I still remember as a kid wondering why teams didn’t just always run the ball. Running backs average around four yards per carry, and three times four is greater than 10, so you can just march the ball down the field every drive... can’t you?
If only it were that easy.
The problem of course is that you get to that four yards per carry with a whole lot of short runs (in 2018 43 percent of runs went for two yards or less) and an occasional chunk play. Those chunks are a big part of what makes the yards per carry stat largely useless - it can be hugely skewed by a couple long runs. Let’s take probably the best example of 2018: Derrick Henry.
It’s easy to look at the overall stats and tell the story that Henry had an elite 2018 season. He broke 1,000 rushing yards, ran for 4.9 YPC (11th among running backs) and 12 touchdowns. Now, there’s no doubt that Henry had a fantastic final four games of the year (to make up for the lackluster first 12), but let’s focus on that 4.9 yards per carry.
Henry had 216 carries in 2018, one of which went for 99 yards against a Jaguars team that looked like it would rather not bother tackling that day. Take away that one run and his impressive yards per carry stat drops to 4.47 YPC - still good but far from elite.
Meanwhile, 40 percent of his runs went for two yards or fewer and another 14 percent were 3-yard carries but each of those have a negligible effect on yards per carry compared to the long runs. Good season, sure but the issue with focusing on yards per carry is that it’s very sensitive to those few outlier plays. “Success Rate” tries to fix that issue.
The original incarnation of success rate was a little cumbersome. There were different criteria for success depending on the down (on third down only a play that gets a first down was successful for instance) and those criteria were somewhat arbitrary.
More recently an expected points-based approach has become more common. Basically, if a team is better off after a play than before it (ie: the play had positive EPA) than that play was a success. If not, it wasn’t. Outliers are effectively ignored, making a 99-yard touchdown run on first and 10 is a success just like an 8-yard-run would be in the same spot. By success rate for running backs, Henry sits just below the NFL average and right in the middle of the pack at just under 40 percent. Not a terrible season but hold off on putting him among the elite backs just yet.
The leaders in success rate? Aaron Jones (on a somewhat smaller sample size than most), Todd Gurley, Philip Lindsay, Marlon Mack and Sony Michel are the Top 5 of 2018, with James Conner, Christian McCaffery, Nick Chubb, Austin Ekeler (again smaller sample) and Alvin Kamara rounding out the Top 10.
Quick aside: success percentage backs the continued rise of the passing game in the NFL. Passing plays average more yards and were successful more than 5 percent more often than runs in 2018.
In the chart above we see that some running backs (like Ekeler and Isaiah Crowell) had yards per carry numbers that far outstripped their success rates. Their yards per carry are skewed by a few long runs, while others like Jordan Howard had an acceptable success rate (better than Saquon Barkley) but low yards per carry due to a lack of big plays.
Big plays tend to be somewhat random. Leonard Fournette broke a few in 2017 to post a respectable YPC, but topped out at 25 yards in 2018 and his YPC reflects that. Lastly, we’re left wondering why LeGarrette Blount is still an NFL running back. But I digress.
So, success rate can be useful because it takes some of the noise out of very noisy data, as per carry stats can be skewed by just one or two plays. Now let’s see where success rate may help us find an edge for the NFL Divisional Round.
The Colts defense stymied Deshaun Watson & Co. but face a much tougher test this week. The Chiefs finished the regular season ranked third in success rate running the ball and second in passing. Meanwhile, while the Colts stop unit was solid against the run, they ranked second from the bottom in success rate allowed versus the pass. Look for Pat Mahomes to have another big day and Kansas City to exorcize some of its playoff demons.
The Dallas defense gets a lot of credit, and rightfully so. But they go up against the NFL leaders in offensive success rate in L.A. this weekend. The Cowboys have been particularly stout against the run (and were again against the stubborn Seahawks, who seemed intent on running no matter how ineffectively) but rank 22nd in the NFL in success rate allowed against the pass.
Jared Goff had his struggles down the stretch, particularly in the cold weather, but the Rams still sit tied for fourth in passing success rate and Goff should avoid those issues back home in Southern California. The Cowboys offense has been able to put up points in spurts but faces a Rams defense that sits 12th in success rate and has improved since Aqib Talib’s return.