Min minute over the last 11 seasons of FBS. The bulk of the teams run about 2.0-2.5 Plays/Min. There are some slow teams (<2.0 Play/Min) – again run-first clubs like Army, Stanford, Georgia Southern. Then, there are some other offenses that go really fast (>3.0 Play/Min) with the 2016 Missouri team setting the record at 3.4. The averages are: median = 2.32 and mean = 2.35. For the sake of this discussion, we’ll classify “very fast” as anything in the Top 10% (Top 150 since 2009) as >2.69 Plays/Min and “very slow” as <2.06 Plays/Min (bottom 10%).

So how do our guys stack up? Well, buckle up Nittany Nation; this will likely be the fastest Penn State offense we’ve ever seen. The fastest team of the last 12 seasons was the 2013 squad that averaged 2.53 Plays/Min. Ricky Rahne’s 2018 and 2019 teams were also slightly above average at 2.45 Play/Min combined. As you can see below, Kirk Ciarrocca never led particularly fast units, as his average is only 2.13 Plays/Min (2.34 for 2020 Penn State). 

Contrast all of these to Mike Yurcich’s teams that have averaged 2.65 Plays/Min. Yurcich’s slowest team, 2014 Oklahoma State, still ran 2.48 Plays/Min which is not that far off of the fastest Penn State team. Yurcich does seem to be a bit binary where he has 4 seasons at 2.57 or less and 3 at 2.76 or higher. Time will tell which direction he chooses to go with the 2021 Penn State squad but I would bet my house that it will be above 2.5 Plays/Min and approaching the “fast” zone. 

The one potential drawback of this is that Yurcich has never controlled an offense that held the ball for an average of 30 minutes per game (range of 27.1 – 29.3). This year, we hoped that the Ciarrocca ball-control offense would help keep the defense fresh. Yurcich will absolutely be a departure from that. So they’d better score in bunches to make up for it.

Havoc Avoidance Rate (HAR)

Reminder, HAR is the percentage of plays an offense runs where something bad happens (fumble, interception, sack, or tackle for loss) and is a derivative of Bill Connelly’s Havoc stat for defenses. As an offense, you want to minimize Havoc plays (obviously) thus maximizing your HAR. So the higher the HAR, the better for an offense. Average HAR performance is 86% and good (Top 10%) is 90% while bad (Bottom 10%) would be 83% or less. Havoc Avoidance was supposed to be a hallmark of the Ciarrocca offense at Penn State. Prior to 2020, his teams had an average HAR of 88% with the best being the 2016 WMU squad with an impressive 92% HAR. No such luck this season. Early on, the Penn State offense struggled mightily with turnovers. The Lions turned it around post-Iowa but still only managed a total HAR of 85.8% that isn’t bad per se but it was not what was expected. 

You can see from the graph below that this is a potential area of concern for Yurcich’s offense. While he’s only had one season (2014) that was approaching “bad” territory, he’s typically been around average in HAR. I say “potential area of concern” mainly because of both Sean Clifford and Will Levis’ struggles to avoid sacks and protect the ball. Less bad stuff happening is a priority next season.

Let’s look one layer deeper at the Havoc plays per game – which helps us ID what’s contributing to these HAR percentages. Impressively, Bill O’Brien’s teams only allowed 1.4 turnovers and less than 6 TFL and sacks combined per game. That’s why they had HAR’s that averaged 90%. Conversely, the 82% average HAR of John Donovan yielded 1.7 turnovers and 10.6 TFL and sacks. I say this every time…while those differences may not seem significant, as an offense you have 70 plays per game or so. Yielding 3 or 4 or more additional Havoc plays than your competitor can be disastrous. You need to protect the ball.

With that in mind, we see that Yurcich’s teams have allow slightly more tackles and sacks than Moorhead or Ciarrocca, but fewer turnovers. It’s not O’Brien good, but it’s not Donovan bad either. 

Explosivity

Explosivity is the measure of yard and point generation per 100 plays. It’s highly correlated to points-per-game, as illustrated in the plot below, but also allows us to measure the overall productivity of an offense. The graph below shows all of the 1510 FBS teams since 2009 and you can clearly see the correlation. The average explosivity is 1.6. If a team falls far below the line, they are generating a lot of yards-per-play but a lower than expected points. If they’re above the line, the opposite is true (i.e. higher than expected points on fewer yards per play). A top 10% team has an explosivity of >2.39 and a bottom 10% team is <0.9. 

Thankfully, for a seasonal average, no Penn State team has been in the bottom 10%. The four years of Moorhead and Rahne saw above average Explosivity with close to expected production. To prove the point in relation to the trendline, notice the Oregon logo (Moorhead 2020). That team had a reasonably good explosivity >2.1 but generated about 3 points less per game than expected. This is reinforced by their high yard-per-play value mentioned earlier. 

For Coach Yurcich, his teams produced BIG TIME. He had four squads that were Top 10% Explosivity from 2009-2020 with the 2017 Cowboys being the 21st most explosive offense since 2009. Only Joe Moorhead has coached Penn State seasons that rival Yurcich with Yurcich having 6 seasons with better explosivity than any Penn State coordinator not named Joe Moorhead. Notably, Kirk Ciarrocca had several good years as well (2015 and 2016 WMU; 2019 Minnesota) and I have no doubt he would’ve figured it out at Penn State too, but that will only happen in a parallel universe.

Red Zone Scoring

One of the big frustrations for Penn State fans during 2020 was the poor red-zone performance early in the season. Whether it was turnovers and missed field goals against Indiana, fade after sad fade against Nebraska, red zone efficiency left something to be desired. For the year, Penn State scored on 75.6% of their red zone attempts (51.3% TDs; 24.3% FG’s) and as you can see, that’s pretty far below the global average of 82%. For what it’s worth though, Ciarrocca teams are typically at least marginally better than this and he holds an average of 85% excluding 2020. 

That being said, we appear to have found another strength for Mike Yurcich. His teams scored on average 89% of the time they’re in the red zone. The lowest performance of his career was the 2020 Longhorns with an 85.7% scoring rate (66.1% TD’s and 19.6% FG’s). Outside of 2020 and 2013 (86%) his squads have never been below 88.5%, which, again, is outstanding.

Offensive Effectiveness

To conclude, we will end with Overall Offensive Effectiveness (OE). This is my attempt to combine ball control, efficiency, and scoring effectiveness into one number. An average offense has a value of 29.1 and a Top 10% offense since 2009 is >41.4. We’ll ignore the Bottom 10% for this discussion since it’s not relevant.  The best Penn State team since 2009, and probably the best ever with the exception of 1994, was the 2017 squad who produced an OE of 45.3. Led by Saquon and Trace, that team ranks 90th overall since 2009. This is Penn State’s only Top 150 offense of the last 12 years (2016 was close and ranks 161st). Penn State’s third best team was 2018 (36.0, 328th).

New offensive coordinator Mike Yurcich orchestrated two clubs better than that epic 2017 PSU offense (2017 Oklahoma State and 2020 Texas) and an ADDITONAL three teams in the Top 10%! Out of his 7 years as an OC, he’s had 5 season that were better than any PSU season except for 2017. His 6th best season (2013) ranks 195th overall. Only the 2014 OSU team was below average.

And for those who are still skeptical about the differences in the Big 12 versus Big Ten and the new challenges that Yurcich will face up north, I would like to present some SP+ (from Bill Connelly, currently at espn.com). Bill’s SP+ system is regarded as one of the gold standards in college football analytics. It is conference-adjusted but I don’t have his 2020 data since they’re now behind a paywall but we can still look at 2009 to 2019. For SP+, we can rank the Yurcich years at Oklahoma State to the Moorhead and Rahne years at Penn State. This correlates all that you’ve read so far with a system that has been around for over a decade and has national prominence. Interestingly, SP+ has three Yurcich teams above the 2017 Penn State squad, whereas OE has only one higher. There are other minor rating and rank differences but nothing substantial (except OE ranks the 2017 Cowboy team as the best of the group where SP+ has the 2016 team marginally higher). The overall point, though, is that Mike Yurcich is a great hire, and that we as fans should be very excited for the future offense at Penn State no matter the ranking system.

Pattern photo created by jcomp – www.freepik.com