You are here
Mon., Dec. 18, 2017 6:00 PM to 7:00 PM CST
Tue., Dec. 19, 2017 2:30 PM to 3:00 PM CST
RTN: Statistical Proof Why Running Game Will Improve
Last year, Vikings running back Adrian Peterson turned in one of the greatest seasons in NFL history, averaging 131.1 rushing yards per game. Undoubtedly the premiere running back in the league, Peterson was aided by the Vikings’ defense, which kept the team in games and allowed AP to average 22.9 carries over the final 11 contests, and Minnesota’s willingness to give Peterson the ball in situations when they could have passed it.
The Vikings ended the season as the league’s top rushing team in terms of yards per carry (YPC), due primarily of course to Peterson’s 6.0 YPC. However, there’s pretty good reason to believe that the Vikings weren’t all that great of a rushing team, and definitely not the top team in the league, despite their other-worldly running back. Could it really be that Minnesota’s rushing game was actually only slightly above average?
A Skewed Stat
Earlier this week, I predicted that the Cowboys wi
ll have a substantially better running game in terms of YPC in 2013. The reason is that, although they’re hardly dominant on the ground, Dallas isn’t quite as poor as most think. They ended the season 31st with just 3.6 YPC, but the Cowboys didn’t run the ball as often as some other teams in high-upside situations, such as first-and-10. They actually passed the ball 58.3 percent of the time on first-and-10, compared to the league average of just 47.9 percent.
On down-and-distances such as first-and-10, offenses have an easier time running the ball. All NFL teams collectively averaged 4.41 YPC on first-and-10 last year, but only 4.26 YPC on all runs in any situation. On short-yardage runs with either one or two yards to go, offenses checked in at just 2.97 YPC.
So you can see that YPC is heavily influenced by the situations in which teams run. The Vikings were one of those that kept it on the ground in high-upside situations, allowing Peterson to rack up lots of yards and great efficiency. Other teams, like the Cowboys, don’t necessarily run the ball as much when there are more than a few yards to go for a first down.
So YPC is skewed based on game situations. But a different metric, called “success rate,” accounts for those situations. Success rate is the percentage of plays during which an offense increases the points they can expect to score. Whereas YPC downgrades a team for a 2-yard run on fourth-and-1, success rate rewards it because such a play would increase the probability of scoring. It better captures true rushing ability, and it doesn’t penalize teams for (astutely) passing the ball often in situations like first-and-10.
When we think about rushing effectiveness in terms of success rate, it can really change our perception of the game. Remember, the goal for an offense isn’t to maximize their YPC at all costs. It’s to utilize their running game to help them score touchdowns. When all is said and done, success rate measures an offense’s ability to translate the running game into points, whereas YPC doesn’t always do that. We can’t keep penalizing offenses for gaining first downs in short-yardage situations.
Around the League
To determine which teams are better (and worse) than the traditional numbers suggest, I compared their rank in YPC to their rank in rushing success rate. Below, each team is sorted by the difference between the ranks. Those at the top ranked much higher in success rate, meaning they were most likely to have had a running game that’s superior to what their YPC suggested. Each team’s success rate was calculated by Advanced NFL Stats. Read
In 2012, the Cowboys had success – meaning they increased their expected points – on 40.2 percent of their runs. That mark isn’t outstanding, but it was good enough to rank them 18th in the NFL, 13 spots higher than their YPC rank. That’s the third-largest positive difference in the league.
Those numbers suggest that the Cowboys really did pass the ball in most high-upside situations, as they should, saving their runs for times when they weren’t likely to gain a lot of yards. While that obviously makes it difficult for them to have a high YPC, it helps to maximize yards per play, which is much more important.
Looking to 2013
When we think about the running game, we shouldn’t consider what a rushing play does for an offense, but rather what they lose in calling it. In short-yardage situations, you might lose the ability to secure a big play through the air. But most of the time, that’s OK, as the goal is to move the chains and ultimately increase the probability of scoring.
However, in situations such as first downs, offenses can lose the ability to move down the field in a hurry if they run the ball too often. Continual runs of five and six yards are awesome, but sometimes you get stuffed for one yard or no gain. The ultimate result might be a high YPC, but you also lose the ability to 1) consistently generate big plays, 2) maximize yards per play, and 3) score as many points as possible. We don’t want rushing efficiency at all costs; we want scoring efficiency at all costs.
We all want the Cowboys to gain as many rushing yards as possible on any given play, but the goal shouldn’t be to call plays in an effort to maximize YPC. If you’re running an offense for the sake of helping the running game at all costs, you can lose sight of the bigger picture: You need to get the ball into the end zone. The running game can help with that, but it shouldn’t be implemented in such a way that decreases your overall offensive efficiency.
The Cowboys will have an improved rushing attack in 2013 no matter how they use it, but they don’t necessarily need to have the highest YPC. You want the greatest efficiency you can get, obviously, and you’d hope they can improve upon 3.6 YPC, but only if they aren’t sacrificing the ability to score. If we see Dallas end up in the middle of the pack in YPC without substantially increasing the rate of first down runs or decreasing the number of short-yardage runs, then that’s a major win. Read