Section I: Phantom Offense
The reason I call 2018 season bizarre was because:
- It was it was so bad it ranked in the bottom decile of any statistical measure see: S&P+ (118th out of 130, sandwiched between Northern Illinois and East Carolina)/OFEI (122nd out of 130, between UTEP and Kent State) ranking. In 2017 the offense wasn’t world breaking but it did its job staying on the field and scoring points (62nd in S&P+ and 87th in OFEI), but in 2018 it was a phantom of itself, in a regression so extreme despite a lack of a coaching change on the offense and most of the starters from 2017 remaining.
- And more importantly (from my point of view) it had a run/pass ratio below the conference average and below the way the scheme was supposed to work with a multiple spread based passing attack by OC Baldwin. This went against the usual mantra of analytics of favoring the pass over the run.
There are many factors:
- Ross Bowers, the incumbent starter was injured right before the UNC game after taking the lion’s share of 1st team snaps in Spring and Fall camp.
- The return of a 3 down back in Patrick Laird who ended up leading the team in runs attempts and catches (!).
- Chase Garbers/Brandon McIlwain having to step in and shuffle in and out of drives as the offense tried to find its identity.
- Chase Garbers’ shoulder fatigued by the lack of available QBs in practice putting additional strain.
By the end of the season, with Ross Bowers injured and electing to transfer and Brandon McIlwain moving to running back, Garbers and Chase Forrest took nearly every snap and stressed their arms in practice.
Garbers would never admit it in the moment, but his right throwing shoulder was shot. He would see play calls asking for 40-yard passes, and he knew he couldn’t get the ball there.
All these factors above are explainable and most importantly not part of the fundamentals of the offense. Of course there are concerns about the fundamentals of the offense:
- Uninspired play-calling over the last two seasons.
- Lack of high-upside football talent in the skill positions (see Cal’s last two recruiting classes where only Nikko Remigio was not a 3* recruit).
- Lack of development in the WRs on a year-to-year basis.
In this series of articles I want to look into how Cal looked on a play-by-play basis using play level data gathered by SportsDataStuff.com. He scraped data from ESPN’s database and condensed it into a .csv file that I cleaned up and processed using R in order to arrive at the charts below.
In this part I will focus on the Run/Pass splits by down/quarter and opponent.
Section II: Run-Pass Tendency
Football is complex. It is why I love it so much, the schemes, moves, counters, and counters to counters are what make waiting for each snap worth it. There is so much football thinking happening as teams huddle up and the play-clock winds down. Despite that scheming it all boils down to a binary decision by an offense: to pass or to run, that is the question.
In 2018 Cal passed the ball 422 (high of 43 against Arizona and low of 24 against Washington) and ran the ball 488 times (high of 49 against UNC and low of 30 against TCU). This takes the Run/Pass Ratio to 53% (meaning 53% of the time Cal will run the ball).
Bic picture: Pac-12 teams averaged 7.58 yards per pass and only 4.125 yards per run, yet they passed 5143 times and ran 5559 times (Run/Pass Ratio of 52%), this means that Pac-12 teams (disregarding the game-script) left a lot of yards on the field.
This is the first chart depicting the general decision making of the Cal offense on each down, by opponent. The decision to do what when is often dictated by the distance of each down. What is universal is the 1st and 10: Here we can see the decision making was pretty on par with what the Cal’s opponents have faced throughout the season.
(Note: Football outsiders states that analytically speaking the most efficient play on 1st down is a pass. Passing on 1st down is correlated with lower punting rates, and higher expected points per drive since.)
Only in the case of UW/WSU/USC did Cal run more often on 1st down than passed. In none of the cases did Cal stand out on either end of the distribution on 1st down. What is more interesting is the increase in run plays in 2nd/3rd downs especially early on the season when Cal shuffled Garbers/McIlwain on those downs. However, we can see that during the Arizona game where McIlwain took all of the snaps Cal consistently passed more than the average team, especially on 1st down.
Here is hoping Cal can get more aggressive with passing on 1st/2nd downs with the renewed confidence in the passing game and the QB. There are uncertainties about the receivers but I think in order to give them a chance to show what they can do OC Baldwin/QB Garbers need to throw and give them a shot.
By quarter Cal has been adhering the adage of burning the clock down during a tight lead (see switch between Q3 and Q4 USC, 1st and 2nd half UW and UNC) and the pass heavier offense during catch-up (2nd half of Oregon, Q4 Arizona, Q2 USC, and most importantly Q4 UCLA).
There isn’t a lot of early game aggressiveness that Cal should exhibit in a modern passing league, this hopefully will change with growing confidence in the air attack.
Cal didn’t break the mold in their binary decision of run vs. pass. Which is both comforting:
- we didn’t fold and go full conservative with a relatively run heavy approach post injury and offensive woes,
- on 1st downs we still adhered to a basic game script and probably left additional yards on the table.
However, there are good reasons why the lack of passing on 1st down and reliance on the run made sense during the season, and it has to do with the distribution of yards gained, something I will explore in part II of this series: “It is the work of an enemy defense.”