Tough to Rank: DeAngelo Williams

Time for everyone's favorite short-lived running segment from last year I forgot about: Tough to Rank. In it, I bring up a player that could seemingly go in any direction and throw it to the floor to find out your infinite wisdom.
First up is last year's Fantasy MVP: DeAngelo Williams. In the first half of last season, Williams was the RB19: 620 total yards, five scores. In the second half, he went bonkers: 1,038 total yards, 15 scores!
The Panthers had 500 rushing attempts and 30 rushing scores last year. Those overall numbers are likely to fall with a tougher schedule that includes both Eastern Divisions.
So which guy do we get this year? Jonathan Stewart had 89 fewer carries than Williams last year; will he close the gap?
Early Mock Draft data has Williams going as a mid-to-late first-round pick. My tone here probably makes it obvious that I see that as tough to swallow. But obviously we can't dismiss last year's top scorer. Let me know where you think he fits.





Comments
From the overall tenor of the comments, it sounds like D. Williams will be underrated going into next year. I think you can really only make an argument for 6 backs ahead of Williams (Peterson, MJD, Turner, Jackson, Forte, C. Johnson), and the only two you could really feel confident in are MJD and Purple Jesus.
Arguments against sharp regression:
1. You can argue over whether "Dude" is right in saying Williams is subjectively the best back in football, but he was clearly the best RB in football in '08 and it wasn't particularly close. He was #1 in DVOA, #1 in DYAR, #1 in TDs, #1 in fantasy points, #2 in TD percentage inside the 3, and #3 in what iWon calls big play rushes. That adds up to overwhelmingly the best back objectively speaking.
2. A season ago, Williams and Stewart combined for 28 TDs and ranked 2nd and 4th in TD percentage inside the 3. Even with a tougher schedule the Panthers are a strong bet to be at or near the top in RB touchdowns in '09. Strong success by both backs in '08 would suggest a continued split of goal line and red zone carries.
3. A season ago, Williams and Stewart combined for over 2300 rushing yards and ranked 1st and 12th in efficiency (DVOA).
4. Not only a great rushing team, the Panthers ranked 9th in passing DVOA a year ago. With Steve Smith and a strong offensive line, it's virtually impossible to successfully load up against the RBs.
Even if you apply every correction (regression, more difficult schedule, lost carries to Stewart) you are left with a near certain top 5 back. It's important to keep in mind that some of the best workhorse backs are operating in offenses where even a full portion of the touchdowns may yield a (significantly?) smaller piece of the TD pie (Gore, SJax, Forte, Portis).
It seems to me that commenters are figuring risk incorrectly. When you draft a RB in the top 5, you want to maximize your potential upside and minimize potential downside. Williams has already won a RB points title so the upside potential is proven. His healthy downside is almost certainly no worse than finishing slightly behind J. Stewart in fantasy production. If you take their combined fantasy production from a year ago and give Williams a 48% split, I'd argue that's still a low end RB1. (If you go for extreme downside, you could take 10% off their combined 2008 performance and then assign Williams a 45% split. By last year's numbers, that would translate into RB18, hardly an unmitigated disaster.)
The unhealthy downside for Williams is the same as for all RBs, getting hurt and missing the season, but it is possible to project odds on his getting hurt. His injury history suggests a back less prone to injury than average, but let's ignore that.
There is plenty of evidence debunking the Curse of 370, but if we simply assign every RB carry a set likelihood of injury, then we can project a RB's relative chances to get through a season. Let's say Williams averages 17 carries a game and SJax or AP average 25. Since the latter two backs average 50% more carries and every carry has the same chance of injury (if anything, each subsequent carry should represent a slightly greater risk), then they have a 50% greater chance of getting injured in any given game.
Therefore if we're drafting a back in the hopes that the runner will achieve our most optimistic projections, it makes sense to draft a back that achieves that production by virtue of a lower number of carries. Turner, Peterson, SJax, and MJD are all likely to outperform Williams if they make it through the year unscathed, but they are significantly less likely to do so.
(This is also a strong argument for C. Johnson, but the case for regression is stronger in Tennessee where the loss of Haynesworth has profound implications for '09 TOP and total RB carries. Moreover, C. Johnson and AP have high stuff rates because opposing defenses do not have to respect the pass. The Vikings have made much stronger strides to remedy this in '09 than have the Titans.)
Really, I think you can make a compelling case to take DeAngelo Williams #1.
Posted by: The Dude Abides | May 5, 2009 03:41 AM
Wow! Good question. Carolinas run schedule is not as favorable this year as opposed to last year. And with a healthy Stewart, you have to think that Fox will want to get Stewart in the mix more. The one thing that will help Williams is the lack of upgrade to the passing game. I believe that Williams HAS elevated his game, but a first rounder?? As tempting as it would be to pick him up early, it is still hard to get excited about 1 year wonders.
Posted by: Tommy C | May 5, 2009 08:33 AM
To "The Dude Abides"
WOW! Excellent post. Very well written and we agree completely on D-Willy. Is there risk? yes, but all RB's have risk and the risk/reward on DeAngelo is a far better bet than most RB's give you.
Posted by: Shawn | May 5, 2009 09:07 AM
Everybody wants a sure thing. Fantasy football mocks owners who think this way ('08 Brady, anyone?). Mr. Rosenthal, I submit to you that every owner (yourself included), when deciding where to rank a player, consider how much risk/reward quotient he/she is happy with inside each position tier (RB1,2,3...WR1, etc.) in accordance with their league scoring rules and personal managing preferences. You like a RB1 with high upside but are ok risking a possibly strong RBBC situation, rank him higher. You like less gamble with possibly less upside, rank others ahead of him. Remember, you are competing against the other owners in your league, and only you know which players you'd like to beat them like a drum with! That's where the fun is!
Posted by: Qsurdaddy | May 5, 2009 09:38 AM
I personally always liked DeAngelo, especially when he was behind Foster....but I agree that this Stewart kid is the better back/more talented of the two especially with Carolina's smash mouth system.....DeAngelo benefited from D's playing him easier than they played Stewart....D's knew when the rook came in they were planning to run and called plays accordingly....DeAngelo hit alot of homeruns last year, I am not so sure D's don't game plan for him this year....making the employment of Stewart all the more appealing to Coach......DeAngelo is a top 15 back based on talent, top 3 based on last years stats and ignoring th efact that Stewart was injured twice....Stewart is a top 10 back based on talent, top 20 based on last years performance (close to 10 TD's if I am not mistaken)
Posted by: Raju | May 5, 2009 10:03 AM
williams outperformed foster in his first two years when he got carries...
the panthers coaching staff seems to almost use seniority sometimes and with some positions. expect williams to maintain much of what he did last year with the loss of a few TD's to stewart and some drop off due to the tough schedule.
Posted by: appanther | May 5, 2009 10:24 AM
It was all about the improved O Line last year in Charlotte. His value to me is only as healthy as his O-Line is, which on draft day should be 100%. Do you take him before LT? Weird.
Posted by: Fantasy Phil | May 7, 2009 12:02 AM