Brian Fremeau is back for his 10th season of college football advanced stats at Football Outsiders, with a lesson on average teams (like Arkansas State) and updated FEI projections for 2016.
27 Aug 2012
by Brian Fremeau
For the past nine years, I have been interested in researching and defining new measures of drive efficiency success. This had led to the development of the Fremeau Efficiency Index (FEI) ratings and other supplemental statistics published here at Football Outsiders. Readers that have followed my work closely know that I’ve been just as interested in finding more effective ways to share data, especially through graphics.
I recently introduced the FEI Profile, a comprehensive visual summary of offensive, defensive, and special teams possession efficiency collected for a given team. I have posted a few FEI Profiles on my blog thus far and plan to continue featuring teams in the format going forward.
As this project continues to develop, I have realized that some additional context is needed. It’s terrific to immerse yourself in hundreds of data points neatly organized, but how does one team compare to another? What is great and what is merely average?
We can answer those questions by investigating recent championship team FEI Profiles against the backdrop of national averages. We'll focus this inquiry on first-down success, new data that I have been collecting and will be focusing on in the year ahead.
Part of the reason I like working with drive data is that it is a relatively easy data set to manage and analyze. Success on a given play can certainly be measured, but down and distance variables make those measures more complex. Success on a particular drive is a bit more straightforward, though a limited drive data set creates more sample size issues.
But what if we treated first-down situations as an independent event? There are many more first-down opportunities than drives, four or five times as many for some teams. The chart above provides a nice comparison. The blue plot, "First Down Series that Result in a Touchdown at End of Drive," is data we've explored before, representing expected success rates on a given drive. The white plot, "First Down Series that Result in a New First Down or Touchdown," is new data I'm diving into this year. Across the field, the average likelihood that a first down series will successfully generate a new first-down opportunity is relatively stable. A team deep in its own territory is less successful due to play calling in that area of the field. Success in the red zone is also impacted by play calling and the challenge of working with a more compact field.
As demonstrated in the already published FEI Profiles, success from starting field position is not as consistent on a team-by-team basis. Part of the reason is sample size, especially near the end zones where first-down situations are less frequent. But this data may also reveal areas of the field where teams struggle to move the ball, or potentially areas where they dominate. Below are first-down success charts for each of the last five BCS championship offenses.
Before running this data, I expected each offense to be consistently above-average across the length of the field. I was surprised to find that several of these championship teams were merely average or below-average in certain situations.
Do any of these individual team profiles provide a road map for championship success? What jumps out to me is that though there are certainly areas of the field in which each team had only modest success, every championship team had elite success in one or two sections of the field. Obviously, first-down success begets traditional success measures like total offense and scoring offense, but it is also important to see how field position advantages can be impacted.
Alabama's 2011 chart is a great example. The Crimson Tide were well above-average deep in their own end of the field; they were able to dig themselves out of poor field position regularly. Alabama started 15 drives last season from at or inside their own 10-yard line. The Crimson Tide gained more than 50 yards on average on those drives, scoring five touchdowns. Between their own 30-yard line and midfield, Alabama was merely average, but drives that stall in that area of the field typically don't put a defense in a bind. Alabama excelled again in moving the ball between midfield and the opponent's 30-yard-line, a critical area of the field to push into field-goal range. Alabama wasn't stellar on long field goal attempts (and almost cost themselves a championship opportunity in a regular season loss to LSU due to that), so execution in this part of the field was especially important.
Combining these five offensive profiles into a summary chart provides the road map we're seeking. Across the field, championship offenses have averaged between 10 and 20 percent more first-down success than an average college football offense.
These five offenses also excelled at limiting mistakes, especially in critical areas of the field. The chart below illustrates the distribution of drive results of every first-down situation across the field for each of the last five national champions. Combined, these teams had 2043 offensive series. They turned the ball over in only 6.1 percent of those situations. (The national average was 13.5 percent last year.) Championship teams were even stingier in the red zone, where those same five teams combined for only three turnovers on 342 series (0.9 percent), a frequency much lower than the national average of 7.5 percent.
It should come as no surprise that protecting the football and consistently generating first downs are important ingredients to fielding a championship football team. But offensive success is only one part of the equation, and the SEC championship teams have made an even bigger impact on defense. We'll look at that later this week.