It's only several days before 2009 NFL Draft. Today I am very curious about this: which team did "well" in recently years' NFL drafts?
Well the term "well" absolutely needs further definition. Without much background knowledge (and also because I don't have much time to make more delicate analysis), the most intuitive way is to calculate how many games each drafted player plays every year. For example, If your drafted players plays an average stable 14 games per year, well you made a pretty good job.
Generally in my humble opinion, we need at least two aspects of adjustment to make this model more accurate: 1. "# games played" is not a very good indicator to measure team's draft success: Lions WR Calvin Johnson played 31 games in two years, recording 126 catches, 2087 yards and 16 TDs; Dolphins Ted Ginn Jr. played 32 games in two years, however only contributes 1210 yards and 4 TD. It's true that we should take playes' overall performance into consideration. But since those performance are position-related (how you compare, let's say a wide receiver with an offense tackle?), we omit such work here.
Also, "# games played" doesn't reflect players' role in games; one can be an average player in special teams, comparing to a franchise quarterback. But still, since we don't have good way to consider it quickly, we just use this metric to see what team did in recent years.
The other factor, however, is easier to handle: different draft positions usually implies different quality of players. Thus, we summarize the performance of drafted players and come up the following graph:
So we see that, high draft picks always means more played games (which is natural). So to evaluate the quality of player at specific draft position, we just consider the difference between his # of games played with the average level at his position.
For example, for 1st draft choice in 2007, Raiders selected JaMarcus Russell who finally played 19 games in two years, i.e. 9.5 games per year. Average 1st draft choice played (according to the black trendline in above graph) 13.2 games per year. So Raiders scores -3.7 for JaMarcus. Use this way we calculate each team's average draft score between 2002 (Texan's first year in NFL) to 2007 (latest data we can get from pro-football-reference.com). The result is:
Team Score Players
NYG 1.881 42
NYJ 1.174 44
BAL 1.081 51
SDG 0.927 47
BUF 0.887 44
DAL 0.838 48
ATL 0.705 46
SEA 0.670 48
IND 0.558 51
CHI 0.271 50
SFO 0.206 54
JAX 0.185 51
OAK 0.135 48
ARI 0.019 40
TEN 0.003 61
CAR -0.057 48
KAN -0.124 42
MIN -0.138 44
CLE -0.168 46
PHI -0.247 50
MIA -0.269 41
GNB -0.302 49
NWE -0.336 48
STL -0.573 52
HOU -0.626 49
PIT -0.731 46
CIN -0.746 46
TAM -0.837 50
WAS -0.951 33
NOR -0.998 43
DEN -1.303 45
DET -1.487 43
For example, each (totally 42) New York Giants' drafted players from 2002 to 2007 played 1.881 games more than average level. Each Detriot Lions players played 1.487 games less than average level. Although these two extreme cases fits their team strength, we see some counter exampls as well: e.g., my dear, returnning world championship Pittsburgh Steelers lies at 26th in the table, with a score -0.731.
But Steeler's situation could also be explained by, Steelers are very conservative on the development of new players. They always protect rookies before they are sure that new guys are ready for play. 2008 high drafted Rashard Mendenhall and Limas Sweed are not immediate starters, so as to Timmons and Woodley in 2007. It's more of a, teams strategy towards the use of new players.
But I think we should still be impressive on the work done by Giants scouts panel (and other top-rated teams). And Lions, take care of your #1 overall picks this year!
So good luck all teams in 09 draft!
