Great stuff in this thread---it's my daily required reading. Fantastic insight on players from some really knowledgeable people.
I have a question about Jonathan Stewart vs. DeAngelo Williams. From what I've read in this thread and seen in F&L's rankings, Stewart is looked upon very highly, despite being a backup. He's definitely produced a fair amount over his first 2 years in the league, particularly with scoring TDs, but something's been bothering me about his performance.
Last year, he averaged 4.5 ypc, and this year he's increased that to 4.8 ypc. Deangelo, behind the same line, averaged about 5.5 ypc last year, down to 5.1 ypc this year. Given that Deangelo's the workhorse back and sees more carries, shouldn't his ypc be lower? I know on F&L's rankings, Stewart is ranked 13th to DWill's 7th, but is that due to a strict talent comparison? I know that you can't put everything on ypc, but I feel that there's a pretty strong correlation between ypc and general talent/production in FF.
I guess what I'm asking is a little insight as to the talent level of Stewart. If he got the chance to be a feature back, who would he be comparable to? Steven Jackson? Ray Rice? Ronnie Brown or Pierre Thomas? I'm talking about pure talent level, all else being equal. I'm assuming that most people think that DWill is the more talented back, if they were the same age.
Thanks in advance for any insight.
It's hard to do a straight ypc comparison, because ypc isn't just team and player dependent, it's also SITUATION dependent.Imagine two hypothetical RBs. One RB averages 2 yards per carry on 3rd and 1, and averages 5 yards per carry on 1st and 10. The second RB averages 0 yards per carry on 3rd and 1, and 4 yards per carry on 1st and 10. The first RB has a better ypc in both situations, but if the second RB gets all of the first down carries and the first RB gets all of the third down carries, then the second RB will actually wind up with the higher overall ypc! This phenomenon (that a player can have a higher ypc in every single situation, but a lower overall ypc) is called "Simpson's Paradox", and is a very real effect that depresses the overall ypc of any RB who gets an inordinate about of red zone or short yardage carries.
In addition to Simpson's Paradox, the ypc statistic always favors big-play RBs over consistent chain-movers. For instance, looking back at Denver, Tatum Bell always posted a sterling ypc (the best on the team in every season, iirc), despite the fact that he was viewed as mediocre and expendable. The reason was because he'd pad his total with a handful of huge runs. A guy who has 9 runs for 2 yards and a run for 32 yards is averaging 5.0 ypc, but a guy who runs for 4 yards half the time and 5 yards the other half is significantly more valuable.
The final variable at play is where the runs are taking place. If a player only ever got the ball at the 1 yard line, the highest ypc he could possibly post would be 1.0 ypc. The closer to the end zone the runs are coming, the less potential there is for a long run that brings up the average ypc- after all, it's impossible to break an 80 yard run from midfield.
Anyway, to see what's at play in Carolina, I did a bit of digging into splits. The data suggests that Stewart converts a higher percentage of short yardage carries, while Williams has a higher percentage of long runs, which seems to fit with my subjective take on their respective skills. Stewart's probably never going to be the kind of guy who can put up a ypc in the same ballpark as Williams', just like Steven Jackson is never going to be a 5+ ypc kind of guy (even if his unbelievable 2006, he only averaged 4.4 ypc). To be honest, though, Stewart doesn't have to put up a 5+ ypc in order to be a fantasy stud.
In terms of physical attributes, the best comparisons for Stewart are Ronnie Brown and Steven Jackson. Those two are pretty much the only guys who have entered the NFL this decade with a comparable package of size and speed, although Stewart's size/speed combo is actually better than either of theirs. Williams is a more dynamic talent and a bigger threat in the open field, but Stewart is better suited to be a bellcow like Peterson, Jackson, or MJD. I really think that Williams benefits from Stewart's presence in a lot of ways, and that he wouldn't be doing as well without Stewart around. It's hard to say which RB is the more talented of the two, because they're very different backs, but I do think that both RBs are among the 5-8 most talented RBs in the league today.