Big group of RBs with uncertain roles (1 Viewer)

[kal-el]

Below I've pasted current ADP data for RBs from antsports.com (I'm assuming it's OK to post this here since it's public information). I've included breaks to separate the backs into four pretty distinct tiers (you could move one or two guys up or down at each break):

1-11: Backs with (1) clear talent, (2) undisputed starter roles, and (3) no major injury concerns.

12-18: Backs with (1) a pretty high level of talent, (2) starter roles but with some threat to that role, and (3) some injury concerns. (Only 2 or 3 apply to some of the backs here.)

19-40: Backs with (1) pretty good talent, (2) starter roles with a threat OR a solid threat to become the starter, and (3) either age/injury concerns OR unproven ability at NFL level.

41-56: Back-ups who would become useful only if the starter in front of them were injured.

First Name Last Name Pos Team ADP High Low Std Dev Drafts

 1.     Larry Johnson RB KCC 1.01.35   1.01 1.03 0.65 23      

 2.     LaDainian Tomlinson RB SDC 1.02.04   1.01 1.03 0.47 23      

 3.     Shaun Alexander RB SEA 1.02.75   1.01 1.05 0.94 24      

 4.     Clinton Portis RB WAS 1.04.46   1.04 1.06 0.72 24      

 5.     Tiki Barber RB NYG 1.05.14   1.03 1.08 1.28 22      

 6.     Steven Jackson RB STL 1.07.04   1.05 2.01 1.43 23      

 7.     Edgerrin James RB ARI 1.07.70   1.06 2.02 1.58 23      

 8.     Ronnie Brown RB MIA 1.08.61   1.05 2.03 1.64 23      

 9.     LaMont Jordan RB OAK 1.08.79   1.05 2.05 2.65 24      

 10.     Rudi Johnson RB CIN 1.10.91   1.04 2.06 2.94 23      

 11.     Carnell Williams RB TBB 2.01.35   1.07 2.05 1.99 23      

12.     Brian Westbrook RB PHI 2.05.55   2.02 3.02 2.58 22      

 13.     Willis McGahee RB BUF 2.10.96   2.02 3.10 5.83 23      

 14.     Domanick Davis RB HOU 3.01.76   2.06 4.07 5.07 21      

 15.     Julius Jones RB DAL 3.05.95   2.05 4.10 5.94 22      

 16.     Kevin Jones RB DET 3.07.32   2.10 4.07 4.06 22      

 17.     Willie Parker RB PIT 3.10.10   3.01 5.05 5.81 21      

 18.     Chester Taylor RB MIN 4.01.71   2.10 4.10 5.26 24      

 19.     Reggie Bush RB NOS 4.03.96   2.10 5.01 6.51 23      

 20.     Jamal Lewis RB BAL 4.05.75   3.07 4.09 4.34 24      

 21.     Reuben Droughns RB CLE 5.01.41   3.09 6.08 8.32 22      

 22.     DeShaun Foster RB CAR 5.02.91   3.08 7.03 8.91 23      

 23.     Warrick Dunn RB ATL 5.03.73   2.08 6.08 8.33 22      

 24.     Corey Dillon RB NEP 5.05.67   3.08 7.03 10.38 21      

 25.     Cedric Benson RB CHI 6.01.00   4.09 7.10 8.87 22      

 26.     Tatum Bell RB DEN 6.01.11   4.01 7.04 9.53 19      

 27.     Joseph Addai RB IND 6.02.46   5.02 8.02 9.17 24      

 28.     Deuce McAllister RB NOS 7.06.17   5.08 9.05 10.95 18      

 29.     Fred Taylor RB JAC 7.07.76   6.01 10.06 12.53 17      

 30.     Thomas Jones RB CHI 8.01.44   5.01 9.09 12.69 16      

 31.     Ahman Green RB GBP 8.03.00   4.01 12.02 16.98 18      

 32.     Dominic Rhodes RB IND 8.03.39   6.10 10.01 7.79 18      

 33.     Frank Gore RB SFO 8.03.60   6.05 11.08 14.25 20      

 34.     Ron Dayne RB DEN 8.06.82   6.02 11.06 13.86 17      

 35.     DeAngelo Williams RB CAR 8.07.91   5.10 10.04 13.28 23      

 36.     Curtis Martin RB NYJ 9.03.33   7.03 11.08 14.58 9      

 37.     Laurence Maroney RB NEP 9.04.81   8.09 10.07 5.38 21      

 38.     Chris Brown RB TEN 9.05.58   6.10 11.03 9.81 19      

 39.     Marion III Barber RB DAL 9.09.65   6.07 12.02 13.92 17      

 40.     Lendale White RB TEN 10.03.47   9.03 11.06 6.68 17      

 41.     Sam Gado RB GBP 11.01.83   10.02 13.09 8.74 18      

 42.     Mike Bell RB DEN 11.03.44   7.04 14.03 26.26 9      

 43.     Greg Jones RB JAC 12.04.06   9.10 15.01 14.77 16      

 44.     Chris Perry RB CIN 12.09.00   10.01 15.09 21.92 7      

 45.     Mike Anderson RB BAL 12.10.20   10.08 16.10 16.63 15      

 46.     Kevan Barlow RB SFO 13.03.30   10.08 15.07 15.33 10      

 47.     Michael Bennett RB NOS 13.06.90   11.01 15.10 16.51 10      

 48.     Cedric Houston RB NYJ 13.07.50   10.10 15.08 15.87 12      

 49.     Michael Turner RB SDC 14.02.00   12.09 16.09 12.81 13      

 50.     T.J. Duckett RB ATL 14.03.17   11.07 16.02 14.93 12      

 51.     Mewelde Moore RB MIN 14.04.82   12.02 16.07 12.82 11      

 52.     Duce Staley RB PIT 14.07.67   11.08 16.04 17.87 9      

 53.     Ryan Moats RB PHI 14.07.75   12.02 16.10 16.24 12      

 54.     Ladell Betts RB WAS 14.10.29   11.02 16.10 19.10 7      

 55.     Michael Pittman RB TBB 15.02.80   14.01 16.10 12.15 5      

 56.     Brandon Jacobs RB NYG 15.09.13   12.06 16.09 15.37 8  

My question to how we judge the relative value of the backs in the third tier (which contains 22 RBs). I really have a hard time differentiating the value of Bush and Lewis at the top end and Brown, Barber, and White at the bottom end. All have a substantial probabilities of (1) being very useful starters for the bulk of the fantasy season or (2) having very little value at all for the bulk of the fantasy season.

I think FBG does a VERY GOOD job and projecting stats. The problem is that they're projecting average outcomes (expected values, in formal probablity terms). With RBs like these, though, there's a very high variance among potential outcomes. These RBs will have similar values to mid-round WRs under the VBD system. But the potential variance among the WRs will be lower. So the temptation when using the VBD system will be to take a WR with slightly higher value, even though the RB may end up with more fantasy value since he has a good chance of being the starter for a portion of the season. The WR, on the other hand, may be a player will never really post fantasy starter numbers over the course of the season.

So the two big questions are:

1) How do you differentiate among the third tier of RBs--those who are in 50/50 or 60/40 situation in terms of whether they'll end up being the main RB for their team? (This answer probably boils down to judgement/intuition.)

2) How do you value these RBs versus players at other positions under the VBD system in order to avoid undervaluing the RBs due to the large amount of uncertainty regarding the backfield situations of many teams this year? This is the key question I'd like to get at.

Obviously, projecting full outcome distributions for each player is out of the question, but there must be some quantitative way to get a handle on this question.

 

[Z-Dog]

The easy answer is to add a risk coefficient. Problem is how to determine the risk coeffecient. The answer to that is to use historical data (sort of an AVT) approach. Problem with that is that you're no longer evaluating individual players, you're forcing them into pre-existing boxes. Nonetheless, it provides a framework.

The real answer is that VBD has massive holes in it, of which this is only one. I abandoned any attempt to fix VBD. I use it to create an initial dept chart, but I monkey with the resulting list considerable.

Btw, this problem that you're having is significant in re-draft, but in auction is is even larger. Not only do you need to decide who is better than whom, you need to quantify by exactly how much.

 

[kal-el]

The easy answer is to add a risk coefficient. Problem is how to determine the risk coeffecient. The answer to that is to use historical data (sort of an AVT) approach. Problem with that is that you're no longer evaluating individual players, you're forcing them into pre-existing boxes. Nonetheless, it provides a framework.The real answer is that VBD has massive holes in it, of which this is only one. I abandoned any attempt to fix VBD. I use it to create an initial dept chart, but I monkey with the resulting list considerable.Btw, this problem that you're having is significant in re-draft, but in auction is is even larger. Not only do you need to decide who is better than whom, you need to quantify by exactly how much.

At the end of the day, maybe this question is sufficiently complicated that an intuitive judgement is the only solution. I've been meaning to read "Blink"; maybe I'll have to do that before my first draft.