Wisdom of Crowds Review: Running Backs
by Bill Barnwell
This week, we finish up our review of this year's Wisdom of Crowds feature by looking at how the pool of running backs performed.
For each back, I'll provide the average predicted performance by our Twitter panel from before the season, as well as their actual 2010 output. If the player participated in fewer than 16 games, I will also pro-rate his statistics for 16 games. You can read the initial articles covering the predictions here, here, and here.
Projected: 329 carries, 1643 rushing yards, 13 touchdowns
Actual: 316 carries, 1364 rushing yards, 11 touchdowns
With his absurd number of big plays from 2009 regressing towards the mean in 2010, Johnson's yards per carry were nearly seven-tenths of a yard below the average projection from 2009. The other contributing factor was the absence of Vince Young. We've found in the past that quarterbacks with a propensity for running increase the efficiency of their halfbacks, and that was certainly the case in Tennessee. With Young in the lineup, Johnson averaged 4.7 yards per carry, but when he was forced to work with Kerry Collins and Rusty Smith, Johnson produced just 3.7 yards a pop.
Projected: 280 carries, 1215 rushing yards, 9 touchdowns
Actual: 185 carries, 766 rushing yards, 2 touchdowns
Pro-rated: 197 carries, 817 rushing yards, 2 touchdowns
Greene lost his starting job after fumble issues against the Ravens in Week 1 and ended up in a timeshare with LaDainian Tomlinson. A decline in the run-blocking abilities (and health) of the Jets' offensive line didn't help, either. Greene's still a very good rusher, but he's still seen as erratic by the coaching staff in pass protection and as a receiver.
Projected: 247 carries, 1185 rushing yards, 11 touchdowns
Actual: 230 carries, 1467 rushing yards, 5 touchdowns
I thought Charles would hit his maximum projection -- 310 carries, 1600 rushing yards, and 12 touchdowns -- because I was naive enough to think that the Chiefs would give him the ball near the goal line. That didn't happen. He's not going to approaching 6.5 yards per carry again, but Charles will really be a fantastically difficult player to project in 2011. Is his role going to increase, and what will his rushing average be at 300 carries? Or will he remain in his current role: Greatest platoon back in league history?
Projected: 263 carries, 1051 rushing yards, 7 touchdowns
Actual: 237 carries, 1069 rushing yards, 6 touchdowns
Those numbers look awful similar to me at first glance, but it's actually off by a full half-yard per carry. I pegged him at halfway between his projected line and his maximum projection, so a season-long rate of 285 carries for 1142 yards and 11 touchdowns. The Bears gave the now-departed Chester Taylor too large of a role for that, and the team remained preposterously bad near the goal line during the first half of the season before bouncing back towards league-average. He's still going to be undervalued next season.
Projected: 294 carries, 1338 rushing yards, 11 touchdowns
Actual: 203 carries, 853 rushing yards, 3 touchdowns
Pro-rated: 295 carries, 1241 rushing yards, 4 touchdowns
Gore was a player that KUBIAK projected for a particularly poor season before the year, partly due to his injury history. In this case, it ended up being right. The young offensive line and carousel of quarterbacks helped keep Gore's rushing average down, and a fractured hip ended his season in Week 12.
Projected: 245 carries, 1047 rushing yards, 7 touchdowns
Actual: 207 carries, 1080 rushing yards, 7 touchdowns
Pro-rated: 221 carries, 1152 rushing yards, 7 touchdowns
I played around with a Ray Rice comp for McCoy before the season, as a versatile back who might make a significant jump in his second season. That ended up being the case, but it owes a lot to Michael Vick. McCoy averaged more than two additional yards per carry with Vick in the lineup as opposed to Kevin Kolb. And even without having expected Vick to stop by, it's clear that he was a better option than the guys who were just ahead of him in August 2010 ADP: Cedric Benson and Pierre Thomas.
Projected: 327 carries, 1430 rushing yards, 10 touchdowns
Actual: 330 carries, 1241 rushing yards, 6 touchdowns
For some reason, Jackson continues to avoid rushing touchdowns like the plague. After running in 13 scores in 2006, he has just 22 touchdowns on 1,144 rushing attempts in 55 games over the four subsequent seasons. He also had five receiving touchdowns through 2006, with only two afterwards. Both the Twitter constituency and I expected Jackson's rushing touchdown total to rise with a better quarterback and the likelihood of more wins, but that just didn't happen. It was also Jackson's first season with a rushing average below 4.0 yards per carry.
Projected: 221 carries, 1014 rushing yards, 7 touchdowns
Actual: 83 carries, 269 rushing yards, 2 touchdowns
Pro-rated: 221 carries, 717 rushing yards, 5 touchdowns
Thanks in part to a tough schedule, Thomas got off to a slow start. Before his schedule could ease up, he suffered an ankle injury that ended up costing him most of his season. A lost year, but he should be undervalued heading into 2011, even if he only ends up maxing out around 150 carries.
Projected: 325 carries, 1529 rushing yards, 14 touchdowns
Actual: 283 carries, 1298 rushing yards, 12 touchdowns
Pro-rated: 302 carries, 1385 rushing yards, 13 touchdowns
As the committee expected, AD was able to arrest the steady decline in his yards per carry, but his rushing touchdowns also fell off from a league-leading figure of 18 in 2009. Realistically, has any player gone through a more dramatic shift in context while playing on the same team as Peterson has during his four-year career? He came into the league with a terrible passing game and an excellent offensive line, and while the line's shifted down to mediocre, the passing game went all the way up to superb and back down again. And yet, he's been consistently effective, with the injury issues from college rarely popping up.
Projected: 313 carries, 1437 rushing yards, 13 touchdowns
Actual: 299 carries, 1324 rushing yards, 5 touchdowns
Pro-rated: 342 carries, 1513 rushing yards, 6 touchdowns
I had Jones-Drew at somewhere around those pro-rated figures, but with twice as many rushing touchdowns. Although we don't have final charting numbers available yet, it seems like the Jaguars played off Jones-Drew's effectiveness in the red zone and went play action more frequently. In 2009, the Jags had 19 rushing touchdowns and 15 receiving touchdowns; in 2010, they had 14 rushing touchdowns and 26 receiving touchdowns. Rashad Jennings vultured some, with four scores on 84 carries, and David Garrard had five of his own on 66 attempts. That should change in 2011, as MJD should get back to double-digit scores with a similar workload.
Projected: 252 carries, 1015 rushing yards, 7 touchdowns
Actual: 54 carries, 227 rushing yards, 2 touchdowns
Pro-rated: 173 carries, 726 rushing yards, 6 touchdowns
A piece of advice: don't separate your groin. I really thought Portis had some bounce-back potential this year, assuming he would be healthy and behind a much-improved offensive line, but the line wasn't all that great, either. I still think he has the potential to contribute in the right system, but it might not be as a featured back anymore. I'm fond of noting that he's yet to turn 30, but he already has more carries on his odometer than O.J. Simpson, Shaun Alexander, and Earl Campbell.
Projected: 308 carries, 1388 rushing yards, 11 touchdowns
Actual: 324 carries, 1273 rushing yards, 13 touchdowns
Even with the annoying insistence on using Ike Redman as a short-yardage back for part of the season, Mendenhall got both the workload and rushing touchdowns that many of us expected before the season. (It's worth noting that no other Steelers running back scored a rushing touchdown this year.) On the other hand, he just wasn't as effective as his projections by either KUBIAK (which famously projected him for a 344-1550-10 line) or Twitter, which each had him at 4.5 yards per carry. You can chalk it up to blown assignments on the offensive line, if you'd like, as the Steelers were 27th in the league in stuffs. With no other back on the roster taking significant reps away, an improvement in the play of the offensive line could turn Mendenhall into the league's best back in 2011.
32 comments, Last at 18 Mar 2011, 5:31pm
#1 by Theo // Mar 07, 2011 - 12:56pm
"We've found in the past that quarterbacks with a propensity for running increase the efficiency of their halfbacks"
Now that's something I'd like to read more about.
#2 by JasonK // Mar 07, 2011 - 1:02pm
I don't remember seeing the math, but logically, it's not rocket science. When defenses have reason to fear the bootleg, they're a step slower in converging on the point-of-attack.
#7 by Dean // Mar 07, 2011 - 3:28pm
If you want anecdotal evidence, just look at the performances of the various Atlanta RBs when the dogkiller was/was not in the lineup. The trend continued with Shady McCoy this year.
#9 by Theo // Mar 07, 2011 - 4:04pm
I understand the logic that football is a team sport. The thing is that I'd like to know the research/numbers behind it.
#17 by Aaron Schatz // Mar 08, 2011 - 8:22am
If you look in the ESPN archives, I think this was one of our MNF features before our stuff was Insider, back in 2006 I think.
#3 by joon // Mar 07, 2011 - 1:39pm
Realistically, has any player gone through a more dramatic shift in context while playing on the same team as Peterson has during his four-year career?
larry fitz, maybe.
#5 by big_jgke // Mar 07, 2011 - 2:36pm
#30 by Andrew Potter // Mar 10, 2011 - 2:10pm
First person I thought of too.
#4 by are-tee // Mar 07, 2011 - 2:11pm
Shonn Greene's lack of carries was really more a function of LT's resurgence. Tomlinson won the starting job more so than Greene lost it. As the season (and Tomlinson) wore down, Greene's role picked up. I expect him to be the Jets' lead back next season.
#6 by jmaron // Mar 07, 2011 - 3:17pm
Vick's winning pct is Atlanta was good - particularly compared with the guys who played when he was injured. Yet he was invariably very low in passing stats.
Perhaps guys like Vick are a good deal more valuable to their teams than their passing stats would suggest they are.
#8 by tony (not verified) // Mar 07, 2011 - 3:54pm
Unless I've missed something Chester Taylor is still with the Bears.
#15 by justanothersteve // Mar 08, 2011 - 12:13am
You are correct. Though why is another question.
#18 by BigCheese // Mar 08, 2011 - 12:14pm
Because they paid him something like 90% of his 9 Million 2-year deal in year one, so why not? Although there were some rumblings about him being released about a month ago.
If handled correctly, I think he should stay. And by that I mean if used far less than last year, but still enough to spell Forte a bit. You're probably not going to find another veteran back for $1 million that has the 3-rd down skills (the argument against going with a low-round rookie).
#22 by Thomas_beardown // Mar 08, 2011 - 2:31pm
I don't think he does anything that can't be better done by the combination Wolfe and Bell.
#10 by JonFrum // Mar 07, 2011 - 4:28pm
So is it fair to say that the Wisdom of Crowds overestimates RB yardage?
#12 by cisforcookie (not verified) // Mar 07, 2011 - 5:17pm
I think this has a lot to do with the expected shape of performance distribution in football. Let's say you're a good running back. Your expected performance is 275 carries for 1200 yards and 10 touchdowns. That assumes that you stay healthy and start 16 games. Now, in what direction is randomness going to affect you? unless herman edwards is your coach, you probably won't see more than 325 carries. that estimate is for around 4.36 ypc, but even if you fluke up to 4.7 ypc you're still below 1300 yards at 275 carries even though that's a HUGE improvement. Plus how many more touchdowns might you score even in the best season? jamal lewis, corey dillon, and curtis martin had a COMBINED 7 seasons over 10 touchdowns between their respective careers, so even if you're a great running back you probably aren't going to see that kind of performance consistently. As a result, the upside of a player who can, assuming good health, put up the above numbers (pretty common numbers for the 3 guys I just mentioned), really has very little upside and fairly low likelihood of actually reaching it. As a result, we see very few big whiffs on the low end with WOC.
The same analysis applies looking at downside risk, however there you get the opposite conclusions. Nagging injuries, injuries to teammates, bad weather, other players stealing carries, shifts in scheme, defenses paying more attention to a star, etc, all conspire to bring down performance without a commensurate reduction in games played. Except whereas it's difficult to improve from already really good, it's frighteningly easy to fail and often very difficult to recover from early setbacks. Sprain your ankle, a rookie takes half your carries, the coaches are thinking they'd rather not pay your incentives, and boom, your season is numerically toast. These sorts of random cram are difficult if not impossible to predict, however, so WOC doesn't take them into effect, resulting in a positive bias to predictions that won't be reflected in the end of season numbers.
#11 by ChicagoRaider // Mar 07, 2011 - 5:15pm
What about the Wisdom of Crowds on Darren McFadden. Oh, he wasn't in the list. Who's bad is that?
#19 by BigCheese // Mar 08, 2011 - 12:16pm
Who's bad? I'd say the person who can't use apostrophes correctly.
#24 by Mountain Time … // Mar 08, 2011 - 5:36pm
STOP TRAMPLING ON MY RIGHT TO BE IGNORANT!
#13 by BaronFoobarstein // Mar 07, 2011 - 5:32pm
Is the "the annoying insistence on using Ike Redman as a short-yardage back" from the perspective of not matching projections or a commentary on personnel selection and usage from a team effectiveness point of view. My subjective viewpoint had Isaac Redman as a reasonable choice and effective short yardage back (especially and surprisingly catching in short yardage situations), but I'd be very interested to read a better analysis, especially if it contradicts that observation.
#14 by Phil // Mar 07, 2011 - 8:41pm
Agreed. I thought Redman was a very effective short-yardage back this season.
#23 by Anonymous Coward (not verified) // Mar 08, 2011 - 4:48pm
As a Steelers fan I was annoyed that Mendenhall got goal line carries ahead of Redman
#16 by slipknottin // Mar 08, 2011 - 1:04am
Honestly, it's getting annoying that every article about the chiefs or Charles is just about how you feel the chiefs didn't give him enough carries. We get it already. Why not talk about something more useful
#20 by BigCheese // Mar 08, 2011 - 12:19pm
Like for example.... what exactly?
Seriously, what information relating to the projections vs actual numbers of Jamaal Charles' rushing stats do you feel could have been brought up that's even slightly more useful or relevant than what was brought up?
#26 by Jimmy // Mar 09, 2011 - 3:08pm
What I find bizarre is that FO came up with the 'Curse of 370' argument and then a team starts heavily spelling its extremely productive tailback (presumably) in an attempt to lengthen his career. It is almost a natural experiment for FO to observe; if the coaches in KC stay the same or if Charles usage remains consistent and he rushes for 1400+ yards on 230 carries a season for the next five years you might say that it proves the over use theory correct (insert obvious caveat about sample size). Of course that does mean waiting five years. From the point of view of winning a championship trying to save your best offensive player's touches for the postseason might help you win a Lombardi even if it costs you the odd regular season game.
Maybe the point FO should be making is that the Chiefs need to replace Thomas Jones (unless they already have that is).
#27 by Theo // Mar 09, 2011 - 6:08pm
I think the coaches understand that they need to keep their best player at their best.
History tells us so.
...and yes running into a wall hurts.
That's why we invented 3rd down backs.
#21 by BigCheese // Mar 08, 2011 - 12:20pm
Anyone notice that the projection got Johnson's rushing yards' numbers exactly right, just not in the correct order?
#25 by dbostedo // Mar 09, 2011 - 12:31pm
#29 by Bobman // Mar 09, 2011 - 11:51pm
You have to admit, though, that 31 TDs would be pretty impressive, if THOSE numbers got switched around.
Or 631 carries.
NO. That would just be disturbing.
#28 by LouiseLopz (not verified) // Mar 09, 2011 - 6:24pm
Quite a summary. Certainly sums up what the running backs achieved. Obviously, these stats don’t just point to individual numbers but showcase team performance.
#31 by DGross (not verified) // Mar 13, 2011 - 11:55am
Mendenhall will get that 4.5 average if the line can just open up a few more holes. He just doesn't have a lot of room to work with. Improve that line and watch him fly.
#32 by Mark Pitcavage (not verified) // Mar 18, 2011 - 5:31pm
I don't understand the comments re Pittsburgh running back Redman--he actually performed quite well during the season in a limited role and I strongly wished they had given him the ball more often--either rushing or coming out of the backfield to catch a pass.