Back in high school, I authored a satirical website about my school's football team. It didn't end well once a
freedom hater concerned student alerted the administration.
One of the running jokes was that a friend of mine, who was not on the football team, was holding out for more money. I joked that he once scored 5 touchdowns on 3 touches to go along with 400-plus yards, an impossible feat. You can't net 400 yards on 3 touches. You can't score more touchdowns than times you touch the ball.
But the more times you touch the ball, the more times you should score. That's just how the world works. And the math backs it up. So we can use that to figure out which running backs should've scored more often in a season than they actually did.
Running the ball isn't efficient. That's just a fact. It's not an efficient way to move the sticks and increase a team's chances of scoring.
According to our Net Expected Points (NEP) metric -- the points a player adds to his team's expected output -- the average running back carry in the NFL actually loses a team expected points (a loss of around 0.02 points per carry in 2017). When you rack up 350 carries of "average" production, you're actually going to lose your team around a touchdown's worth of production.
So rather than rely on expected points, we'll turn to straight up volume to identify touchdown regression candidates for 2018. The relationship between rushing yards and rushing touchdowns is decent on its own, but total touchdowns and total yards actually has a higher R-Squared value.
Photo by Keith Allison