Wisdom of Crowds in Review: RB
by Bill Barnwell
On Monday, I took the predictions provided by my Twitter followers before the 2009 season and saw how they related to the actual performance of quarterbacks.
Today, I'll be doing the same thing for running backs. Again, remember that these predictions assumed that each player would make it through a full 16-game season without missing any time due to injuries. The panel was asked to predict strictly the rushing statistics for each player. You can read the three original articles on this topic here, here, and here.
Predicted: 226 carries, 1004 rushing yards, 7 TD
Actual: 247 carries, 947 rushing yards, 7 TD
Moreno missed time at the beginning of the season with a sprained MCL, but he made it into all 16 games. Those numbers look pretty similar at first glance, but the crowd predicted he'd average 4.4 yards per carry, and he was actually at 3.8. I was very high on Moreno coming into the season -- higher than that prediction -- and I'll probably feel the same way going into next year. That's just too good of an offensive line in front of him.
Predicted: 281 carries, 1267 rushing yards, 13 TD
Actual: 313 carries, 1391 rushing yards, 15 TD
KUBIAK's darling, MJD actually outscored Adrian Peterson through the first 16 weeks of the season, only to see Peterson to pull ahead in the final game of the year with a nice day against a disinterested Giants defense. Jones-Drew actually beat the predicted projection by 10-15 percent across the board, and the true "expected" prediction from the average fan would probably be even lower, since we didn't allow people to assume a possibility of injury.
Predicted: 299 carries, 1304 rushing yards, 10 TD
Actual: 258 carries, 929 rushing yards, 4 TD
As a disappointed Matt Forte fantasy owner myself, I don't exactly want to re-visit that actual line. (Especially considering I saw the predicted projection and said that the TD total was low.) It was an awful year, but I still have confidence in Forte; his offensive line in 2010 can't possibly be as bad as it was last year. That touchdown total is obviously going to rise, too. Forte had 19 carries inside the opposition's five-yard line last year, and he scored as many times as he fumbled: twice. An average back behind an average line, given the same opportunities, would've scored 7.7 touchdowns. That difference of 5.7 touchdowns is remarkable; the last back with such a difference is the 2007 version of Thomas Jones, who saw a dramatic upswing the following season.
Predicted: 205 carries, 876 rushing yards, 6 TD
Actual: 104 carries, 357 rushing yards, 1 TD
I was Darren McFadden's biggest cheerleader before the season. For all that Oakland does wrong, they do have a reasonably solid run-blocking line, and while McFadden had split time and struggled with a turf toe injury as a rookie, 2009 would be his chance to play the featured back and have a breakout year. I was wrong.
Marion Barber III
Predicted: 228 carries, 1029 rushing yards, 10 TD
Actual: 214 carries, 932 rushing yards, 7 TD
Barber's predicted touchdown total was high because of the perception that he has a "Nose for the End Zone", but no such thing exists. He's played significantly worse since signing his long-term extension, but it's not like he's obviously dogging it. It could be the expansion of his role; Barber dislocated his toe last year, and he pulled his quadriceps and broke his thumb this year. You can probably make a case for any of the three Dallas backs being the best of the bunch at this point, which is pretty remarkable. I had him at 260/1042/8, thinking that the team wouldn't throw the ball as frequently as they ended up doing.
Predicted: 279 carries, 1181 rushing yards, 12 TD
Actual: 223 carries, 730 rushing yards, 12 TD
The year started with a sprained ankle and stayed pretty miserable. That touchdown total is a product of his usage in short-yardage; he got 28 carries inside the five, producing nine of his 12 rushing touchdowns, but he was still 2.6 touchdowns below what you would expect from an average back in the same situations. Of course, the San Diego offensive line was injury-riddled and struggled last year. Teams seem to be looking towards Tomlinson as a Correll Buckhalter-type backup this offseason, but I think he's probably best as a short-yardage back.
Predicted: 244 carries, 977 rushing yards, 6 TD
Actual: 98 carries, 398 rushing yards, 0 TD
The panel was within .06 yards of his actual per-carry average, so since readers weren't allowed to predict injury, I think that's a pretty solid job. Parker, of course, lost his job to Rashard Mendenhall and is wandering the streets as a free agent this offseason. He was linked to Tampa Bay in a rumor that made no sense whatsoever, and then Thomas Jones signed with Kansas City before Parker could visit. There's not a significant role out there for him, outside of maybe Detroit or Seattle.
Predicted: 270 carries, 1350 rushing yards, 11 TD
Actual: 358 carries, 2006 rushing yards, 14 TD
While we didn't request Adrian Peterson or Michael Turner predictions, the esteemed followers that made up the prediction panel suggested Chris Johnson would have more rushing yards than any other back. I don't recall seeing anyone predict that Johnson would lead the league in rushing yards before the season, but nearly 360 carries later, that's exactly what he did. I didn't believe 1350 was attainable, figuring that Johnson would split more of his time with LenDale White, but White was essentially a non-factor.
Predicted: 268 carries, 1247 rushing yards, 10 TD
Actual: 216 carries, 1117 rushing yards, 7 TD
Anyone could have seen that rushing TD total declining, but falling off all the way to seven was a surprise. If you round up from 4.97, Williams has averaged five yards per carry in three consecutive seasons; in the history of the NFL, only four backs have ever done that while carrying the ball at least 100 times in each of those seasons: James Brooks, Marshall Faulk, and the mid-60s Browns combination of Jim Brown and Leroy Kelly. (Funny how that worked.) No back's ever made it to four.
Predicted: 152 carries, 609 rushing yards, 5 TD
Actual: 70 carries, 390 rushing yards, 5 TD
Bush's raw numbers look worse than his advanced metrics; whether that's due to the efficiency of the Saints' offense is up to you. Worth noting: Bush has 1940 rushing yards through his first four seasons as a pro. A comparable total from a comparable player? Tiki Barber, who had 1941 rushing yards through his first four years. Barber achieved that figure in 52 fewer carries, though, and while he had a big senior season after three years of extremely limited play, Bush has had a steady role across his professional career.
Predicted: 194 carries, 837 rushing yards, 6 TD
Actual: 219 carries, 828 rushing yards, 10 TD
Again, the figures seem similar, but it's a difference of about a half-yard per carry. He does have a weird Bret Saberhagen-style touchdown count going on alternate seasons. I was bullish on Addai's figures because I was concerned about Donald Brown's ability to pick up the offense and serve as an effective pass-blocker.
Predicted: 228 carries, 1022 rushing yards, 7 TD
Actual: 114 carries, 409 rushing yards, 1 TD
Derrick Ward was considered to be an actual catch on the free agent market last offseason, and while he's still a talented back, playing on a dismal team behind a remarkably-healthy Cadillac Williams left him pretty much as a backup back on a terrible team. Insert your joke about the last two weeks of the Giants' season here.
Predicted: 291 carries, 1263 rushing yards, 9 TD
Actual: 282 carries, 1253 rushing yards, 11 TD
That's pretty much spot-on, huh? Grant had a flukily-low performance inside the five in 2008, so I thought his touchdown total would hit double-digits. On the other hand...
Predicted: 317 carries, 1320 rushing yards, 10 TD
Actual: 124 carries, 494 rushing yards, 1 TD
...Portis had a freaky-great performance inside the five in 2008, so it was reasonable to suggest his touchdown total would decline in 2009, albeit not to the point that it actually did. His numbers went down with the offensive line, but his health woes were another data point for the idea that the "workhorse back" doesn't exist.