Ken's NFL Forecasts
Model Description
My model is a neural network-based classifier,
designed to
estimate the probability of a particular NFL
team winning a matchup.
The dataset used to build the model is provided by Pro Football
Reference.
Week 3 Odds
Pick Distribution (based on sportsbook odds): 8 Favorites, 2 Underdogs
Model Favorite
|
Model Odds
|
Book
Odds
|
Edge
|
Model Underdog
|
Model Odds
|
Book
Odds
|
Edge
|
| Buffalo Bills |
-456 |
-245
|
11%
|
Washington Commanders |
+456 |
+200
|
-15%
|
| Miami Dolphins |
-203 |
-250
|
-4%
|
Denver Broncos |
+203 |
+205
|
0%
|
| Jacksonville
Jaguars |
-669 |
-345
|
9%
|
Houston Texans |
+669 |
+275
|
-14%
|
| Detroit Lions |
-317 |
-166
|
14%
|
Atlanta Falcons |
+317 |
+140
|
-18%
|
| Baltimore Ravens |
-400 |
-360
|
2%
|
Indianapolis Colts |
+400 |
+285
|
-6%
|
| New England Patriots |
-108 |
-142
|
-7%
|
New York Jets |
+108 |
+120
|
3%
|
| Minnesota Vikings |
-233 |
-118
|
16%
|
Los Angeles Chargers |
+233 |
-102
|
-20%
|
| Cleveland Browns |
-100 |
-166
|
-12%
|
Tennessee Titans |
+100 |
+140
|
8%
|
| Green Bay Packers |
-150 |
-118
|
6%
|
New Orleans Saints |
+150 |
-102
|
-10%
|
| Seattle Seahawks |
-163 |
-218
|
-7%
|
Carolina Panthers |
+163 |
+180
|
2%
|
| Dallas Cowboys |
-614 |
-675
|
-1%
|
Arizona Cardinals |
+614 |
+490
|
-3%
|
| Kansas City
Chiefs |
-1900 |
-650
|
8%
|
Chicago Bears |
+1900 |
+470
|
-13%
|
| Pittsburgh
Steelers |
-170 |
+124
|
18%
|
Las Vegas Raiders |
+170 |
-148
|
-23%
|
| Philadelphia
Eagles |
-300 |
-218
|
6%
|
Tampa Bay Buccaneers |
+300 |
+180
|
-11%
|
| Cincinnati
Bengals |
-1329 |
-142
|
34%
|
Los Angeles Rams |
+1329 |
+120
|
-38%
|
Teams with a positive
expectation of profit (model odds are significantly greater than book odds) are
highlighted in gold.
Bold font indicates the best picks.
Forecasts
published on Sept. 24th, 2023 at 10:19am.
Week 3 Final
Results:
- Model Accuracy: TBD.
- Return: TBD.
Disclaimer: This project is in
early development and may generate skewed odds. It should not be used
to
place bets.
Previous Forecasts: