In the last installment of this series, we looked at Pittsburgh Steelers quarterback Ben Roethlisberger. He was a great choice to use for some simple projections, because of the large sample size and mostly consistent improvement, year on year.
Running back Le’Veon Bell, on the other hand, is an odd case. On the one hand, his is a very small sample size — just four seasons. On the other hand...well...take a look.
We saw this chart in the series intro. The jagged, blue line represents Bell’s actual rushing yardage total, from one year to the next. The straight, blue line is the linear regression trend line.
The red lines, though, are remarkable. The dark red line is his normalized yards -- the yardage he would have gained over a full 16 games, based on his per-game average. It nearly perfectly flows along the trend line, showing an almost immaculately steady improvement over four years. So, while the sample size is exceedingly low, it’s also been exceedingly dependable. Based on that, it stands to reason that we could create some fairly accurate projections.
The first, and most intriguing, thing we find when we apply normalization and linear regression to Bell’s key numbers is that his yardage projection for 2017 is pushing up against NFL record territory. Eric Dickerson holds the record of 2,105 yards, and six others have bumped past the 2,000-yard mark. Bell’s numbers for 2017 would project to be just a shade under that, at 1,904 yards. With a standard error of 54 yards, that puts him in a range of anywhere from 1,850 to 1,958.
With numbers like that, it’s not entirely difficult to see why Bell turned down a reported offer of more than $12 million per year on July 17th, risking instead that he will have a season for the ages in 2017.
Bell is certainly a workhorse. When normalized for 16 games, his carries bottom out at 290 in any season. In 2016, his 261 carries in 12 games normalized to 348 in 16 games, and his 2017 number would be almost identical, at 349.
Scoring, though, is where Bell’s argument for more money might fall apart, as he has never scored more than eight in a season on the ground. That normalizes to 10, as a rookie. His trendline for touchdowns is actually a downward slope, and that is reflected in 2017 with eight touchdowns in a 16-game season.
Bell’s greatest value, though, may not be as a runner. It is likely his versatility, which allows him to run, catch out of the backfield, and even line up as a legitimate number-two or two-A receiver. He’s carried a massive workload in the receiving game, and was on a 100-reception pace in 2016, had he played the entire schedule. That projects out to 105 catches in 2017.
Those 105 receptions would net Bell some serious yardage, although this is, like his touchdowns, a downward trend. In 2017, under this basic statistical model, he projects to amass 757 yards, good for 7.2 yards per catch. That average is a full yard below his 2016 mark of 8.2 yards per catch.
Finally, Bell would likely pull in just three touchdown receptions. This isn’t at all out of line with his past history, as he only has five career receiving touchdowns: three in 2014, and two in 2016.
Bell, like Roethlisberger, has missed a good bit of time. In Bell’s case, he has missed 17 games in four seasons, or more than a full season’s worth of games. Five of those were the result of two suspensions, which we will normalize out, like we did with Roethlisberger. Based on that, he has missed 12 games total. The projections say he will miss three in 2017, due largely to the massive outlier in 2015 when he missed eight games due to a season-ending knee injury. This is a situation where small sample size really hurts, but it’s all we have to go on.
Taking all of these projections, and three missed games, here are the projections for Bell in 2017: