Game theory and the spread offense

Over at Inside Texas I have a piece up I’ve been putting together since Big 12 media days with quotes from Tom Herman and Sam Ehlinger that have illuminated the apparent strategy for the Longhorn offense in 2019. In short, the plan is to put a greater emphasis on RPOs this coming season, giving Ehlinger more pass options on their run plays in order to punish opponents for the way they play the run.

As I noted on Twitter:

The goal on offense this year will be to use Ehlinger's expertise and the speed at WR to isolate Keaontay Ingram and the OL on light boxes. Sort of an inversion of what they did last year, holding attention in the box to throw it downfield to Humphrey and Johnson. https://t.co/QqeGbUDYaG

— Asst to the Minister of Culture (@Ian_A_Boyd) July 25, 2019

I realized after writing that tweet that I’d just hinted at a sort of major and largely unexplored truth about the way that spread offenses tend to function.

It was over the course of the 2018 season, writing weekly previews of college football’s top six games over at Football Outsiders, that I realized that both Texas and Iowa State had inverted production relative to their strategies.

Texas ranked 27th in offensive S&P+ and Iowa State was 59th. The Longhorns were 98th in rushing S&P+ and 43rd in passing S&P+. Similarly the Cyclones were 102nd in rushing S&P+ and 21st in passing S&P+.

But if you watched the games you’d notice that both Texas and Iowa State had run-centric approaches to the games, both running the ball regularly and often throwing on RPOs or play-action and both regularly employing big, blocking TEs to aid their approach.

Here’s how Texas defended the Cyclones:

The Longhorns would consistently rotate both safeties to the strong side to get an extra defender in the box, leaving their corners on islands against the big Cyclone wideouts Hakeem Butler and Matthew Eaton in order to ensure they had the numbers to stop the ISU run game. They also put in auto-blitzes like this one to help themselves out when Iowa State would use motion to try and manipulate the matchups. There’s seven guys in or near the box here against six blockers.

The Longhorn structure was unique because of their safety rotations and ability to use multiple overhangs but the idea that the best way to handle Iowa State was to attack their run game and limit their play-action rang true in that game and was evident in the approaches of other teams as well.

Here’s how TCU played Texas:

This was one of the more conservative approaches the Longhorns saw in terms of defenders devoted to run defense. The SS/nickel Innis Gaines is darting around trying to present an unclear read on run vs pass but he’s fitting the run on the edge off the TE’s block to give the Frogs an extra guy. The LBs are hard-charging to the run and while TCU is playing cloud to either side they have both the boundary CB and the weak safety’s eyes on the backfield. Texas couldn’t run the ball well on the Frogs with their base run game and Tre Watson and Keaontay Ingram combined for 23 carries that yielded 96 yards at 4.2 ypc and a score. As was typical for 2018, Texas’ longest runs for their RBs were 14 and nine yards apiece.

But Ehlinger threw 32 passes for 255 yards at 8.0 ypa with two TDs. Collin Johnson chewed up TCU’s frequent deployment of the weak safety to help against the run or Lil’Jordan Humphrey with seven catches for 124 yards and a TD.

Because Texas and Iowa State committed blockers and emphasis to the box to run the ball, opponents did likewise, which opened up opportunities to throw the ball outside. Hakeem Butler, Lil’Jordan Humphrey, and Collin Johnson were all close to or above 1k yards on the year and that was the main thrust of their offenses.

Passing to set up the run

The 2019 Cyclones will probably look like the 2018 Cyclones but moreso, continuing and expanding use of 12 and even 13 personnel sets that put big bodies on the field to command attention in the run game and set up the passing attack for PFPurdy.

In 2019, Texas will have more speed at WR and better experience in Ehlinger but less of a matchup weapon in the passing game without Lil’Jordan Humphrey on the field. However, Keaontay Ingram flashed a lot at RB in 2018 as a 200 pound, oft-injured freshman. In 2019 he’ll be a 220 pound sophomore and backed up by 220 pound, 5-star freshman Jordan Whittington.

The offensive line will also feature preseason All-B12 C Zach Shackleford, likely postseason All-B12 LT Sam Cosmi, transfer All-ACC OG Parker Braun, and then some players they’ve been developing over recent seasons. It looks a lot like the 2014 Ohio State team that was starting over with four new offensive linemen, sliding RT Taylor Decker to LT, and plugging in sophomore RB Zeke Elliott with Carlos Hyde moving on. Whether or not Ingram will be Zeke or if Texas’ interior OL will prove to be as fantastic as the new starters for the Buckeyes (future Rimington award winners at either guard spot) remains to be seen. However, at the very least it looks like the run game could be quite potent in a league that isn’t known for having big, talented defensive fronts.

The question then becomes how to best make Big 12 defenses come to grips with Texas’ size and athleticism in the run game? The answer is RPOs from spread sets. If Texas can execute vertical pass options they can force defenses to hang back and maintain width even if they’re seeing the Longhorn OL firing downhill. With that accomplished they can create 5-on-5 situations for the OL against lighter boxes and buy Ingram or Whittington extra instants before run support DBs are able to reach the box after initially sitting back on pass options.

Clemson utilized a pass-first approach last season with aggressive RPOs and RB Travis Etienne went for 1500+ rushing yards because up until the playoffs opponents couldn’t handle their run game without taking defenders away from the Clemson wideouts, which was too costly.

This game theory approach to offense is basically what propelled the Art Briles machine at Baylor. They’d use wide splits, RPOs, play-action, and tons of deep bombs to force teams to drop everyone back, then you couldn’t stop them in the run game.

If you want the game to be focused on the perimeter, you either need to really flood it with personnel, or you need to divert defensive attention to the box. If you want the game to be focused in the trenches, you either need to flood it with blocking personnel or else you need to divert defensive attention to the perimeter. Smart defenses will always commit numbers where they need to for you to have to beat them “left handed” unless you show the left jab so hard that they have to commit to stopping it and open themselves up for the big right. Texas will be looking to set up their right hand this season and it should yield some fascinating results.