We’re headed down the home stretch of the college football season which means most of the attention during the next week will be centered on how to assess which four teams in the nation are the best and will ultimately meet in the College Football Playoff. When the Selection Committee makes their announcement on Sunday, there may be questions raised as the national picture is anything but clear and Alabama being great is the only thing everyone can agree on.
While the committee’s work is done behind the scenes, some of the evaluation measures they use are well-known including record (overall and conference), conference championships, strength of schedule, other factors that could have affected performance like weather or injuries, and the all-important “eye test.” What is unknown however is what role numbers play in the evaluation and more specifically, which measures, if any, are used.
There may not be a “silver bullet” mathematical solution that solves the problem of deciding which four teams are deserving of playing for national supremacy, but data can help in making the selection easier.
For my own analysis, I wanted to highlight two statistics in particular that a good job in illuminating the performance of college football teams over the course of an entire season: Yards per play and yards per point. Yards per play simply measures the explosiveness of a particular offense on a play-by-play basis, while yards per point is a tad more nuanced.
Yards per point demonstrates which teams are efficient with their ball movement, transferring yards gained into points on the scoreboard. A team with fewer yards per point is thought to be more efficient for a number of reasons. They could be capitalizing on good field position. It could be a matter of a team converting more of their scoring opportunities. It also could be due to the exemplary performance of a team’s defense and special teams creating scoring chances. For whatever reason, yards per point is a good barometer of a team’s offensive efficiency.
One other underlying component of both of these statistics is the role luck plays in each. Yards per play, by its very nature, is a more stable statistic over the course of a season. It is a straightforward measure of an offense’s ability to generate yards consistently being less prone to fluctuation because it possesses the requisite talent and / or scheme to continue doing so. Yards per point on the other hand includes a number of factors where luck could play a role. One example is the randomness of fumbles and the resulting situation (field position, possession, score) from each.
Each measure on its own doesn’t paint a clear enough picture to accurately assess each team, but combining them can be truly effective in identifying top performers. A team that is explosive and lacks efficiency may be a threat on a per play basis, but may have to work too hard to generate their points by creating fewer opportunities through defense, special teams, or simply not converting enough scoring chances. A team that is efficient, but lacks explosiveness, often finds itself relying on other facets of the game to score points because they don’t pose the same play-to-play threat that others do. The best offenses are thought to possess both of these traits.
Below is a graph representing each NCAA FBS team’s explosiveness (yards/play, x-axis) and efficiency (yards/point, y-axis) for the 2016 regular season (through November 28). The intersecting x- and y-axes represent the national averages of each data set.
Those 10 teams most likely to appear in the College Football Playoff are represented by green dots while all other teams are in red. Not surprisingly, all of the playoff candidates are above average in efficiency while only Colorado and Wisconsin fall below average in explosiveness. Pittsburgh reigns as the most efficient team in the country and Louisiana Tech is the most explosive, but the standard bearers for the playoff participants in those areas are Michigan and Oklahoma, respectively.
Looking at offense alone doesn’t tell the whole story. Both figures can be flipped to the defensive side of the ball to show which teams are best at limiting explosiveness and creating inefficiency. Below is a graph like the offensive one above, however, where the best offenses were found in the bottom right quadrant of the previous graph, the best defenses can be found in the top left.
Not surprisingly, Alabama resides comfortably as the best defense in the country with the best balance of limiting yards per play (less than four) and forcing more yards per point. Alabama is joined by seven of the other ten potential playoff teams in the top left of the chart with Oklahoma State (good efficiency, poor explosiveness) and Oklahoma (poor in both areas) the lone exceptions. It’s interesting that Oklahoma’s defensive unit appears to be the only one of the twenty units amongst the potential playoff group that rates as poor.
After looking at both charts, Louisville seems to have the most unique situation as they had one of the strongest units in the country on both sides of the ball as it relates to explosiveness, but lagged behind with their defensive efficiency.
In order to measure overall team performance and not simply show how strong a team is on each side of the ball, I’ve combined offense and defense to create net yards/play and net yards/point. The difference in this chart is that the top right quadrant is where the best teams should reside. The data here seems to bear that out.
It’s important to note that none of these figures are adjusted for schedule or opponent. Still, the data gives a clear indication of the top four teams based on this measure and it’s not far off base from the national conversation. Alabama resides as far and away the best team with three others right behind: Michigan, Ohio State, and Washington. Clemson appears to be the most viable candidate to sneak in (solely on this chart because right now, if they win, they’re in the playoff) to join those four with the other five lagging behind.
In addition to the playoff contenders, there are other interesting bits of information from this graph that may better explain something that occurred this season or potentially predict something else to come. Louisville had the best net explosiveness in the country but turned out to be barely above average in efficiency. Down the stretch, it was turnovers that allowed their opponents yards per point to drop. One interesting and unexpected team in amongst the national powers is Miami, while Maryland rated as the most “average” team in the country and Rutgers as the worst.