Is Oklahoma State’s defense more important than its offense?

With the departure of under-appreciated offensive coordinator Mike Yurcich, I was curious how his offenses stacked up against the other eight offenses in the Mike Gundy era. While doing that, I decided I had to see how new defensive coordinator Jim Knowles faired vs the kings of “bend don’t break” (Glenn Spencer and Bill Young) who covered nine of Gundy’s 14 seasons (SPOILER ALERT…it didn’t fair well).

There are a lot of things that go into determining how good or bad an offensive or defensive unit is. In this day and age of high-speed spread offenses, points-per-drive is one of the favorite numbers cited.

While there is a difference between scoring a touchdown and a FG, I would like to look at the percentage of drives in a game that DON’T result in a score, utilizing conference games only and ignoring end of half/game and OT possessions.

This data doesn’t easily exist anywhere, so I relied on game box scores going all the way back to 2005. I reviewed the drive charts and team stats manually so I may have missed an eligible drive or two, but overall the data matches up to W-L records so I’m comfortable that I have a good data set to review.

What I found was pretty fascinating, and confirmed the importance of a tried and trusted data point…turnover margin…in explaining why teams with similar numbers post different records.

Let’s look at offenses first.

Not surprisingly, 2011 comes in at No. 1, not scoring on 45.4% of the eligible drives. Likewise the 2005 squad, responsible for the only losing season in the Gundy era, holds down the bottom spot having not scored on 75.5% of their drives. Only three offenses have managed to score on more than 50% of the eligible drives (2008, 2011, 2017), and only four offenses have scored on less than 40% of their drives (2005, 2009, 2013, 2014).

If we rank the offenses top to bottom based on percentage of non-scoring drives (low is good, high is bad), we get this:

While Yurcich’s offenses have been pretty good, four of his units finish in the bottom half of this Gundy Era ranking. It’s not a surprise to see 2014 here, but 2013 and 2015 are shockers. The top seven (lowest percentage of non-scoring drives) had an overall conference record of 41-19. The bottom seven came in at 31-29, so there is definitely a correlation between these percentages and W-L records which should be no surprise.

One of my pet peeves with the non-scoring drives is going “three and out,” so I decided to look at the percentage of non-scoring drives involving three or fewer plays.

Imagine my disappointment when I discovered that the percentage of non-scoring drives involving three or fewer plays virtually doesn’t matter. The overall records of the top seven vs. the bottom seven are almost identical, making the fact that the team didn’t score more important than the number of downs played when not scoring.

How do new offensive coordinator Sean Gleeson’s Princeton offenses stack up against these historical numbers? Utilizing only Ive League games, the Tigers posted the following over the past two seasons with Gleeson as the OC:

• 2017 and 2018
  11.1 eligible drives per game (only 2006, 2007, & 2008 offenses averaged less than 12 eligible drives per game under Gundy)
• 2018...4.6 eligible non-scoring drives per game
• 2017...5.1 eligible non-scoring drives per game

So Gleeson’s offenses failed to score on 46.2% of their eligible drives in 2017 and just 41% of their eligible drives in 2018, which would be numero uno on OSU’s list by quite a bit. We’ll see how that translates to the Big 12 and a much faster style of play, but...

It takes two to tango, so let’s not leave the defense out any longer.

The greatest defense in the Gundy Era is 2013 and it’s not even close.

If you want to know why OSU almost brought home a Big 12 title that season, look no further than defensive side of the ball. It’s actually astonishing how good they were when paired with the fourth worst non-scoring offense under Gundy. Because the offense played quickly and gave the ball back without scoring 63.6% of the time, the 2013 defense did the following:

· Faced 14.8 eligible drives per game (next closest was 13.9 in 2005)

· Held opponents scoreless 76.7% of those drives (11.3 scoreless drivers per game)

· Produced 25 takeaways in Big 12 play, which is #2 in the Gundy Era (2011 with 33 is No. 1… more on this in a minute)

If we rank the defenses by non-scoring drive percentage (high is good, low is bad):

Remember Bill Young? Bend don’t break?

The top seven defenses had a combined record of 46-15 while the bottom seven posted 26-33 which is an even stronger correlation to winning and losing than the offensive production. Like the offenses, the defense’s ability to get off the field in three or fewer plays on non-scoring drives meant nothing, with the seven best combining for a record of 36-25. The seven worst? 36-23…

So essentially what we are seeing is a lean towards the old axiom that defense wins championships. It is clear from the data that if one side of the ball has to be more productive than the other it’s best if it’s the defense. In other words if your offense isn’t great, the defense can help them make up for that in volume of possessions.

The wild card in all of this? Turnover margin.

This just goes to the point that the greatest season in OSU history, to a large degree, was a fluke. We all know about Brandon Weeden, but let’s take a look at one of the craziest turnover margin numbers in the history of college football.

Dave Hudson posted this back in 2013 on how turnover margin can impact a team’s season. While it does not correlate perfectly to wins and losses, it does indicate the potential for additional wins above what a team is capable of and can impact results positively or negatively.

A great example of this is OSU’s 2012 and 2017 teams.

How good was the job Todd Monken did in 2012? He took inexperience and a revolving door at QB and produced the 4th best non-scoring drive percentage in the Gundy Era. Bill Young engineered the 4th best defense under Gundy in that category, but unfortunately a -1 turnover margin in Big 12 play doomed that squad to be just what they were and nothing more.

The 2017 season is probably the comp for what would have happened to the 2011 team had that defense not come home with an historically absurd number of takeaways. Offensive numbers in this study are virtually identical. Weeden and Company gave the ball away 16 times in conference play, while Rudolph’s unit did so 17 times. The HUGE difference maker was that the 2011 defense got the ball back twice as many times whereas the 2017 defense posted the Gundy Era average of 16 takeaways in Big 12 play.

How do you influence turnover margin? Marginally at best, and in Hudson’s stats turnover margin was only 37% responsible for win totals.

Defenses can practice stripping the ball and working on tip drills for interceptions. Offenses can focus on how the ball is carried and securing the ball once a pass is caught. However, much of the turnover game is just pure luck.

OSU had a turnover margin of +17 in conference play in 2011 as the defense pulled in 33 takeaways. That’s almost two extra possessions per game. The problem? The defense wasn’t very good when not producing a turnover. Let’s compare 2011 to 2013, which posted a turnover margin of +11 with 25 takeaways.

While the 2013 and 2011 defenses produced the Nos. 1 and 3 non-scoring drive percentages (76.7% and 73.1%), these numbers came about in very different ways.

The 2011 defense posted 87 eligible non-scoring drives in Big 12 play. Of those, 34.5% (30) ended with a takeaway, meaning 65.5% (57) ended with a punt or downs. On average, the 2011 unit allowed 5.5 plays and 20.8 yards and were on the field 2:12 per non-scoring drive, which all rank near the bottom of the defenses in this study.

The 2013 defense posted 102 eligible non-scoring drives in conference play. Of those, only 20.6% (21) ended in a takeaway while 79.4% (81) ended with a punt or downs. On average, the 2013 squad allowed 4.6 plays and 15.3 yards and were on the field 1:48 per non-scoring drive, which of course all rank near the top.

Over the whole of Big 12 play, while facing 15 more non-scoring drives than the 2011 unit, the 2013 defense faced fewer total non-scoring plays (467 to 477), gave up fewer non-scoring yards (1557 to 1811), and spent less non-scoring time on the field (3:03:24 to 3:11:10).

How good are those numbers? Look at it this way…

Arguably the greatest offense in OSU history needed a washed up minor league pitcher who decided to go back to college and become a football walk-on, and one of the most ridiculous turnover margins in the history of college football in order to accomplish the greatest season in OSU history.

The best defense in the Gundy Era single-handedly came THAT close to lifting the fourth worst non-scoring offense in the Gundy Era to a conference title, and they did it largely by getting off the field in the traditional manner. This squad was a couple of plays from having a Big 12 title and quite possibly the same record as 2011.

I think the point is clear and I don’t think it’s any surprise that having a defense capable of getting off the field by forcing shorter drives and punts should be a team’s No. 1 goal. Turnovers are great, but you can’t depend on those for the wins.

Let’s hope Jim Knowles’ second season at DC shows significant improvement. His first go round was abysmal in this statistical study. His non-scoring percentage of 53.2% ties 2005 as the worst in the Gundy Era. His defense also posted the lowest number of turnovers under Gundy in conference play, leading to the 3^rd worst turnover margin in that time.

IMHO, Gundy’s hire of Gleeson for offensive coordinator won’t be nearly as important as whether or not his recent choice for defensive coordinator pans out. (calm down, I didn’t say offensive coordinator WASN’T important)

There’s a lot more to breakdown here but I’m not looking to write a novel. Here are the files I constructed for both offensive and defensive data. Enjoy!