Do Returning Offensive Line Starts Matter?

With a couple plates and a handful of screws the 2016 season of starting junior tackle Jake Campos is likely over before it even begins. With Campos’ injury disappears 24 starts over two years and only one other offensive lineman — Nick Fett — has started a game in a Cyclone uniform.

As we wrote yesterday, there are schematic ways to combat the lack of experienced linemen and there is no doubt that head coach Matt Campbell and his staff are working on this very thing through the rest of fall camp.

Yet I kept coming back to the fact that Toledo had zero returning offensive linemen who started games prior to the 2015 season and they allowed five sacks the entire season. Surely Tom Manning’s magic has bounds, so something else must be afoot.

The Data

Phil Steele, he of magnificent preview magazine fame, does a fantastic job keeping track of returning starters and as a component tracks returning offensive lineman starts. If a team returns all five linemen from a season where they all started 12 games, that would be a combined 60 returning starts.

Typically a number of teams will return a season’s worth of starts with the 2015 average for returning offensive line starts coming in at 65.5. Certain teams, such as Bowling Green and North Carolina last year, returned 100+ starts (125 & 105, respectively). Those two teams went a combined 21-7 in 2015 and clocked in at 8th (North Carolina) and 31st (Bowling Green) in CFB Analytics’ 2015 Ratings. They also both fielded Top 10 offenses in CFA’s Ratings, with North Carolina’s Adjusted Offensive Points per Possession of 3.28 being tops in the country.

It’s natural to look at those stats and think that more starts immediately equates to more points which equates to more wins. And, of course, you’d be wrong.

SMU returned 104 starts in 2015 and went 2-10 in Chad Morris’ first year at the helm. Indiana and Ball State, two other schools with 100+ starts returning, finished a combined 9-16 and failed to crack the CFA Top 60.

The Model

Seeing this trend, I decided to put together a quick linear regression model to see how much influence returning offensive line starts alone will have on a team’s win percentage. I grabbed Steele’s data from 2011 through 2015 and went to work. The results were quick and not at all severe. Below are the adjusted r-squared amounts for the three combinations that were run.

Career Starts & Wins: .05138

Career Starts & CFA’s AdjO: .1002

Career Starts & CFA’s Pyth: .0461

Do your basic high school math on those amounts and you get a range of 4.6% to 10% that is explained by a given single variable model. The word “weak” doesn’t even begin to describe this correlation and any sensible modeler would either throw this variable out immediately or focus on building something more complex.

What Matters If Returning OL Starts Doesn’t?

Good question, and this just goes to show the intricacies of the game that have yet to translate themselves statistically like in baseball, basketball, and even hockey.

For one, teams don’t play just five linemen each game. Most will rotate upwards of eight or nine, and as Ben Bruns stated on KXNO Monday, offensive line coaches are always looking for their “top 8” guys to get on the field in a game. In reality, it’s not returning starts that matter so much as returning snaps. But there’s no way any journalist or blogger that wants to have a life could chart that for all 128 teams in a season.

Second, the data doesn’t account for whether or not these linemen stay healthy. Iowa State could easily have walked into the season with Campos’ 24 returning starts and had him drop on the first snap against UNI. Yet that won’t show up in this data as Steele compiles this information mid-summer before fall camp opens.

Lastly, the model I ran was quick and dirty and didn’t account for prior season success (or lack thereof). A more complex model would evaluate a team’s prior season statistics and account for an expected lift from returning experience. That said, I had already done something like this with returning starters in order to do our 2016 Projected Ratings over at CFA, and the results were only slightly more promising than this analysis. It’s unlikely that a combination of returning starters with returning offensive line starts would yield a compelling and predictive model.

That’s Cool and All, but What About the Cyclones?

A lot of fanfare has been given to how Tom Manning took a Toledo line with zero returning experience and turned it into one of the best in the country, but the man himself credits it to the system more than anything else.

Unfortunately that system has been in place in Ames for less than nine months, and expecting a line to succeed while being patched together with a walk-on, possibly a true freshman, and some other guys is asking for too much.

Last year’s teams that returned less than 30 offensive line starts included the aforementioned Toledo, Florida State, Florida, Louisville, Miami (FL), and Oklahoma. That’s good, but the bad is that list also includes Boston College, Kansas, UT-San Antonio, and UL-Monroe; all teams who won three games or less in 2015.

As for the Cyclones themselves, they don’t have a lot of success in this area dating back to 2011.

2011: 54 returning starts, 6-7 record, Pinstripe Bowl loss

2012: 53 returning starts, 6-7 record, Liberty Bowl loss

2013: 70 returning starts, 3-9 record, most of the offensive staff fired

2014: 87 returning starts, 2-10 record, I don’t even know anymore

2015: 76 returning starts, 3-9 record, Rhoads packs up his hard hat

This is mind numbingly terrible. Not a single Cyclone offense in that time cracked an AdjO rating of 2.0, which would still be below the FBS average. As I stated above, this also doesn’t account for injuries, of which there were plenty in this five year timeframe.

It’s hard to make sense of where this will all go, but there are two things I know from this analysis: the season isn’t over in August, and Tom Manning has his work cut out for him.