2014 Season‎ > ‎

2014-09-28

Scroll to the bottom for notations regarding adjustments.

1 UCLA 0.446134
2 Florida St. 0.430746
3 Nebraska 0.403527
4 Alabama 0.379091
5 Mississippi St. 0.377779
6 Auburn 0.367007
7 Arizona 0.350331
8 Oregon 0.345429
9 Texas A&M 0.311833
10 Oklahoma 0.301018
11 Brigham Young 0.300013
12 Georgia Tech 0.246959
13 Notre Dame 0.243393
14 Ole Miss 0.240431
15 North Carolina St. 0.205172
16 LSU 0.157198
17 TCU 0.152580
18 Marshall 0.144607
19 Southern California 0.142507
20 Missouri 0.129120
21 Baylor 0.127832
22 Minnesota 0.108522
23 Georgia 0.096024
24 Oklahoma St. 0.087407
25 South Carolina 0.086215
26 East Carolina 0.076637
27 Louisville 0.076443
28 Michigan St. 0.064352
29 Stanford 0.061634
30 Penn St. 0.057751
31 Ohio St. 0.051282
32 Washington 0.049115
33 Kansas St. 0.047454
34 Cal (Berkeley) 0.046428
35 Wisconsin 0.046084
36 Wyoming 0.031726
37 Arizona St. 0.027682
38 Florida 0.022135
39 Maryland 0.018732
40 Colorado St. 0.017979
41 Kentucky -0.000007
42 Arkansas -0.000608
43 Rutgers -0.018629
44 Air Force -0.020616
45 Northern Illinois -0.030686
46 Cincinnati -0.033938
47 Nevada (Reno) -0.035701
48 Oregon St. -0.038128
49 Louisiana-Monroe -0.047871
50 Duke -0.053122
51 Virginia Tech -0.056094
52 Illinois -0.060227
53 Virginia -0.062322
54 Texas -0.070060
55 Middle Tennessee -0.088183
56 Memphis -0.088321
57 Tennessee -0.092068
58 U. Miami -0.102829
59 Georgia Southern -0.108260
60 Clemson -0.112420
61 Boston College -0.114806
62 Western Kentucky -0.118334
63 Boise St. -0.132960
64 Iowa -0.138208
65 Utah -0.140350
66 Texas Tech -0.158787
67 North Carolina -0.165146
68 Temple -0.168917
69 Toledo -0.180426
70 Northwestern -0.180457
71 Old Dominion -0.180576
72 West Virginia -0.185464
73 Bowling Green -0.191020
74 Indiana -0.197269
75 UTEP -0.200541
76 Florida Atlantic -0.206060
77 Washington St. -0.217462
78 Arkansas St. -0.223651
79 Syracuse -0.236155
80 Central Florida -0.249646
81 Pittsburgh -0.255739
82 South Alabama -0.260972
83 Akron -0.277040
84 Fresno St. -0.296609
85 Texas St. -0.300435
86 San Jose St. -0.316342
87 Utah St. -0.316583
88 Rice -0.320758
89 Texas-San Antonio -0.321693
90 Colorado -0.338075
91 San Diego St. -0.341123
92 Ohio -0.348163
93 South Florida -0.349975
94 Navy -0.352963
95 Michigan -0.382228
96 Eastern Michigan -0.384871
97 North Texas -0.388015
98 Western Michigan -0.404504
99 Houston -0.408412
100 Alabama-Birmingham -0.422848
101 Southern Mississippi -0.444234
102 Louisiana Tech -0.452652
103 Florida International -0.453813
104 Louisiana-Lafayette -0.505153
105 Iowa St. -0.528820
106 Buffalo -0.547783
107 Hawaii -0.553049
108 Tulsa -0.563240
109 Kansas -0.588017
110 New Mexico -0.592328
111 Purdue -0.602257
112 Army -0.607784
113 Appalachian St. -0.609570
114 Central Michigan -0.625645
115 Georgia St. -0.633307
116 Ball St. -0.666064
117 Wake Forest -0.676495
118 Kent St. -0.710634
119 New Mexico St. -0.723071
120 Southern Methodist -0.834743
121 UNLV -0.847129
122 Vanderbilt -0.888603
123 Connecticut -0.995395
124 Tulane -1.099871
125 Miami U. -1.271845
126 Massachusetts -1.315203
127 Idaho -1.329672
128 Troy -1.775871

The following list is a normal function of my formula, but it's still something I consider worth noting.  If I don't adjust these ratings, very large negative numbers are produced by certain losses, and it gives too much weight to that one week.  These teams have been given the minimum rating as opponents (not above) in place of their actual rating.  These are only teams with at least one FBS/I-A win.  More teams are added to this list as teams have more losses relative to (usually, the one) win.
Duke
Temple
Western Michigan

The following also had off-the-charts (weak) strengths of schedules that had to be adjusted:
Baylor
Duke 
Temple

I will try to break it down further, but it is a little complex.  there are three major steps to my formula.  Step 1 computes the strength of schedule (SoS).  Higher numbers indicate worse strengths of schedule, so the number I get is subtracted from a second number that is the same for every team.  Step 2 is when I use the SoS to get an opponent's rating.  This is winning percentage times (x-SoS).  If the team has at least one FBS win, I then supply a minimum rating of 1 to that team.  Step 3 is when I add and subtract numbers based on the strengths of wins and weaknesses of losses.  I never alter Step 3.  The first list above applies to Step 2 and the second applies to Step 1.  As the season goes on, typically there will be no teams that need adjustments in Step 1.

I'll lose people here, but if you want to try to understand the math... Since losses involve taking inverse numbers (this is so that I subtract less for losing to good teams), if you look at a graph of 1/x (http://www.sagemath.org/calctut/calctut-pix/onesided05.png), you would understand that a number x of less than 1 can get ridiculously large.  In short, one bad game can wipe out a whole season of good wins.  Losing to a team with an opponent's rating of 1 does enough damage.  It can make the difference of up to 60 spots this early in the year.  Later in the year it could be anywhere from two or three spots to around 30 spots depending on where in the rankings an affected team is.  It's a subtraction of 0.5, to be exact.
Comments