Too Early to Cry Over Spill-t Milk?

Photo: Josh Schultz. License

(Note: Check out the new updated version of the article at our new location with a brand new look and feel: Blog-a-Bills)

The Bills have had an interesting recent past when it comes to the drafting of skill position players in the 1st round (in the same way that a train wreck is interesting). Names such as Willis McGahee, Marshawn Lynch, and (for some) Lee Evans bring shudders down the spines of many Bills fans as wasted first round picks. (This phenomenon isn’t limited to RBs or WRs, lest we forget JP Losman or Mike Williams).

Across the league, luminary running backs such as Trung Canidate, Tim Biakabatuka, and William Green have all served as cautionary tales for taking multiple-threat RBs, like CJ Spiller, in the 1st round of the NFL draft.

We all bore witness to CJ’s less-than-stellar rookie year with the Bills. His play and his stats did nothing to ease those bust concerns:

G	Rushes	Average	Rec	Yds	Yds/Rec	TDs	Yds/Sm
14	74	283	24	157	6.5	1	440

Due in large part to his performance, the fact that many Bills fans were incredibly displeased with what many saw as a “luxury pick”, and Buffalo’s (over?)-sensitivity to busts since the whole Aaron Maybin debacle (wait – he’s still on the team?), the questions have already been asked by fans and the media: Is CJ a bust?

It’s always been my opinion that to judge a draft pick (and, by extension, an entire draft) accurately, one needs to let at least 3 years pass. The NFL has seen so many late bloomers at so many different positions (Thomas Jones being an ideal example at RB) that to hastily declare this player or that player is a bust after one year has always seemed incredibly reactionary to me.

That, however, doesn’t mean we can’t take a deeper look into the numbers, and see if there is any trend we can identify based on previous RBs drafted in the first round since 1990 to determine what we might be able to expect from CJ this year, and in the future.

Multiple Threat RBs

There is an unmistakable difference in aura about a team that has an RB who is a threat both as a receiver and as a rusher. At their best, running backs like Marshall Faulk and Thurman Thomas were dangerous every single time they touched the ball, whether it was on a draw from the K-gun or a swing pass from Kurt Warner. It’s that aura (I believe) that causes so many teams to take chances on RBs despite the hit-or-miss nature of drafting them early. That set of game-breaking skills that helped Spiller get drafted where he did, but it’s also what made his performance last year so frustrating.

In order to see what hope we might have for CJ in future, I decided to look back at RBs drafted in the 1st round since 1990, and analyze their first and second years in the league to see if there’s any trend that may indicate what we can expect from CJ this year. In order to make the analysis more relevant, we analyzed different subsets of the RBs drafted.

The first such group, as you might imagine, is the rushing/receiving type of RB that CJ embodies. Obviously, identifying which RBs do or do not have similar skill sets to Spiller is an incredibly subjective exercise, but we did our best to identify similar players by looking at college stats and/or measurables. We found 27 players fitting that archetype drafted in the 1st round since 1990 (not including those drafted last year):

Year	Pk	Tm	Name
2009	12	DEN	Knowshon Moreno
2008	4	OAK	Darren McFadden
2008	22	DAL	Felix Jones
2008	23	PIT	Rashard Mendenhall
2008	24	TEN	Chris Johnson
2007	12	BUF	Marshawn Lynch
2006	2	NOR	Reggie Bush
2006	27	CAR	DeAngelo Williams
2006	30	IND	Joseph Addai
2004	24	STL	Steven Jackson
2003	23	BUF	Willis McGahee
2003	27	KAN	Larry Johnson
2002	16	CLE	William Green
2001	5	SDG	LaDainian Tomlinson
2000	31	STL	Trung Canidate
1999	4	IND	Edgerrin James
1999	5	NOR	Ricky Williams
1998	5	CHI	Curtis Enis
1998	1	JAX	Fred Taylor
1998	29	MIA	John Avery
1997	12	TAM	Warrick Dunn
1996	8	CAR	Tim Biakabutuka
1995	18	OAK	Napoleon Kaufman
1994	2	IND	Marshall Faulk
1990	2	NYJ	Blair Thomas
1990	20	ATL	Steve Broussard
1990	25	SFO	Dexter Carter

Having identified a list of similar RBs to Spiller (in terms of versatility), we then decided to compare how this list of players performed in their rookie years to Spiller’s rookie year:

	Gms	GS	Runs	Yds	TDs	Yds/Rush	Rec	Yds	TD
Others	12.85	7.037	174.3	720.2	4.851	4.166	27.69	234.0	0.769
Spiller	14	1	74	283	0	3.8	24	157	1

It doesn’t take a rocket scientist, or even a Billy Joe Hobert, to see that Spiller did not compare well to the average 1st year performance of similar RBs. Spiller was behind said average in almost every category save two.

Even if we were to argue that Spiller was somehow more of a polished receiver than the rest of these RBs coming out of college, and that was what he was drafted for, his stats in the receiving department show no upgrade over the average.

The next logical question would seem to be: How much did these RBs improve from their 1st to their 2nd year?

	Gms	GS	Runs	Yds	TDs	Yds/Rush	Rec	Yds	TD
2nd Yr	13.63	9.85	197.15	855.89	5.78	4.39	34.15	277.12	1.27
Avg Improvement	0.78	2.81	22.81	135.59	0.93	0.22	6.46	43.04	0.5

The average improvement here is not terribly impressive, especially when looked at over the course of a season (i.e. 135 rushing yards over the course of 16 games is about 8.5 yards more per game). If Spiller were to follow a similar trajectory (which is no guarantee), he wouldn’t be doing much at all to help his cause as a non-bust in the eyes of Bills fans.

Of course, given how much better this list of RBs performed in their 1st year when compared to CJ, there is obviously less room for improvement than on a season like the one Spiller had last year, so the relatively miniscule improvements could be a function of that fact.

1st Round RBs with Poor 1st Years

To account for that, we took another list of RBs drafted in the first round since 1990, but this time looked for those RBs who had similar (i.e. poor) first seasons to Spiller’s. Below find that list of players:

Year	Pk	Tm	Name
2009	27	IND	Donald Brown
2008	4	OAK	Darren McFadden
2006	27	CAR	DeAngelo Williams
2005	4	CHI	Cedric Benson
2004	24	STL	Steven Jackson
2002	18	ATL	T.J. Duckett
2001	23	NOR	Deuce McAllister
2000	7	ARI	Thomas Jones
2000	19	SEA	Shaun Alexander
1999	5	NOR	Ricky Williams
1998	5	CHI	Curtis Enis
1995	1	CIN	Ki-Jana Carter
1995	17	NYG	Tyrone Wheatley
1995	18	OAK	Napoleon Kaufman
1995	19	JAX	James Stewart
1994	25	KAN	Greg Hill
1993	21	MIN	Robert Smith
1991	21	KAN	Harvey Williams
1990	20	ATL	Steve Broussard

Again, we have here a list featuring RBs that went on to have productive, even near-great careers, but also a list featuring gigantic disappointments like Ki-Jana Carter and Curtis Enis.

The first piece of analysis we conducted here was to again take the average first year production for the above list of RBs and compare it to Spiller’s first year. Obviously we would expect this average to be much closer to Spiller’s output than the previous list.

	Gms	GS	Runs	Yds	TD	Yds/Rush	Rec	Yds	TD
Similar 1st Year RBs	13.54	3.95	127.27	504.68	3.1363	4.0727	16.909	136.22	0.3636
CJ Spiller	14	1	74	283	0	3.8	24	157	1

Looking at the above table, we can be reasonably content with the similarity in the numbers between the two groups. It might not be perfect, but it is indicative of a group of RBs who arguably had as poor, or poorer, 1st years than Spiller did.

The main reason to break out these players this way is, as mentioned above, to determine if there is any sort of upward trend between year one and year two with these RBs that would give us some added measure of hope for Spiller this year. With that said, below find the average improvement for first round RBs who had similar first years to Spiller’s:

	Gms	GS	Runs	Yds	TDs	Yds/Rush	Rec	Yds	TD
2nd Yr	14.47	6.68	168.89	683.68	5.21	4.01	23.74	184.74	0.89
Avg Improvement	1.32	3.21	59.37	247.42	2.84	-0.08	6.63	48.68	0.58

As we’d expect, the average improvement from year one to year two for this group of RBs is greater than it was for the first list we looked at, as by definition, there is much more room for improvement for players who had poor first years when compared to a group of players whose performance varied quite a bit in their rookie season.

Even with that being the case, though, would we be pleased if Spiller were to improve on his numbers by these averages?

	Gms	GS	Runs	Yds	TDs	Yds/Rush	Rec	Yds	TD
Projected Spiller 2nd Yr	15.32	4.21	133.37	530.42	2.84	3.72	30.63	205.68	1.58

Of course, this isn’t necessarily the most scientific method to project a RBs performance, but this isn’t to be used so much as a projection as it is a “what-if.”

After accounting for an average jump in statistics for CJ, I doubt there would be many Bills fans who would stop discussing whether or not CJ is a bust. 4 total TDs and fewer than 1,000 yards from scrimmage wouldn’t do much to help anyone’s stock in Buffalo.

As I mentioned earlier, I prefer to wait at least 3 years before judging a draft pick because of the potential for a premature declaration of a bust (Thomas Jones, anyone?). But that’s just a personal guideline of mine. Does it actually make any statistical sense for us to wait that long for RBs to develop?

When Do First Round RBs Peak?

Looking at all RBs taken in the first round since 1990, we identified when each RB had their “peak” season (defined in this case by using pro-football-reference.com’s Approximate Value ranking for each player.) If a RB had multiple seasons of an equal AV, the first season in which he hit his peak AV mark was used as his peak season. RBs who have yet to garner 3 years experience in the league were left out of the study.

Year	Pick	Team	Name	# Seasons to peak
2007	7	MIN	Adrian Peterson	3
2007	12	BUF	Marshawn Lynch	2
2006	2	NOR	Reggie Bush	1
2006	21	NWE	Laurence Maroney	2
2006	27	CAR	DeAngelo Williams	3
2006	30	IND	Joseph Addai	1
2005	2	MIA	Ronnie Brown	4
2005	4	CHI	Cedric Benson	5
2005	5	TAM	Cadillac Williams	1
2004	24	STL	Steven Jackson	3
2004	26	CIN	Chris Perry	2
2004	30	DET	Kevin Jones	1
2003	23	BUF	Willis McGahee	2
2003	27	KAN	Larry Johnson	3
2002	16	CLE	William Green	1
2002	18	ATL	T.J. Duckett	2
2001	5	SDG	LaDainian Tomlinson	6
2001	23	NOR	Deuce McAllister	2
2001	27	MIN	Michael Bennett	2
2000	5	BAL	Jamal Lewis	3
2000	7	ARI	Thomas Jones	9
2000	11	NYG	Ron Dayne	7
2000	19	SEA	Shaun Alexander	6
2000	31	STL	Trung Canidate	2
1999	4	IND	Edgerrin James	1
1999	5	NOR	Ricky Williams	4
1998	5	CHI	Curtis Enis	2
1998	9	JAX	Fred Taylor	1
1998	18	NWE	Robert Edwards	1
1998	29	MIA	John Avery	1
1997	12	TAM	Warrick Dunn	1
1997	23	BUF	Antowain Smith	5
1996	6	STL	Lawrence Phillips	1
1996	8	CAR	Tim Biakabutuka	4
1996	14	HOU	Eddie George	4
1995	1	CIN	Ki-Jana Carter	2
1995	17	NYG	Tyrone Wheatley	6
1995	18	OAK	Napoleon Kaufman	3
1995	19	JAX	James Stewart	6
1995	21	CHI	Rashaan Salaam	1
1994	2	IND	Marshall Faulk HOF	6
1994	25	KAN	Greg Hill	2
1993	3	PHO	Garrison Hearst	1
1993	10	RAM	Jerome Bettis	5
1993	21	MIN	Robert Smith	6
1992	9	CLE	Tommy Vardell	2
1992	19	ATL	Tony Smith	1
1992	21	NOR	Vaughn Dunbar	1
1991	14	NWE	Leonard Russell	3
1991	21	KAN	Harvey Williams	5
1990	2	NYJ	Blair Thomas	1
1990	17	DAL	Emmitt Smith HOF	3
1990	19	GNB	Darrell Thompson	4
1990	20	ATL	Steve Broussard	1
1990	24	NYG	Rodney Hampton	3
1990	25	SFO	Dexter Carter	2
			AVERAGE:	2.89

On average, 1st round running backs tend to hit their peak season in the 2nd-to-3rd season of their careers. This, of course, does not leave much time for Spiller to start performing at a higher level before cries of “bust” are averted. While you have the odd situation with someone like Thomas Jones or Cedric Benson becoming much better running backs much later in the careers, they are the exception, not the rule. This average mark also does not account for the fact that some players were impact running backs from day one, but went on to have even better seasons later in their careers (Emmitt Smith, Jerome Bettis, and Adrian Peterson come to mind here).

Interestingly (but perhaps not surprisingly), when we perform the same study on RBs who had poor 1st years in the league, the average number of years before those RBs peaked was 3.35, slightly higher than the overall population of 1st round RBs. However, it still implies that some level of peak production can be reasonably expected of Spiller sometime in the next year or two.

Soooooo, it’s too Early?

The numbers we’ve looked at thus far, while not promising, do indicate some level of patience with Spiller may be warranted. However, the group of RBs who had rookie seasons as poor as Spiller’s are not, on the whole, the most scintillating group of RBs I’ve ever seen. A quick look at the names on the list show that less than half of those RBs went on to have careers befitting a first round draft pick (a subjective call, but for the purposes of this post, one I’m comfortable with).

We all know that Spiller has talent falling out of his ears, and will likely make a play or two that will draw out the “oohs” and “ahhs” from the fans. Whether or not he’ll be able to maintain a high level of play is severely in doubt right now, but plenty of evidence exists to suggest that at least one or two more years of patience must be shown with Spiller before making a true judgment of his value.