As noted by Drinen, one measure of the importance of a player in the contest results is the %Live which is the percentage of entries having a specified player who are still alive as of a given week.
Another measure, using a different combination of the same variables that he publishes weekly, is the degree of concentration (or dilution) of rosters having that player (as of a given week). A “Concentration/Dilution Ratio” (CDR) can be calculated as:
CDR(i) = [np(i)/n(i)] / [NP/N]
Where:
n(i) = total number of live rosters in week i
np(i) = number of live rosters in week i having a certain player
N = starting total number of rosters
NP = starting number of rosters having a certain player
A CDR(7) of 1.5 (for a specified player) indicates that the player was on a much higher proportion of rosters at the end of Week 7 (compared to the average player). Similarly, a CDR(7) of 0.5 indicates that relatively few rosters containing that player survived to Week 7.
As of the end of Week 7, McNabb, Kevin Jones, and Roy Williams had the highest CDR ratios by far and presumably were the biggest contributors to their owners surviving through the 1st 7 weeks of the contest (column 1 below).
But things changed substantially in Weeks 8 through 11. Roy Williams had an average impact, but McNabb and Kevin Jones both had a negative influence on survivorship (column 2).
Obviously, Tomlinson had the biggest impact in Weeks 8-11 by a large margin. Tomlinson's CDR of 3.01 (for Weeks 8 to 11) quantifies how big that impact has been. The other big contributors to rosters were Lee Evans, Willie Parker, Miami DST, and Braylon Edwards.
For Weeks 1 to 11 combined (column 3), the players with the largest CDR ratios were:
1. 3.46 Tomlinson RB
2. 2.10 Evans WR
3. 1.81 Parker RB
4. 1.73 Edwards WR
5. 1.61 Wayne WR
6. 1.58 R. Williams WR
7. 1.52 Jennings WR
8. 1.51 Losman QB
9. 1.50 Palmer QB
10. 1.49 Miami DST
11. 1.44 Vick QB
12. 1.43 P. Manning QB
13. 1.39 TJ Houshmanzadeh WR
14. 1.38 B. Watson TE
15. 1.37 T. Glenn WR
16. 1.35 F. Gore RB
From Weeks 1 to 7, the strongest roster would have contained:
QB - McNabb, Kitna, Vick
RB – K. Jones, Parker, Gore
WR – R. Williams, TJ Housh, L. Evans
TE – LJ Smith
DST - Minnesota
From Weeks 8 to 11, the strongest rosters included:
QB – Palmer, Losman, P. Manning
RB – Tomlinson, Parker, Gore
WR – Evans, Edwards, Wayne
TE – Watson
DST - Miami
Player WKS 1-7 WKS 8-11 WKS 1-11 (Rank)
D. McNabb QB 1.72 .58 1.00
K. Jones RB 1.65 .73 1.20
R. Williams WR 1.52 1.04 1.58 (6)
TJ Houshmanzadeh WR 1.38 1.01 1.39 (13)
L. Evans WR 1.37 1.53 2.10 (2)
G. Jennings WR 1.37 1.11 1.52 (7)
B. Berrian WR 1.36 .63 .86
LJ Smith TE 1.34 .76 1.01
J Kitna QB 1.34 .90 1.20
R. Brown WR 1.34 .52 .69
A. Johnson WR 1.32 .87 1.15
Minnesota DST 1.29 .95 1.23
T. Glenn WR 1.25 1.10 1.37 (15)
W. Parker RB 1.21 1.50 1.81 (3)
F. Gore RB 1.21 1.12 1.35 (16)
B. Edwards WR 1.20 1.44 1.73 (4)
M. Vick QB 1.18 1.22 1.44 (11)
B. Watson TE 1.16 1.19 1.38 (14)
P. Manning QB 1.16 1.23 1.43 (12)
R. Wayne WR 1.15 1.40 1.61 (5)
L. Tomlinson RB 1.15 3.01 3.46 (1)
P. Rivers QB 1.12 1.21 1.35
JP Losman QB 1.11 1.36 1.51 (8)
C. Palmer QB 1.07 1.40 1.50 (9)
Miami DST 1.01 1.48 1.49 (10)
P.S. Hope the math is correct --