March 18, 2019

# Yards Still Matter

If you spend any time on this site you know that one of my favorite college football metrics is also one of the most straightforward – total yards.  With the exception of yards per play, and it’s close, you’d be hard pressed to find a metric that correlates more to winning than total yardage differential between two teams.  Why?  Because yards = points.

Every time I hear someone say yards don’t matter – and provide an example of a game where a team with less yards won – my head wants to explode.  Yes, there are exceptions to the rule.  Just like there are exceptions to “You have to win the turnover battle” or “You have to run the ball to win”.  For every example you provide of a team winning despite having less total yards, I can provide more where a team won the turnover battle and lost.

If you made it through freshman stats then the graph below should tell you a story.  I’ve plotted every college football game from 2011 through 2015 (D1 vs. D1 only, 3,580 games and 7,160 data points) with yards and points.  Notice the slope of the line that goes through the data points.  It’s pointing up, as in more yards means more points.

This is significant for Clemson because one of the narratives of the offseason is that the Tigers defense lost a lot of talent and will have to “rebuild”. Fair enough.

In the period covered, teams that reached the magic 500 yard mark won 79% of the time, without regard to any other metric.  Clemson has reeled off 10 straight (and counting) 500 yard games.  Just by the fact that your offense is gaining 500 yards means you are highly likely to win.

But it gets better.

The Clemson defense gave up 313.0 yards per game in 2015.  For arguments sake, lets say the Tigers 2016 defense regresses to “average”, which in NCAA terms in 2015 was 400 yards per game.  That’s an additional 87 yards per game given up (27.7% increase).

In our mythical game Clemson gets 500 (or more) yards and gives up 400 yards (regressed to average).  What are the chances the Tigers win?  Over the same time period (2011-2015) teams with this profile won 94% of the time (671-43).

To recap, the Clemson offense is likely going to be so good that the Tigers can absorb a defensive regression to “average” and still have a high probability of winning.

Obviously, there are no guarantees and every game is an independent data point on a graph such as the one above. The Tigers may reach 400 yards in one game (reducing the odds of winning) and 600 (very high probability of winning) in the next.

The point is an offense that’s likely to reach 500 yards in any given game and give up an “average” amount in the same game is still likely to win and is also why our early win probabilities have the Tigers favored in all 12 games.

As we saw in the championship game 500 yards of offense a game is not a 100% guarantee of a win, but 500 yards of offense combined with holding your opponent to 400 or less is about as close as it gets.