NFL Power Rankings: Week1
By Keith Myers
Those of you who visited this site last season probably remember my mathematical power rankings that I updated each week. Well, I spent some time over the offseason delving much deeper into the statistics of what it takes to win in the NFL, and now much each of those factors effect the probability of coming away with a W.
The ranking have no bias, since all they are is a collection of stats mashed together into in a very complex way into one number, and then re-normalized so that the maximum is 100 and the minimum is 0. I’ve included 4 of the 30-some stats that are included in my mathematical model so you can see a little of what goes into creating the end result.
I had thought about not unveiling these until after week 4, or even later. Since they are stats based the results will swing wildly early in the season due to small sample size variance and the effects individual opponents can have, but then decided that there would be no fun in waiting for here there for week 1.
It should come as no surprise that the Ravens are on top this week, after they completely undressed the bangles in every phase of the game. You might notice that the Bangles are at the bottom of the rankings too, for the very same reason. Because we have only one game in the books so far, week one opponents are mirrored top and bottom. (The Jets played the Bills, the Bears played the Colts, etc.)
The Seahawks, who check in at #18, end up as the 2nd best of the losers half of the ratings. Only Cleveland lost ended up with a higher statistical rating. When you consider that the Seahawks were only 4 yards away from a victory, that seems about right in the end.
| Rank | Team | PtDif | YDS | TO | ST | Power |
| 1 | Baltimore | 31 |
2.7
2
-0.4
93.2
2NY Jets
20
-0.6
3
90.5
90.5
3Chicago
20
0.4
4
43.6
82.9
4New England
21
1.1
2
18.8
82.4
5Atlanta
16
1.1
3
-12.5
72.5
6Houston
20
0.3
4
-62.9
69.9
7Denver
12
2.1
0
-28.1
65.7
8Dallas
7
2.6
0
-8.5
61.1
9San Diego
8
0.1
1
-0.2
60.8
10Washington
8
0.9
3
-6.5
60.8
11San Francisco
8
0.8
1
-8.04
59.2
12Minnesota
3
2
0
17.8
57.0
13Detroit
4
1.8
-3
-5.38
55.8
14Tampa Bay
6
-1.9
2
-33.16
51.9
15Arizona
4
0.8
0
-29.34
51.7
16Philadelphia
1
1.6
-1
-18.6
50.7
17Cleveland
-1
-1.6
1
18.6
47.8
18Seattle
-4
-0.8
0
29.34
47.2
19Carolina
-6
1.9
-2
33.16
46.5
20St. Louis
-4
-1.8
3
5.38
44.4
21Jacksonville
-3
-2
0
-17.8
42.6
22Green Bay
-8
-0.8
-1
8.04
40.2
23Oakland
-8
-0.1
-1
0.2
38.9
24NY Giant
-7
-2.6
0
8.5
38.7
25Pittsburgh
-12
-2.1
0
28.1
37.5
26New Orleans
-8
-0.9
-3
6.5
36.6
27Kansas City
-16
-1.1
-3
12.5
31.3
28Miami
-20
-0.3
-4
62.9
29.0
29Tennessee
-21
-1.1
-2
-18.8
18.5
30Indianapolis
-20
-0.4
-4
-43.6
15.2
31Buffalo
-20
0.6
-3
-90.5
14.8
31Cincinnati
-31
-2.7
-2
0.4
5.5
PtDif = point differential
YDS = yards – this is a collection of offensive and defensive yards per play stats. Due to massive small sample size problems, I had to use a similar form of my algorithm this week for this rating.
TO = turn over differential
ST = special teams rating, calculated from kicking, punting and returning stats
The model also uses 3rd down efficiency, time of possession and a host of other stats that aren’t included in the chart above.