Alternative College Football Week 2 Predictions
I allowed the computer ratings (over?)react more to week 1 games. How much do the predictions shift for week 2's top matchups?
In a previous article, I discussed how my ratings for week 2 of college football were different if I allowed the rating system to give more weight to the results of week 1. There’s always some crazy outcomes in week 1, and it’s easy for fans to overreact to a single game. I discussed how changing the influence of week 1 games affected the ratings, some of which didn’t have too much of an impact while others resulted in some absurd overreactions. That article was already quite long without including any of the predictions from those alternative ratings, but I thought it might be worth showing how the predictions shift if week 1 games influence the ratings more.
Why Generate Alternative Ratings?
I've used my current concept for rating college football teams for three seasons now, going through various iterations along the way. One of the biggest challenges I had early on was how to handle lopsided matchups and blowout games, specifically with thresholds for what is a blowout game. That's not an issue with my current system because it's all determined from the data. The two biggest challenges, in my opinion, are calculating schedule strength and how to rate teams near the start of the season. I'll post a separate article in a week or two going into more detail about schedule strength, because that's a surprisingly complex issue.
I don't have data to explicitly show this, but it seems very intuitive that ratings and predictions are less accurate near the start of a season. Many of the roster changes happen during the offseason, and there aren't any preseason games in college football that might provide some indication about the relative strength of teams. It's possible to update the ratings using the percentage of returning production along with some subjective assessments of incoming recruits and transfers. But there's just no substitute for playing a few games. Until that happens, the ratings are based heavily on games from the prior season. That at least has some predictive power when many players return from the prior season. The real issue is how to balance the impact of games from the prior season with those during the first couple of weeks of the new season.
My current system is really conservative in this respect, and it's entirely possible that it's too conservative. Instead of weighting each game from the prior season the same as the previous season, I've also calculated some alternate ratings where each prior season game is weighted at 20%. This sounds like it mostly ignores the previous season, but that’s not really the case. A typical FBS team might have played 13 games that season, and the influence of those 13 games does add up to still have a big impact. Overall, 13 prior season games weighted at 20% each and one fully weighted game from the current season means that the current season is responsible for approximately 28% of a team's rating. The other roughly 72% is still based on games from last season. Just to see what happens, I'll probably run this alternate system for a few weeks, tapering the 20% down to 16% after week 2, then lowering it 4% in each successive week until it's completely gone after week 6.
If you want to know more about how my ratings work, I'll refer you to my articles introducing the system, explaining why games need to be weighted, and discussing a few subsequent updates to the system. The weighting I'm discussing here is about how my system handles older games from the prior season and isn't directly related to trying to rank teams from different divisions against each other.
There are definitely some shifts in the predictions, notably the Kansas-Missouri and Iowa-Iowa State games, which don't look as closely matched when the ratings give more importance to week 1 games. It's possible that these predictions might turn out to be more accurate, which is why I'm doing this experiment. If these predictions really do turn out to be more accurate, I'll adopt this as the default approach in future seasons. Of course, I may also be biased in favor of these alternative ratings because this Mizzou alumnus prefers a 72.66% probability of beating Kansas than the 61.92% from the original ratings.
The Alternative Team Ratings
Alternative Overall Ratings
Home advantage: 2.54 points
Mean score: 26.29 points
Rank Rating Team Offense Defense
1 79.68 Ohio State 36.58 43.17
2 76.97 Ole Miss 40.98 35.99
3 75.60 Notre Dame 36.34 39.35
4 75.33 Texas 33.74 41.62
5 75.30 Oregon 41.83 33.43
6 73.29 Georgia 36.57 36.69
7 71.26 Tennessee 37.28 34.05
8 70.35 Miami 42.01 28.44
9 70.22 Penn State 35.01 35.08
10 69.33 Indiana 36.29 32.99
11 69.05 USC 38.81 30.09
12 68.85 South Carolina 32.70 36.19
13 68.48 BYU 35.73 32.58
14 67.70 Iowa State 33.84 33.81
15 65.64 LSU 29.56 36.18
16 65.58 Alabama 34.49 30.81
17 65.44 Florida 32.82 32.73
18 65.12 TCU 34.75 30.26
19 64.58 Utah 29.40 34.94
20 64.53 Louisville 37.69 26.70
Rank Rating Team Offense Defense
21 63.70 Auburn 30.54 33.04
22 63.13 Texas A&M 32.85 30.33
23 62.77 Missouri 30.27 32.63
24 62.32 Arizona State 33.95 28.00
25 61.71 Clemson 30.74 30.73
26 60.40 Tulane 29.00 31.40
27 60.19 Colorado 30.36 29.71
28 59.96 Georgia Tech 28.70 30.89
29 59.64 SMU 31.67 28.23
30 59.57 Baylor 34.69 24.90
31 59.53 Oklahoma 27.10 32.58
32 58.42 Virginia Tech 25.91 32.33
33 58.09 Michigan 23.40 34.69
34 57.68 Arkansas 28.73 28.77
35 57.56 Illinois 26.26 31.05
36 57.44 Minnesota 23.04 34.39
37 57.43 Kansas 28.53 29.06
38 57.19 UCF 26.43 30.79
39 57.12 Boston College 31.16 25.75
40 56.85 Iowa 27.01 29.79
Rank Rating Team Offense Defense
41 56.68 Nebraska 23.22 33.23
42 56.26 Kansas State 29.26 27.01
43 56.01 Vanderbilt 28.60 27.28
44 55.48 Pittsburgh 32.11 23.38
45 54.60 Texas Tech 35.48 19.12
46 54.49 Cincinnati 25.39 29.07
47 54.44 Syracuse 32.49 21.84
48 54.41 Wisconsin 21.05 33.06
49 53.91 Memphis 29.54 24.37
50 53.24 Kentucky 21.11 32.04
51 53.11 California 23.41 29.48
52 52.82 Florida State 21.36 31.34
53 52.62 Virginia 24.60 28.15
54 52.61 Maryland 26.44 25.95
55 52.35 Washington 23.78 28.61
56 51.83 South Florida 26.47 25.33
57 51.39 West Virginia 28.26 23.17
58 50.94 Arizona 25.51 25.29
59 50.93 Army 23.92 27.01
60 50.89 Navy 27.19 23.80
Rank Rating Team Offense Defense
61 50.84 Rutgers 30.08 20.60
62 50.66 Houston 16.43 34.30
63 50.39 Texas State 30.39 19.87
64 49.91 UNLV 27.62 22.40
65 49.64 James Madison 23.91 25.77
66 49.59 Duke 25.15 24.32
67 48.99 UConn 27.48 21.44
68 48.68 Old Dominion 22.54 26.14
69 48.64 Oklahoma State 24.76 23.88
70 48.29 Boise State 24.95 23.32
71 47.84 Mississippi State 26.21 21.45
72 47.50 North Texas 29.53 17.97
73 47.48 Ohio 22.69 24.79
74 47.46 Western Kentucky 25.43 21.86
75 47.23 Jacksonville State 24.60 22.91
76 46.08 Marshall 24.07 22.11
77 45.98 Michigan State 16.40 29.86
78 45.29 NC State 22.84 22.30
79 45.15 UTSA 27.62 17.53
80 44.66 UCLA 20.69 23.94
Rank Rating Team Offense Defense
81 44.53 Toledo 22.00 22.51
82 44.29 Washington State 23.56 20.74
83 43.84 Fresno State 20.61 23.32
84 43.51 Northwestern 16.41 27.09
85 43.24 North Carolina 25.24 18.05
86 42.83 Louisiana 20.07 22.58
87 42.81 Miami (OH) 15.10 27.59
88 42.48 East Carolina 21.61 20.81
89 41.94 Rice 15.11 26.76
90 40.94 App State 21.32 19.57
91 40.91 South Alabama 24.24 16.73
92 40.64 Bowling Green 17.29 23.37
93 39.63 Buffalo 19.70 19.84
94 39.25 Northern Illinois 14.58 24.68
95 38.94 Air Force 15.73 23.32
96 38.56 San Diego State 17.16 21.41
97 38.54 New Mexico 28.07 10.61
98 38.48 Stanford 19.97 18.43
99 37.52 Louisiana Tech 12.40 25.14
100 37.49 Sam Houston 18.65 18.96
Rank Rating Team Offense Defense
101 37.31 San José State 17.75 19.41
102 37.28 Purdue 17.17 20.10
103 37.24 Colorado State 20.47 16.80
104 37.09 Temple 17.19 19.90
105 37.03 Wyoming 10.63 26.40
106 36.82 Liberty 16.13 20.68
107 36.79 Wake Forest 16.46 20.11
108 36.53 Utah State 23.80 12.67
109 36.53 Troy 20.61 15.91
110 36.29 Nevada 18.46 18.00
111 35.89 Oregon State 18.47 17.43
112 35.57 Arkansas State 21.05 14.48
113 35.50 Georgia Southern 19.13 16.41
114 34.97 Florida International 17.16 17.71
115 34.50 UL Monroe 14.02 20.38
116 33.97 Central Michigan 14.59 19.37
117 33.07 Western Michigan 18.89 14.47
118 31.88 Coastal Carolina 19.19 12.59
119 31.68 Florida Atlantic 16.11 15.61
120 31.61 Hawai'i 11.83 19.84
Rank Rating Team Offense Defense
121 31.29 Georgia State 18.49 12.80
122 30.93 Eastern Michigan 19.30 11.48
123 30.65 Tulsa 16.69 13.68
124 30.35 UAB 25.27 4.95
125 28.75 Charlotte 14.14 14.73
126 27.81 Akron 9.10 18.53
127 26.70 Kennesaw State 7.53 19.21
128 26.54 UTEP 10.68 15.81
129 25.46 Delaware 14.04 11.43
130 23.47 Missouri State 20.16 3.32
131 22.86 Southern Miss 10.48 12.38
132 21.72 New Mexico State 8.31 13.58
133 21.13 Ball State 12.37 8.80
134 20.03 Massachusetts 12.77 7.12
135 17.66 Middle Tennessee 9.77 7.79
136 11.92 Kent State 7.26 4.64 Predicting Week 2 With the Alternative Ratings
#1: Baylor (-2.61, 42.09%) at SMU (2.61, 57.91%)
Estimated score: 31.48 - 34.33, Total: 65.81
Quality: 96.88%, Team quality: 95.88%, Competitiveness: 98.90%
Blowout probability (margin >= 30.0 pts): 2.79%
Close game probability (margin <= 7.0 pts): 40.00%
High scoring probability (total >= 66.0 pts): 49.79%
Low scoring probability (total <= 38.0 pts): 23.02%
#2: Michigan (-3.98, 38.03%) at Oklahoma (3.98, 61.97%)
Estimated score: 15.83 - 19.96, Total: 35.80
Quality: 96.25%, Team quality: 95.66%, Competitiveness: 97.45%
Blowout probability (margin >= 30.0 pts): 3.14%
Close game probability (margin <= 7.0 pts): 39.03%
High scoring probability (total >= 66.0 pts): 21.14%
Low scoring probability (total <= 38.0 pts): 52.33%
#3: Vanderbilt (-4.95, 35.24%) at Virginia Tech (4.95, 64.76%)
Estimated score: 21.29 - 26.18, Total: 47.47
Quality: 95.49%, Team quality: 95.19%, Competitiveness: 96.08%
Blowout probability (margin >= 30.0 pts): 3.48%
Close game probability (margin <= 7.0 pts): 38.12%
High scoring probability (total >= 66.0 pts): 31.15%
Low scoring probability (total <= 38.0 pts): 40.07%
#4: West Virginia (1.37, 54.18%) at Ohio (-1.37, 45.82%)
Estimated score: 28.49 - 27.08, Total: 55.57
Quality: 94.86%, Team quality: 92.53%, Competitiveness: 99.70%
Blowout probability (margin >= 30.0 pts): 2.60%
Close game probability (margin <= 7.0 pts): 40.54%
High scoring probability (total >= 66.0 pts): 39.09%
Low scoring probability (total <= 38.0 pts): 32.05%
#5: Illinois (5.43, 66.11%) at Duke (-5.43, 33.89%)
Estimated score: 26.96 - 21.65, Total: 48.61
Quality: 94.45%, Team quality: 94.03%, Competitiveness: 95.30%
Blowout probability (margin >= 30.0 pts): 3.68%
Close game probability (margin <= 7.0 pts): 37.60%
High scoring probability (total >= 66.0 pts): 32.22%
Low scoring probability (total <= 38.0 pts): 38.91%
#6: Texas State (2.70, 58.20%) at UTSA (-2.70, 41.80%)
Estimated score: 37.88 - 35.30, Total: 73.18
Quality: 94.09%, Team quality: 91.81%, Competitiveness: 98.82%
Blowout probability (margin >= 30.0 pts): 2.81%
Close game probability (margin <= 7.0 pts): 39.94%
High scoring probability (total >= 66.0 pts): 57.56%
Low scoring probability (total <= 38.0 pts): 17.52%
#7: Kansas (-7.88, 27.34%) at Missouri (7.88, 72.66%)
Estimated score: 20.92 - 28.76, Total: 49.68
Quality: 94.01%, Team quality: 95.91%, Competitiveness: 90.30%
Blowout probability (margin >= 30.0 pts): 5.03%
Close game probability (margin <= 7.0 pts): 34.42%
High scoring probability (total >= 66.0 pts): 33.25%
Low scoring probability (total <= 38.0 pts): 37.82%
#8: Western Kentucky (0.39, 51.19%) at Toledo (-0.39, 48.81%)
Estimated score: 27.94 - 27.69, Total: 55.63
Quality: 93.91%, Team quality: 91.01%, Competitiveness: 99.98%
Blowout probability (margin >= 30.0 pts): 2.54%
Close game probability (margin <= 7.0 pts): 40.73%
High scoring probability (total >= 66.0 pts): 39.16%
Low scoring probability (total <= 38.0 pts): 31.99%
#9: Virginia (4.79, 64.29%) at NC State (-4.79, 35.71%)
Estimated score: 27.32 - 22.25, Total: 49.57
Quality: 93.61%, Team quality: 92.28%, Competitiveness: 96.33%
Blowout probability (margin >= 30.0 pts): 3.42%
Close game probability (margin <= 7.0 pts): 38.28%
High scoring probability (total >= 66.0 pts): 33.13%
Low scoring probability (total <= 38.0 pts): 37.94%
#10: Army (-7.87, 27.39%) at Kansas State (7.87, 72.61%)
Estimated score: 21.93 - 29.80, Total: 51.73
Quality: 92.74%, Team quality: 93.97%, Competitiveness: 90.35%
Blowout probability (margin >= 30.0 pts): 5.02%
Close game probability (margin <= 7.0 pts): 34.44%
High scoring probability (total >= 66.0 pts): 35.25%
Low scoring probability (total <= 38.0 pts): 35.77%
#11: UConn (-7.99, 27.08%) at Syracuse (7.99, 72.92%)
Estimated score: 30.66 - 38.61, Total: 69.27
Quality: 92.20%, Team quality: 93.29%, Competitiveness: 90.06%
Blowout probability (margin >= 30.0 pts): 5.10%
Close game probability (margin <= 7.0 pts): 34.26%
High scoring probability (total >= 66.0 pts): 53.46%
Low scoring probability (total <= 38.0 pts): 20.33%
#12: Houston (6.18, 68.18%) at Rice (-6.18, 31.82%)
Estimated score: 14.69 - 8.37, Total: 23.06
Quality: 92.01%, Team quality: 91.06%, Competitiveness: 93.95%
Blowout probability (margin >= 30.0 pts): 4.03%
Close game probability (margin <= 7.0 pts): 36.73%
High scoring probability (total >= 66.0 pts): 12.72%
Low scoring probability (total <= 38.0 pts): 65.42%
#13: Boston College (8.60, 74.43%) at Michigan State (-8.60, 25.57%)
Estimated score: 26.32 - 18.21, Total: 44.53
Quality: 91.63%, Team quality: 93.20%, Competitiveness: 88.56%
Blowout probability (margin >= 30.0 pts): 5.54%
Close game probability (margin <= 7.0 pts): 33.34%
High scoring probability (total >= 66.0 pts): 28.43%
Low scoring probability (total <= 38.0 pts): 43.12%
#14: UCLA (-7.79, 27.59%) at UNLV (7.79, 72.41%)
Estimated score: 23.30 - 31.24, Total: 54.54
Quality: 91.15%, Team quality: 91.45%, Competitiveness: 90.54%
Blowout probability (margin >= 30.0 pts): 4.97%
Close game probability (margin <= 7.0 pts): 34.56%
High scoring probability (total >= 66.0 pts): 38.05%
Low scoring probability (total <= 38.0 pts): 33.03%
#15: Fresno State (5.41, 66.06%) at Oregon State (-5.41, 33.94%)
Estimated score: 28.20 - 22.70, Total: 50.91
Quality: 90.15%, Team quality: 87.67%, Competitiveness: 95.33%
Blowout probability (margin >= 30.0 pts): 3.67%
Close game probability (margin <= 7.0 pts): 37.62%
High scoring probability (total >= 66.0 pts): 34.44%
Low scoring probability (total <= 38.0 pts): 36.59%
#16: Arizona State (11.95, 81.87%) at Mississippi State (-11.95, 18.13%)
Estimated score: 37.52 - 25.77, Total: 63.29
Quality: 88.86%, Team quality: 94.30%, Competitiveness: 78.89%
Blowout probability (margin >= 30.0 pts): 8.67%
Close game probability (margin <= 7.0 pts): 27.69%
High scoring probability (total >= 66.0 pts): 47.13%
Low scoring probability (total <= 38.0 pts): 25.11%
#17: San Diego State (-8.27, 26.38%) at Washington State (8.27, 73.62%)
Estimated score: 21.44 - 29.71, Total: 51.16
Quality: 88.77%, Team quality: 88.46%, Competitiveness: 89.38%
Blowout probability (margin >= 30.0 pts): 5.30%
Close game probability (margin <= 7.0 pts): 33.85%
High scoring probability (total >= 66.0 pts): 34.68%
Low scoring probability (total <= 38.0 pts): 36.35%
#18: Sam Houston (3.35, 60.12%) at Hawai'i (-3.35, 39.88%)
Estimated score: 23.83 - 20.43, Total: 44.26
Quality: 88.69%, Team quality: 84.29%, Competitiveness: 98.19%
Blowout probability (margin >= 30.0 pts): 2.96%
Close game probability (margin <= 7.0 pts): 39.52%
High scoring probability (total >= 66.0 pts): 28.19%
Low scoring probability (total <= 38.0 pts): 43.40%
#19: Miami (OH) (-10.57, 21.01%) at Rutgers (10.57, 78.99%)
Estimated score: 19.52 - 30.04, Total: 49.56
Quality: 88.38%, Team quality: 91.11%, Competitiveness: 83.14%
Blowout probability (margin >= 30.0 pts): 7.22%
Close game probability (margin <= 7.0 pts): 30.11%
High scoring probability (total >= 66.0 pts): 33.13%
Low scoring probability (total <= 38.0 pts): 37.95%
#20: Iowa (-13.40, 15.39%) at Iowa State (13.40, 84.61%)
Estimated score: 18.21 - 31.61, Total: 49.82
Quality: 88.21%, Team quality: 96.25%, Competitiveness: 74.09%
Blowout probability (margin >= 30.0 pts): 10.45%
Close game probability (margin <= 7.0 pts): 25.09%
High scoring probability (total >= 66.0 pts): 33.38%
Low scoring probability (total <= 38.0 pts): 37.68%
#21: UT Martin (1.63, 54.97%) at UTEP (-1.63, 45.03%)
Estimated score: 22.58 - 20.98, Total: 43.56
Quality: 85.92%, Team quality: 79.81%, Competitiveness: 99.57%
Blowout probability (margin >= 30.0 pts): 2.63%
Close game probability (margin <= 7.0 pts): 40.45%
High scoring probability (total >= 66.0 pts): 27.57%
Low scoring probability (total <= 38.0 pts): 44.13%
#22: North Texas (11.90, 81.77%) at Western Michigan (-11.90, 18.23%)
Estimated score: 40.08 - 28.47, Total: 68.55
Quality: 84.64%, Team quality: 87.58%, Competitiveness: 79.05%
Blowout probability (margin >= 30.0 pts): 8.61%
Close game probability (margin <= 7.0 pts): 27.78%
High scoring probability (total >= 66.0 pts): 52.69%
Low scoring probability (total <= 38.0 pts): 20.87%
#23: Liberty (-12.95, 16.20%) at Jacksonville State (12.95, 83.80%)
Estimated score: 18.25 - 31.47, Total: 49.72
Quality: 83.98%, Team quality: 88.51%, Competitiveness: 75.60%
Blowout probability (margin >= 30.0 pts): 9.87%
Close game probability (margin <= 7.0 pts): 25.89%
High scoring probability (total >= 66.0 pts): 33.28%
Low scoring probability (total <= 38.0 pts): 37.79%
#24: South Florida (-16.15, 11.00%) at Florida (16.15, 89.00%)
Estimated score: 18.76 - 35.05, Total: 53.81
Quality: 83.54%, Team quality: 95.17%, Competitiveness: 64.36%
Blowout probability (margin >= 30.0 pts): 14.63%
Close game probability (margin <= 7.0 pts): 20.18%
High scoring probability (total >= 66.0 pts): 37.31%
Low scoring probability (total <= 38.0 pts): 33.74%
#25: Jackson State (1.02, 53.11%) at Southern Miss (-1.02, 46.89%)
Estimated score: 24.16 - 23.27, Total: 47.42
Quality: 83.52%, Team quality: 76.39%, Competitiveness: 99.83%
Blowout probability (margin >= 30.0 pts): 2.57%
Close game probability (margin <= 7.0 pts): 40.63%
High scoring probability (total >= 66.0 pts): 31.10%
Low scoring probability (total <= 38.0 pts): 40.13%
#26: North Carolina (11.95, 81.88%) at Charlotte (-11.95, 18.12%)
Estimated score: 35.53 - 23.65, Total: 59.17
Quality: 82.81%, Team quality: 84.86%, Competitiveness: 78.87%
Blowout probability (margin >= 30.0 pts): 8.67%
Close game probability (margin <= 7.0 pts): 27.68%
High scoring probability (total >= 66.0 pts): 42.81%
Low scoring probability (total <= 38.0 pts): 28.71%
#27: Tulsa (6.39, 68.75%) at New Mexico State (-6.39, 31.25%)
Estimated score: 28.13 - 22.18, Total: 50.31
Quality: 82.60%, Team quality: 77.62%, Competitiveness: 93.54%
Blowout probability (margin >= 30.0 pts): 4.14%
Close game probability (margin <= 7.0 pts): 36.46%
High scoring probability (total >= 66.0 pts): 33.85%
Low scoring probability (total <= 38.0 pts): 37.19%
#28: Sacramento State (-11.44, 19.15%) at Nevada (11.44, 80.85%)
Estimated score: 20.50 - 32.25, Total: 52.75
Quality: 81.43%, Team quality: 81.91%, Competitiveness: 80.48%
Blowout probability (margin >= 30.0 pts): 8.11%
Close game probability (margin <= 7.0 pts): 28.58%
High scoring probability (total >= 66.0 pts): 36.26%
Low scoring probability (total <= 38.0 pts): 34.77%
#29: James Madison (-17.43, 9.31%) at Louisville (17.43, 90.69%)
Estimated score: 22.22 - 39.48, Total: 61.70
Quality: 81.14%, Team quality: 94.63%, Competitiveness: 59.66%
Blowout probability (margin >= 30.0 pts): 16.94%
Close game probability (margin <= 7.0 pts): 18.00%
High scoring probability (total >= 66.0 pts): 45.45%
Low scoring probability (total <= 38.0 pts): 26.47%
#30: Northern Illinois (-15.89, 11.36%) at Maryland (15.89, 88.64%)
Estimated score: 13.64 - 29.32, Total: 42.96
Quality: 81.05%, Team quality: 90.32%, Competitiveness: 65.27%
Blowout probability (margin >= 30.0 pts): 14.20%
Close game probability (margin <= 7.0 pts): 20.62%
High scoring probability (total >= 66.0 pts): 27.04%
Low scoring probability (total <= 38.0 pts): 44.77%
#31: Bowling Green (-16.39, 10.66%) at Cincinnati (16.39, 89.34%)
Estimated score: 13.24 - 29.57, Total: 42.81
Quality: 80.74%, Team quality: 91.06%, Competitiveness: 63.47%
Blowout probability (margin >= 30.0 pts): 15.05%
Close game probability (margin <= 7.0 pts): 19.76%
High scoring probability (total >= 66.0 pts): 26.91%
Low scoring probability (total <= 38.0 pts): 44.93%
#32: Tulane (16.96, 90.09%) at South Alabama (-16.96, 9.91%)
Estimated score: 37.29 - 20.40, Total: 57.68
Quality: 80.61%, Team quality: 92.36%, Competitiveness: 61.40%
Blowout probability (margin >= 30.0 pts): 16.06%
Close game probability (margin <= 7.0 pts): 18.79%
High scoring probability (total >= 66.0 pts): 41.26%
Low scoring probability (total <= 38.0 pts): 30.07%
#33: Southern Illinois (-12.76, 16.55%) at Purdue (12.76, 83.45%)
Estimated score: 18.95 - 31.74, Total: 50.69
Quality: 80.06%, Team quality: 82.05%, Competitiveness: 76.22%
Blowout probability (margin >= 30.0 pts): 9.64%
Close game probability (margin <= 7.0 pts): 26.23%
High scoring probability (total >= 66.0 pts): 34.22%
Low scoring probability (total <= 38.0 pts): 36.81%
#34: Idaho State (-14.50, 13.51%) at New Mexico (14.50, 86.49%)
Estimated score: 34.49 - 49.18, Total: 83.67
Quality: 78.00%, Team quality: 82.18%, Competitiveness: 70.27%
Blowout probability (margin >= 30.0 pts): 12.00%
Close game probability (margin <= 7.0 pts): 23.10%
High scoring probability (total >= 66.0 pts): 68.05%
Low scoring probability (total <= 38.0 pts): 11.27%
#35: Ole Miss (21.19, 94.52%) at Kentucky (-21.19, 5.48%)
Estimated score: 33.96 - 12.68, Total: 46.64
Quality: 75.36%, Team quality: 96.50%, Competitiveness: 45.96%
Blowout probability (margin >= 30.0 pts): 25.06%
Close game probability (margin <= 7.0 pts): 12.26%
High scoring probability (total >= 66.0 pts): 30.36%
Low scoring probability (total <= 38.0 pts): 40.93%
#36: Memphis (20.09, 93.57%) at Georgia State (-20.09, 6.43%)
Estimated score: 41.76 - 21.67, Total: 63.44
Quality: 72.93%, Team quality: 88.16%, Competitiveness: 49.91%
Blowout probability (margin >= 30.0 pts): 22.48%
Close game probability (margin <= 7.0 pts): 13.82%
High scoring probability (total >= 66.0 pts): 47.29%
Low scoring probability (total <= 38.0 pts): 24.98%
#37: Northern Iowa (-17.33, 9.43%) at Wyoming (17.33, 90.57%)
Estimated score: 10.16 - 27.52, Total: 37.68
Quality: 72.44%, Team quality: 79.59%, Competitiveness: 60.01%
Blowout probability (margin >= 30.0 pts): 16.76%
Close game probability (margin <= 7.0 pts): 18.16%
High scoring probability (total >= 66.0 pts): 22.61%
Low scoring probability (total <= 38.0 pts): 50.34%
#38: UC Davis (-21.23, 5.44%) at Washington (21.23, 94.56%)
Estimated score: 9.88 - 31.16, Total: 41.05
Quality: 70.93%, Team quality: 88.25%, Competitiveness: 45.81%
Blowout probability (margin >= 30.0 pts): 25.16%
Close game probability (margin <= 7.0 pts): 12.21%
High scoring probability (total >= 66.0 pts): 25.39%
Low scoring probability (total <= 38.0 pts): 46.78%
#39: Western Carolina (-18.33, 8.25%) at Wake Forest (18.33, 91.75%)
Estimated score: 25.13 - 43.52, Total: 68.65
Quality: 70.51%, Team quality: 78.86%, Competitiveness: 56.36%
Blowout probability (margin >= 30.0 pts): 18.70%
Close game probability (margin <= 7.0 pts): 16.53%
High scoring probability (total >= 66.0 pts): 52.81%
Low scoring probability (total <= 38.0 pts): 20.79%
#40: Arkansas State (-22.11, 4.77%) vs. Arkansas (22.11, 95.23%)
Estimated score: 18.56 - 40.54, Total: 59.10
Quality: 70.24%, Team quality: 90.05%, Competitiveness: 42.73%
Blowout probability (margin >= 30.0 pts): 27.33%
Close game probability (margin <= 7.0 pts): 11.05%
High scoring probability (total >= 66.0 pts): 42.73%
Low scoring probability (total <= 38.0 pts): 28.77%
#41: UAB (-23.08, 4.10%) at Navy (23.08, 95.90%)
Estimated score: 26.49 - 49.80, Total: 76.29
Quality: 66.62%, Team quality: 86.63%, Competitiveness: 39.40%
Blowout probability (margin >= 30.0 pts): 29.85%
Close game probability (margin <= 7.0 pts): 9.85%
High scoring probability (total >= 66.0 pts): 60.76%
Low scoring probability (total <= 38.0 pts): 15.47%
#42: Central Michigan (-24.05, 3.52%) at Pittsburgh (24.05, 96.48%)
Estimated score: 16.23 - 40.30, Total: 56.53
Quality: 65.87%, Team quality: 88.85%, Competitiveness: 36.21%
Blowout probability (margin >= 30.0 pts): 32.45%
Close game probability (margin <= 7.0 pts): 8.75%
High scoring probability (total >= 66.0 pts): 40.07%
Low scoring probability (total <= 38.0 pts): 31.14%
#43: Bryant (-15.64, 11.73%) at Massachusetts (15.64, 88.27%)
Estimated score: 17.91 - 33.68, Total: 51.59
Quality: 65.00%, Team quality: 64.40%, Competitiveness: 66.20%
Blowout probability (margin >= 30.0 pts): 13.78%
Close game probability (margin <= 7.0 pts): 21.07%
High scoring probability (total >= 66.0 pts): 35.10%
Low scoring probability (total <= 38.0 pts): 35.92%
#44: Long Island University (-20.10, 6.42%) at Eastern Michigan (20.10, 93.58%)
Estimated score: 21.73 - 41.52, Total: 63.25
Quality: 64.06%, Team quality: 72.62%, Competitiveness: 49.86%
Blowout probability (margin >= 30.0 pts): 22.52%
Close game probability (margin <= 7.0 pts): 13.80%
High scoring probability (total >= 66.0 pts): 47.09%
Low scoring probability (total <= 38.0 pts): 25.14%
#45: McNeese (-23.76, 3.69%) at Louisiana (23.76, 96.31%)
Estimated score: 11.87 - 35.41, Total: 47.28
Quality: 62.34%, Team quality: 80.74%, Competitiveness: 37.15%
Blowout probability (margin >= 30.0 pts): 31.66%
Close game probability (margin <= 7.0 pts): 9.07%
High scoring probability (total >= 66.0 pts): 30.96%
Low scoring probability (total <= 38.0 pts): 40.28%
#46: Eastern Washington (-24.97, 3.03%) at Boise State (24.97, 96.97%)
Estimated score: 21.51 - 46.42, Total: 67.93
Quality: 61.72%, Team quality: 84.08%, Competitiveness: 33.26%
Blowout probability (margin >= 30.0 pts): 35.03%
Close game probability (margin <= 7.0 pts): 7.78%
High scoring probability (total >= 66.0 pts): 52.05%
Low scoring probability (total <= 38.0 pts): 21.34%
#47: Missouri State (-25.14, 2.95%) at Marshall (25.14, 97.05%)
Estimated score: 23.07 - 48.31, Total: 71.38
Quality: 60.59%, Team quality: 82.43%, Competitiveness: 32.73%
Blowout probability (margin >= 30.0 pts): 35.50%
Close game probability (margin <= 7.0 pts): 7.61%
High scoring probability (total >= 66.0 pts): 55.68%
Low scoring probability (total <= 38.0 pts): 18.78%
#48: Troy (-27.72, 1.90%) at Clemson (27.72, 98.10%)
Estimated score: 14.90 - 42.38, Total: 57.28
Quality: 59.16%, Team quality: 90.56%, Competitiveness: 25.24%
Blowout probability (margin >= 30.0 pts): 43.09%
Close game probability (margin <= 7.0 pts): 5.35%
High scoring probability (total >= 66.0 pts): 40.85%
Low scoring probability (total <= 38.0 pts): 30.44%
#49: Oklahoma State (-29.20, 1.46%) at Oregon (29.20, 98.54%)
Estimated score: 16.35 - 45.51, Total: 61.86
Quality: 57.90%, Team quality: 95.12%, Competitiveness: 21.46%
Blowout probability (margin >= 30.0 pts): 47.56%
Close game probability (margin <= 7.0 pts): 4.32%
High scoring probability (total >= 66.0 pts): 45.62%
Low scoring probability (total <= 38.0 pts): 26.33%
#50: Utah State (-29.14, 1.48%) at Texas A&M (29.14, 98.52%)
Estimated score: 18.49 - 47.74, Total: 66.23
Quality: 56.22%, Team quality: 90.70%, Competitiveness: 21.60%
Blowout probability (margin >= 30.0 pts): 47.38%
Close game probability (margin <= 7.0 pts): 4.36%
High scoring probability (total >= 66.0 pts): 50.24%
Low scoring probability (total <= 38.0 pts): 22.68%
#51: Florida A&M (-24.26, 3.40%) at Florida Atlantic (24.26, 96.60%)
Estimated score: 9.26 - 33.63, Total: 42.89
Quality: 56.18%, Team quality: 70.67%, Competitiveness: 35.51%
Blowout probability (margin >= 30.0 pts): 33.04%
Close game probability (margin <= 7.0 pts): 8.51%
High scoring probability (total >= 66.0 pts): 26.98%
Low scoring probability (total <= 38.0 pts): 44.84%
#52: Charleston Southern (-24.65, 3.19%) at Coastal Carolina (24.65, 96.81%)
Estimated score: 12.25 - 36.94, Total: 49.19
Quality: 55.48%, Team quality: 70.60%, Competitiveness: 34.26%
Blowout probability (margin >= 30.0 pts): 34.13%
Close game probability (margin <= 7.0 pts): 8.10%
High scoring probability (total >= 66.0 pts): 32.77%
Low scoring probability (total <= 38.0 pts): 38.32%
#53: Louisiana Tech (-30.65, 1.12%) at LSU (30.65, 98.88%)
Estimated score: 1.23 - 31.98, Total: 33.21
Quality: 53.24%, Team quality: 91.27%, Competitiveness: 18.12%
Blowout probability (margin >= 30.0 pts): 51.99%
Close game probability (margin <= 7.0 pts): 3.47%
High scoring probability (total >= 66.0 pts): 19.21%
Low scoring probability (total <= 38.0 pts): 55.06%
#54: Akron (-31.40, 0.98%) at Nebraska (31.40, 99.02%)
Estimated score: 0.89 - 32.24, Total: 33.13
Quality: 49.75%, Team quality: 86.32%, Competitiveness: 16.52%
Blowout probability (margin >= 30.0 pts): 54.28%
Close game probability (margin <= 7.0 pts): 3.08%
High scoring probability (total >= 66.0 pts): 19.15%
Low scoring probability (total <= 38.0 pts): 55.14%
#55: Stanford (-32.54, 0.79%) at BYU (32.54, 99.21%)
Estimated score: 12.41 - 44.86, Total: 57.27
Quality: 49.41%, Team quality: 91.82%, Competitiveness: 14.31%
Blowout probability (margin >= 30.0 pts): 57.70%
Close game probability (margin <= 7.0 pts): 2.57%
High scoring probability (total >= 66.0 pts): 40.83%
Low scoring probability (total <= 38.0 pts): 30.45%
#56: North Carolina Central (-31.07, 1.04%) at Old Dominion (31.07, 98.96%)
Estimated score: 7.24 - 38.54, Total: 45.79
Quality: 48.37%, Team quality: 81.10%, Competitiveness: 17.21%
Blowout probability (margin >= 30.0 pts): 53.27%
Close game probability (margin <= 7.0 pts): 3.25%
High scoring probability (total >= 66.0 pts): 29.58%
Low scoring probability (total <= 38.0 pts): 41.81%
#57: Weber State (-31.88, 0.89%) at Arizona (31.88, 99.11%)
Estimated score: 12.74 - 44.48, Total: 57.22
Quality: 47.25%, Team quality: 82.30%, Competitiveness: 15.57%
Blowout probability (margin >= 30.0 pts): 55.71%
Close game probability (margin <= 7.0 pts): 2.86%
High scoring probability (total >= 66.0 pts): 40.79%
Low scoring probability (total <= 38.0 pts): 30.49%
#58: UL Monroe (-33.61, 0.64%) at Alabama (33.61, 99.36%)
Estimated score: 8.22 - 41.67, Total: 49.89
Quality: 46.55%, Team quality: 90.14%, Competitiveness: 12.42%
Blowout probability (margin >= 30.0 pts): 60.89%
Close game probability (margin <= 7.0 pts): 2.15%
High scoring probability (total >= 66.0 pts): 33.44%
Low scoring probability (total <= 38.0 pts): 37.62%
#59: Georgia Southern (-36.09, 0.39%) at USC (36.09, 99.61%)
Estimated score: 14.05 - 49.95, Total: 64.00
Quality: 41.64%, Team quality: 90.78%, Competitiveness: 8.76%
Blowout probability (margin >= 30.0 pts): 67.92%
Close game probability (margin <= 7.0 pts): 1.41%
High scoring probability (total >= 66.0 pts): 47.89%
Low scoring probability (total <= 38.0 pts): 24.50%
#60: Campbell (-33.49, 0.66%) at East Carolina (33.49, 99.34%)
Estimated score: 11.45 - 45.11, Total: 56.56
Quality: 41.24%, Team quality: 74.53%, Competitiveness: 12.63%
Blowout probability (margin >= 30.0 pts): 60.52%
Close game probability (margin <= 7.0 pts): 2.20%
High scoring probability (total >= 66.0 pts): 40.11%
Low scoring probability (total <= 38.0 pts): 31.11%
#61: Howard (-32.88, 0.74%) at Temple (32.88, 99.26%)
Estimated score: 3.21 - 35.95, Total: 39.16
Quality: 40.62%, Team quality: 69.97%, Competitiveness: 13.68%
Blowout probability (margin >= 30.0 pts): 58.72%
Close game probability (margin <= 7.0 pts): 2.43%
High scoring probability (total >= 66.0 pts): 23.81%
Low scoring probability (total <= 38.0 pts): 48.78%
#62: Florida International (-37.79, 0.28%) at Penn State (37.79, 99.72%)
Estimated score: 7.10 - 44.86, Total: 51.95
Quality: 38.17%, Team quality: 90.65%, Competitiveness: 6.77%
Blowout probability (margin >= 30.0 pts): 72.42%
Close game probability (margin <= 7.0 pts): 1.04%
High scoring probability (total >= 66.0 pts): 35.46%
Low scoring probability (total <= 38.0 pts): 35.56%
#63: Lindenwood (-34.75, 0.51%) at App State (34.75, 99.49%)
Estimated score: 9.75 - 44.49, Total: 54.24
Quality: 38.13%, Team quality: 72.28%, Competitiveness: 10.61%
Blowout probability (margin >= 30.0 pts): 64.19%
Close game probability (margin <= 7.0 pts): 1.78%
High scoring probability (total >= 66.0 pts): 37.75%
Low scoring probability (total <= 38.0 pts): 33.32%
#64: Delaware (-37.27, 0.31%) at Colorado (37.27, 99.69%)
Estimated score: 9.34 - 46.49, Total: 55.83
Quality: 37.74%, Team quality: 85.57%, Competitiveness: 7.34%
Blowout probability (margin >= 30.0 pts): 71.07%
Close game probability (margin <= 7.0 pts): 1.14%
High scoring probability (total >= 66.0 pts): 39.36%
Low scoring probability (total <= 38.0 pts): 31.80%
#65: Western Illinois (-35.68, 0.42%) at Northwestern (35.68, 99.58%)
Estimated score: 12.45 - 48.14, Total: 60.58
Quality: 37.04%, Team quality: 73.96%, Competitiveness: 9.29%
Blowout probability (margin >= 30.0 pts): 66.81%
Close game probability (margin <= 7.0 pts): 1.51%
High scoring probability (total >= 66.0 pts): 44.28%
Low scoring probability (total <= 38.0 pts): 27.44%
#66: Northern Colorado (-35.16, 0.47%) at Colorado State (35.16, 99.53%)
Estimated score: 5.61 - 41.05, Total: 46.66
Quality: 36.08%, Team quality: 68.47%, Competitiveness: 10.01%
Blowout probability (margin >= 30.0 pts): 65.35%
Close game probability (margin <= 7.0 pts): 1.65%
High scoring probability (total >= 66.0 pts): 30.39%
Low scoring probability (total <= 38.0 pts): 40.91%
#67: St. Francis (PA) (-35.94, 0.40%) at Buffalo (35.94, 99.60%)
Estimated score: 0.46 - 36.31, Total: 36.77
Quality: 35.34%, Team quality: 70.21%, Competitiveness: 8.95%
Blowout probability (margin >= 30.0 pts): 67.51%
Close game probability (margin <= 7.0 pts): 1.44%
High scoring probability (total >= 66.0 pts): 21.89%
Low scoring probability (total <= 38.0 pts): 51.30%
#68: East Texas A&M (-38.63, 0.23%) at Florida State (38.63, 99.77%)
Estimated score: 2.69 - 41.27, Total: 43.96
Quality: 33.52%, Team quality: 79.72%, Competitiveness: 5.93%
Blowout probability (margin >= 30.0 pts): 74.52%
Close game probability (margin <= 7.0 pts): 0.89%
High scoring probability (total >= 66.0 pts): 27.93%
Low scoring probability (total <= 38.0 pts): 43.71%
#69: San José State (-40.55, 0.15%) at Texas (40.55, 99.85%)
Estimated score: 1.15 - 41.89, Total: 43.05
Quality: 33.13%, Team quality: 91.79%, Competitiveness: 4.31%
Blowout probability (margin >= 30.0 pts): 78.97%
Close game probability (margin <= 7.0 pts): 0.62%
High scoring probability (total >= 66.0 pts): 27.12%
Low scoring probability (total <= 38.0 pts): 44.67%
#70: Middle Tennessee (-39.30, 0.20%) at Wisconsin (39.30, 99.80%)
Estimated score: 1.73 - 40.82, Total: 42.55
Quality: 32.55%, Team quality: 80.51%, Competitiveness: 5.32%
Blowout probability (margin >= 30.0 pts): 76.11%
Close game probability (margin <= 7.0 pts): 0.78%
High scoring probability (total >= 66.0 pts): 26.68%
Low scoring probability (total <= 38.0 pts): 45.19%
#71: Gardner-Webb (-41.85, 0.12%) at Georgia Tech (41.85, 99.88%)
Estimated score: 8.70 - 50.14, Total: 58.84
Quality: 28.73%, Team quality: 82.93%, Competitiveness: 3.45%
Blowout probability (margin >= 30.0 pts): 81.68%
Close game probability (margin <= 7.0 pts): 0.48%
High scoring probability (total >= 66.0 pts): 42.46%
Low scoring probability (total <= 38.0 pts): 29.01%
#72: Kennesaw State (-45.17, 0.05%) at Indiana (45.17, 99.95%)
Estimated score: 0.00 - 44.63, Total: 44.63
Quality: 24.15%, Team quality: 86.94%, Competitiveness: 1.86%
Blowout probability (margin >= 30.0 pts): 87.56%
Close game probability (margin <= 7.0 pts): 0.24%
High scoring probability (total >= 66.0 pts): 28.52%
Low scoring probability (total <= 38.0 pts): 43.02%
#73: Ball State (-45.11, 0.06%) at Auburn (45.11, 99.94%)
Estimated score: 4.35 - 49.30, Total: 53.65
Quality: 23.61%, Team quality: 83.55%, Competitiveness: 1.89%
Blowout probability (margin >= 30.0 pts): 87.46%
Close game probability (margin <= 7.0 pts): 0.25%
High scoring probability (total >= 66.0 pts): 37.15%
Low scoring probability (total <= 38.0 pts): 33.89%
#74: Cal Poly (-45.59, 0.05%) at Utah (45.59, 99.95%)
Estimated score: 2.45 - 47.68, Total: 50.13
Quality: 22.94%, Team quality: 83.85%, Competitiveness: 1.72%
Blowout probability (margin >= 30.0 pts): 88.19%
Close game probability (margin <= 7.0 pts): 0.22%
High scoring probability (total >= 66.0 pts): 33.68%
Low scoring probability (total <= 38.0 pts): 37.37%
#75: Kent State (-45.22, 0.05%) at Texas Tech (45.22, 99.95%)
Estimated score: 13.17 - 58.40, Total: 71.56
Quality: 22.17%, Team quality: 76.82%, Competitiveness: 1.85%
Blowout probability (margin >= 30.0 pts): 87.63%
Close game probability (margin <= 7.0 pts): 0.24%
High scoring probability (total >= 66.0 pts): 55.87%
Low scoring probability (total <= 38.0 pts): 18.65%
#76: East Tennessee State (-47.33, 0.03%) at Tennessee (47.33, 99.97%)
Estimated score: 3.91 - 51.21, Total: 55.12
Quality: 20.91%, Team quality: 86.93%, Competitiveness: 1.21%
Blowout probability (margin >= 30.0 pts): 90.57%
Close game probability (margin <= 7.0 pts): 0.15%
High scoring probability (total >= 66.0 pts): 38.64%
Low scoring probability (total <= 38.0 pts): 32.47%
#77: Austin Peay (-47.85, 0.03%) at Georgia (47.85, 99.97%)
Estimated score: 0.00 - 47.54, Total: 47.54
Quality: 20.31%, Team quality: 87.78%, Competitiveness: 1.09%
Blowout probability (margin >= 30.0 pts): 91.20%
Close game probability (margin <= 7.0 pts): 0.14%
High scoring probability (total >= 66.0 pts): 31.21%
Low scoring probability (total <= 38.0 pts): 40.00%
#78: South Carolina State (-48.16, 0.03%) at South Carolina (48.16, 99.97%)
Estimated score: 1.19 - 49.46, Total: 50.65
Quality: 19.46%, Team quality: 85.09%, Competitiveness: 1.02%
Blowout probability (margin >= 30.0 pts): 91.57%
Close game probability (margin <= 7.0 pts): 0.13%
High scoring probability (total >= 66.0 pts): 34.19%
Low scoring probability (total <= 38.0 pts): 36.85%
#79: North Carolina A&T (-55.51, 0.00%) at UCF (55.51, 100.00%)
Estimated score: 0.00 - 52.63, Total: 52.63
Quality: 9.86%, Team quality: 71.49%, Competitiveness: 0.19%
Blowout probability (margin >= 30.0 pts): 97.23%
Close game probability (margin <= 7.0 pts): 0.02%
High scoring probability (total >= 66.0 pts): 36.13%
Low scoring probability (total <= 38.0 pts): 34.89%
#80: Texas Southern (-59.82, 0.00%) at California (59.82, 100.00%)
Estimated score: 0.00 - 51.91, Total: 51.91
Quality: 6.30%, Team quality: 64.24%, Competitiveness: 0.06%
Blowout probability (margin >= 30.0 pts): 98.69%
Close game probability (margin <= 7.0 pts): 0.01%
High scoring probability (total >= 66.0 pts): 35.42%
Low scoring probability (total <= 38.0 pts): 35.60%
#81: Northwestern State (-66.03, 0.00%) at Minnesota (66.03, 100.00%)
Estimated score: 0.00 - 53.97, Total: 53.97
Quality: 3.41%, Team quality: 63.07%, Competitiveness: 0.01%
Blowout probability (margin >= 30.0 pts): 99.60%
Close game probability (margin <= 7.0 pts): 0.00%
High scoring probability (total >= 66.0 pts): 37.47%
Low scoring probability (total <= 38.0 pts): 33.59%
#82: Bethune-Cookman (-70.84, 0.00%) at Miami (70.84, 100.00%)
Estimated score: 0.00 - 69.28, Total: 69.28
Quality: 2.19%, Team quality: 70.71%, Competitiveness: 0.00%
Blowout probability (margin >= 30.0 pts): 99.86%
Close game probability (margin <= 7.0 pts): 0.00%
High scoring probability (total >= 66.0 pts): 53.47%
Low scoring probability (total <= 38.0 pts): 20.32%
#83: Grambling (-75.19, 0.00%) at Ohio State (75.19, 100.00%)
Estimated score: 0.00 - 60.35, Total: 60.35
Quality: 1.38%, Team quality: 74.95%, Competitiveness: 0.00%
Blowout probability (margin >= 30.0 pts): 99.95%
Close game probability (margin <= 7.0 pts): 0.00%
High scoring probability (total >= 66.0 pts): 44.04%
Low scoring probability (total <= 38.0 pts): 27.65%As always, games are ranked based on the projected quality. This factors in the overall strength of the two teams and the potential for a competitive game. Game quality ratings are not directly comparable between college football and the NFL. NFL games are typically decided by smaller margins than college games, the teams are more balanced in their quality, and there's just not as much scoring in the NFL. Thresholds for close games and blowouts are also different between college and the NFL for the same reasons.
Beside each team, there are two numbers in parentheses. One is the predicted margin of victory (positive) or defeat (negative), the other is the probability of winning. These margins are sometimes larger than what's indicated by the predicted score. That's because there's nothing in the math that prevents a prediction of negative points with a sufficiently lopsided matchup. This is, of course, impossible, so the score is set to zero in those instances. There's no cap on how many points a team can be projected to score, though.
NFL ratings and week 1 predictions will be posted Thursday morning, though I’ll still have to consider how to weight 2024 games after week 1. The NFL is also more complex because of also weighting 2025 preseason games. But I’ll save that discussion for future articles.
This article uses data from collegefootballdata.com.