2014 Stats Update

I still hope to post statistics for the 2014 season. The decision to add rushing and passing statistics has complicated my compilation process because it made me dependent on a database I no longer have access to.

I’m still contemplating how to overcome this obstacle. Your patience is appreciated.

Quick Post

FBS Drive Stats has now been updated to include the following things:

  • Drive stats now go back to 2005. The 2005 season interested me for fairly obvious reasons. Some people may have forgotten that before losing to Texas in the Rose Bowl, many in the national media considered the 2005 Trojans as one of the premier teams in college football history. Evidence suggests that Texas clearly had the best team that season. The Adjusted Efficiency Rankings ranked USC third that season, and it ranked Ohio State second. Some outside Columbus and Austin may have forgotten this, but the Buckeyes and Longhorns played that season. Vince Young threw a game-winning touchdown to Limas Sweed with less than three minutes remaining. The Longhorns wont the game 25-22. Ohio State’s other loss that season came against an exceptional Penn State team in Happy Valley. USC probably had the better team than Ohio State that season, but it’s not totally unfathomable to think the Buckeyes were their equal.
  • I have received some emails from people asking me about how to manipulate team drive charts. I think I have solved this problem by installing pivot tables into the drive charts for all 128 teams. This simplifies filtering and should make it easy to generate drive-based statistics in games against an opponent or set of opponents.
  • The Game Summaries files also have pivot tables. Games can be filtered based on conference affiliation or team. This will be particularly useful for those interested in analyzing inter-conference matchups. (SEC fans have yet another way to intelligently prove their league’s superiority to everyone else.) The time filter can be used to assess how a team or conference performed during a month or set of weeks.

That’s all I have for right now. As always, please free to email with any questions. I respond promptly to all messages.

Rushing and Passing Stats Added

New statistics that measure rushing and passing statistics have been added to FBS Drive Stats. Here is a brief overview of the new statistics:

  • Rushing yards per drive (RYds/Dr)
  • Passing yards per drive (PYds/Dr)
  • Rushing yards per play (RYds/Play)
  • Passing yards per play (PYds/Play)
  • Rushing yards percentage (RYds%)
  • Passing yards percentage (PYds%)

This should make it easier to determine strengths, weaknesses, and style of play. (Triple-option offenses finish high in percentage of available yards gained rushing. On the other end of the spectrum, Mike Leach really does not care about running.) Yards per game may be one of the most useless and misleading statistics in sports, but I have yet to see an alternative for rushing and passing statistics that’s easy to understand, measure, and apply.

I will also mention that blog posts prior to April of this year will not be edited to account for these changes. Editing all these blog posts simply represents too much work at this point. Here are the revised Adjusted Efficiency Rankings for the 2013 season:

A few other things ought to be mentioned:

  • This site caters to a very specific type of football fan who possesses an intense interest in football and the math behind it. Even among this group of people, I don’t want to make navigating this site or a particular document unnecessarily cumbersome.
  • After deliberating on how to present field position statistics, I determined that the best way to share that information would be to focus solely on drives beginning in opponent territory and on the offense’s side of the field.
  • Additionally, fumble percentage and interception percentage have been eliminated because turnover percentage and effective turnover percentage exhibited higher correlations with net points per drive.
  • The team drive charts have been changed to make them easier to navigate and manipulate.

I will edit the “Tips for Analysis” page at some point this week.

This project would not have been possible without the fantastic work of Marty at CFBStats.com.

Five-Year Rankings

The most recent version of the five-year Adjusted Efficiency Rankings can be viewed below:


  • The standings for each league now contain each team’s record against conference foes.
  • The conference rankings now show each league’s record against non-conference opponents. This information needs to be shared because the limited sample of games determine national perception of each league and each team’s strength of schedule. (Every league will go .500 in games they play against themselves. Even the SEC.) However, conference realignment has altered the membership of each league over the past five seasons. No league will have a membership with a .500 record shown in their conference standings.
  • A measurement for luck has now been included in the Adjusted Efficiency Rankings as well. Luck in football can be measured based off of turnovers (particularly fumble recoveries), injuries, and the ability to win close games. FBS Drive Stats will measure luck by subtracting a team’s expected winning percentage from its winning percentage. Expected winning percentage will use a team’s offensive and defensive points per drive to determine what percentage of games they should have won based on their production. For example, BYU went 7-5 in games against FBS opponents. They scored 1.92 points per drive and allowed 1.49 points per drive. This gave the Cougars a winning percentage of .5833 and an expected winning percentage of .6368. This means they won 5.35% fewer games than their stats indicated that they should have won last year. Had they won another game with the same output, they would have been seen as a team with good fortune.
  • This “luck” measurement does not attempt to gauge how many close games a team won, but teams who are lucky win most of their close games. Oklahoma surprisingly finished first in this category last season, but they only went 2-0 in games (against TCU and Texas Tech) with a scoring margin of eight points or less. This happened because the Sooners went 8-0 in games with a margin of less than 1 point per drive. They beat Oklahoma State by nine when the Cowboys slightly outscored the Sooners in terms of points per drive. (OU scored two non-offensive touchdowns on a Jalen Saunders punt return and an Eric Stryker fumble recovery to end the game.)

The Best Teams, the “Luckiest Teams,” and the Toughest Schedules (2009-2013)

The newest Viz portrays each team’s performance over the past five seasons by converting the statistics shown in the Adjusted Efficiency Rankings to z-scores. For those that may have forgotten their elementary statistics course, z-scores measure how many standard deviations a data point is from the mean. Click on the picture below to access the interactive graph:

09 to 13 Rankings

Drive-Based Analysis of the Five Major Conferences

Click on the picture below to access the interactive visual:

Big 5 Viz

The linked chart contains conference drive stats for each team in the ACC, Big 12, Big Ten, Pac-12, and SEC over the past five seasons. Here is an overview of what it contains:

  • The first sheet contains each team’s conference record during the past five seasons. Below that, it contains a bubble chart that lists conference champions in the five power leagues. The 2010 Big Ten (Wisconsin, Ohio State, and Michigan State) and the 2012 Big 12 (Kansas State and Oklahoma) each had multiple conference champions. Consequently, the championships in those leagues do not add up to five.
  • The standings are organized by overall record, home record, away record, and neutral record.
  • The remaining sheets contain highlight tables that showcase how each team performed in certain statistical categories. These categories include starting field position, yards per play, percentage of available yards gained, effective turnover percentage, touchdown percentage, and points per drive.

The reason I’m focusing on the five major conferences has to do with the college football playoff. Teams that do not win their conference championship are not technically precluded from making the national title, but I anticipate that this rule will almost exclusively benefit the SEC. (Recent history suggests the Pac-10 in 2010 and the Big 12 in 2008 were most likely aberrations.) Winning a championship in the ACC, Big 12, Big Ten, and Pac-12 will be essential for serious postseason consideration in most seasons.