Analytics, Basic Stats, and Recent Historic Context: The ABC’s of Penn State Football – Defensive Touchdowns Outside of the Red Zone (D-TORZ) a Defensive Explosivity Metric

Sponsor: For The Blogy’s 2022 preseason coverage is sponsored by FANATICS. Spare the embarrassment of showing up to Beaver Stadium in LAST year’s sweat- & alcohol-stained gear. Grab something fresh from our collection of Penn State jerseys, T-shirts, and more right HERE.

Introduction

In the spring we looked at offensive explosivity from a new lens as the rate a team score touchdowns from outside of the red zone. We referred to that new metric as TORZ. Today, we flip the script and look at a defense’s ability to prevent opposing offenses for scoring beyond the red zone.

Defensive explosivity allowed is something that is not often discussed as a common statistic and metric. But, just as some offenses are more explosive than others, as we outlined in our TORZ post a few months ago, and that this offensive explosivity is reproducible for the best team (see: Ohio State), we would expect that a defense’s ability to PREVENT or LIMIT explosivity is also a repeatable skill. Teams, like Iowa for example under defensive coordinator Phil Parker, have garner big-play-prevention reputations over the years. Which leads us to the $64,000 question: Just because a defense can stop big scoring plays from happening, does that mean they’re actually a good defense?

Data Source and Methods

As usual, the data herein comes from cfbstats.com and includes the 2009 through 2021 seasons. Any calculations will be explained in the body of the text and graphs are created using Tableau Public.

Overall Scoring and RZ Performance

As with the previous article for offense, we want to start with the relationship of D-TORZ to total points allowed. Similarly to the offensive side of the ball, there is a relationship between D-TORZ (labeled TD PORZ) per game in the graph below. To clarify – a value of 7 means a team allows 7 points from touchdowns outside of the red zone per game. You can see that as a team allows more points, they tend to allow more D-TORZ. This is fairly intuitive. The lowest points-allowed defenses would tend to be the most disciplined ones and would not give up long touchdowns regularly.

But, it is not as simple as “good defenses are good at everything.” Looking at the graph below we see that points allowed inside the red zone (PIRZ) is very weakly correlated to D-TORZ with an R-squared value of 0.17. For those astute viewers, yes, there are 2 data points for teams that allowed zero D-TORZ/game. One of those was the 2020 Arizona State Sun Devils that played four games. The other was the 2013 Louisville Cardinals who went 12-1, allowed 12.6 PPG (t-10 best in FBS since 2009), and catapulted Charlie Strong to Austin.

D-TORZ: What’s Good?

Below is a histogram of D-TORZ/game in FBS since 2009. From this you can see that the average is between about 7 and 10 points with those being the three tallest bars on the histogram. If a team allows fewer than 5 D-TORZ/game they are in the top-20%. Those teams in the bottom-20% give up more than 14 D-TORZ/game.

Looking at the average by year we see a similar data point layout as the histogram above – meaning offensive advancements like the RPO and other schemes haven’t affected D-TORZ. During the last 12 years, the annual average has ranged between 8 and 10 D-TORZ/game with a fairly high concentration of data points around that average. Again, this shows that, despite some fairly big offensive changes over the last decade-and-a-half or so, defensive explosivity allowed hasn’t changed that much.

D-TORZ: Penn State and the Big Ten

Now we can narrow our focus and zoom-in on the Big Ten and our Penn State Nittany Lions. Let’s take an analogous look at the Big Ten like we did from a national POV above. Here, we see that the Big Ten average tends to be lower than that of the overall FBS. The lowest year of FBS (2009) was 7.9 D-TORZ/game and the Big Ten had only four seasons exceed that threshold. The 2018 season is an interesting outlier where the Big Ten had zero teams below 5 D-TORZ/game. Typically, there would be several conference teams at that level. This better D-TORZ performance of the Big Ten versus FBS speaks to two things 1. The league tends to have better defenses than many of the other conferences 2. The corn-fed, 3 Yards and a Cloud of Dust stereotype of Big Ten football has a lot of merit.

Either way, if we replace the dots with team logos, we start to see some of the common squads show up in the top of the conference. The graph below is the same as above but now labeled by team. While this is a little muddled with overlapping logos, you see that Iowa, Wisconsin, Ohio State, and Penn State are often in the top tier of the conference when it comes to D-TORX. Yes, there are outliers…2014 Iowa, 2020 Penn State, 2018 Ohio State, 2009 Wisconsin. But, on balance, these are the best in the conference.

If we modify the graph to have the x-axis by team and each dot representing one season for said team, we can better gauge the overall team performance in D-TORZ, which reinforces our hypothesis that performance in this metric isn’t random but a repeatable trend. Coincidentally, Iowa, Ohio State, and Penn State are all tied at an average of 5.6 D-TORZ/game since 2009. Indiana is the worst since that time (but as you can see, there has been marked improvement in the last six seasons under Tom Allen).

To wrap up, we compare a defense’s points/game allowed (x-axis) versus the D-TORZ/game for the Big Ten during the James Franklin Era at Penn State (2014-2021). The expected trendline (R-squared is a weak 0.55) is shown and while we shouldn’t make too many inferences, we can roughly say that teams above the line allow more D-TORZ/game than their total points allowed would portend. And teams below the line are more stingy at explosivity but allow more total points than you would expect. The below-the-line Ohio State, Penn State, and Iowa teams from 14–20 Points/G are good examples of teams not allowing explosivity, which drives towards them being the conference leaders. And while they do allow somewhat more points than explosivity alone would predict, fewer than 20 PPG allowed is in the top 15% of all FBS teams since 2009.

So what can we take away from all this? First, I think there are a couple of reasons for the trends that we laid out. Teams like Iowa, and to a lesser extent Wisconsin, may protect themselves against big plays with softer coverages in the secondary. Possibly, teams like Ohio State have good enough athletes that they’re simply not going to get beat by a ton of long plays for scores. In any case, limiting big scores does tend to show that a defense is “better” than one that does not. It will be interesting when we get to the 2022 season to see when a high explosivity offense plays a low explosivity-allowed defense. Will there be more that we can use to predict outcomes then? We’ll start to get answers in the next couple of weeks.