Simple vs. Complex Offenses, Part 1

Illinois offensive coordinator Bill Cubit developed a bit of a reputation last year from his "bag of tricks", the collection of trick plays that by some accounts were responsible for revitalizing the Illinois offense. I love to watch a well-executed swinging gate just as much as the next person, but I’ve always been a little skeptical of the touted advantages of having complex playbooks.

(Remember the last person making bold claims about "decided schematic advantages"?)

Sure, they’re fun to watch when they work, but asking 19-year olds to memorize 300 pages of your glorious tactical vision is always a tall order. And how do you contrast that with past offensive stalwarts like Texas Tech, who didn't even have a playbook? Or Chip Kelly’s Oregon teams, who had all of three running plays? They’re on the opposite end of the spectrum, espousing well-executed simplicity over tactics.

Kelly would be the first to say that you can be good at football by choosing either, but if you’re designing an offense from scratch, that doesn’t answer which direction you should go. What generally works better, a complex offense that keeps the defense on its heels, or a simple offense that can be boringly practiced to near-perfection? Using the data from the 2013 charting project, we can begin to answer that question.

First things first: let’s try to get some understanding of the relative complexities of offenses. A good place to start is seeing the range of formations in which offenses line up. Formations are defined by three attributes: number of backs, number of wide receivers, and where the quarterback lines up (under center, pistol, shotgun). The basic assumption is the greater diversity in the number of formations, the more complex the playbook. Sure, there are some exceptions -- a standard 1RB, 4WR shotgun lineup gets put in the same bucket as an Emory-and-Henry formation, for example -- but I think it still works for assessing general complexity.

I took the charting data and ran it through the standard garbage-time filter S&P uses. Then, I whittled the data set down to teams with at least five or more games charted to restrict the analysis to offenses with legitimate sample sizes. Here’s a visualization of the range of formations each offense ran in 2013 (click each team to see their graph):

You can see some distinct styles from these graphs. LSU and Oregon State’s predominantly blue/green circles indicate a pro-style offense; Ohio State and Oklahoma have embraced the pistol probably the most of all teams. Some teams have next to no variety in their formations (LOL Washington State), some have a little bit of everything (Northwestern has a circle in just about every possible spot). And others, like Georgia Tech, are just plain different from everyone else.

I think it’s important to start with visualizations like these when assessing the complexity of offensive play calling. There are plenty of ways to quantify these differences with some pretty hairy math -- and don’t worry, we’ll get to those too -- but explaining the formulas doesn’t give nearly the intuitive explanations that just looking at the data does when presented the right way.

In the next post, we’ll take it a step further and try to quantify these graphs to a single complexity metric. That will allow us to start making correlations to offensive output numbers and start to answer whether simple or complex is better.