How accurate were my objective FEIN projections ? You may think that a completely automated football forecasting system won’t be as accurate as human projections such as CBS Sportsline’s , FF Today’s , or ESPN’s . Let’s see how each set of projections stacked up in accuracy for the 2009 season.
I looked at quarterbacks with 200 pass attempts, running backs with 100 rushes, wide receivers with 40 catches, and tight ends with 30 catches this season (selecting only those with a projection from each system). There were 33 QBs, 47 RBs, 54 WRs, and 25 TEs who met these criteria.
I first adjusted each system’s projections up or down to make the average projection equal to actual average of the players in the sample. (For instance, if ESPN had projected 4.5 yards per carry and the average was 4.2, I subtracted three-tenths of a yard from each player’s projection.)
Then I compared each player’s adjusted projection to his actual number. I squared the difference between the two and weighted that number based on his actual number of attempts or catches, and took the square root of the average for each system.
Here are the results.
Read More »
The Week 16 NFL ratings are below. Ratings are calculated based on an Elo backbone using both margin of victory as well as just the result of the team’s game. Final ratings, represented on a points scale—a 21 rating is three points better than an 18 rating on a neutral field—are an average of these two individual scores.
The Elo method rewards teams for performing better than expected and punishes teams for playing worse than expected. For instance, a team that is 15-point favorites and only wins by 10 will see a decline in their rating.
These ratings do not take into account teams who rest players, such as the Colts, who lost 4.74 points on their rating after losing to the Jets by 14 points. As well, Week 17 predictions are based on full-strength teams—the Patriots are three-point favorites by these ratings yet eight-point underdogs to Houston based on Vegas spreads because bookies believe the Pats will rest their starters.
Both New England and Cincinnati, however, are fighting for the No. 3 seed and will, I believe, play the whole game with starters in.
Click on table headers to sort.
Read More »
The Week 15 NFL ratings are below. Ratings are calculated based on an Elo backbone using both margin of victory as well as just the result of the team’s game. Final ratings, represented on a points scale—a 21 rating is three points better than an 18 rating on a neutral field—are an average of these two individual scores.
The Elo method rewards teams for performing better than expected and punishes teams for playing worse than expected. For instance, a team that is 15-point favorites and only wins by 10 will see a decline in their rating.
Five teams that were predicted to win by seven or more points lost last week, and, as such, the ratings see a major shakeup this week. Twelve teams see a three-point swing in their rating from last week, and the Saints fall below No. 2 for the first time this season.
Click on table headers to sort.
Read More »
The Week 14 NFL ratings are below. Ratings are calculated based on an Elo backbone using both margin of victory as well as just the result of the team’s game. Final ratings, represented on a points scale—a 21 rating is three points better than an 18 rating on a neutral field—are an average of these two individual scores.
The Elo method rewards teams for performing better than expected and punishes teams for playing worse than expected. For instance, a team that is 15-point favorites and only wins by 10 will see a decline in their rating.
Click on table headers to sort.
Read More »
Through 13 weeks of football, both the Indianapolis Colts and New Orleans Saints are 12-0 with no huge upset special looming in the final four weeks of the season. Although both of their toughest games (Colts vs Broncos and Saints vs Cowboys) are at home, the chances of an undefeated season aren’t 100 percent or even 90 percent.
If the Colts have a 90 percent chance to win each of their remaining games, their probability of a 16-0 season is .90 x .90 x .90 x .90, or 65.6 percent. Using the Week 13 NFL ratings, we can estimate the win probability for each of their remaining games and, thus, their likelihood of an undefeated season.
Read More »
The Week 13 NFL ratings are below. Ratings are calculated based on an Elo backbone using both margin of victory as well as just the result of the team’s game. Final ratings, represented on a points scale—a 21 rating is three points better than an 18 rating on a neutral field—are an average of these two individual scores.
The Elo method rewards teams for performing better than expected and punishes teams for playing worse than expected. For instance, a team that is 15-point favorites and only wins by 10 will see a decline in their rating.
The weekly weights have changed once again; each week they will be adjusted to make the deviation of ratings the same (set at 6.0). As such, the numbers in the "jump" column represent the change in rating from last week based on the new weights.
Click on table headers to sort.
Read More »
The Week 12 NFL ratings are below. Ratings are calculated based on an Elo backbone using both margin of victory as well as just the result of the team’s game. Final ratings, represented on a points scale—a 21 rating is three points better than an 18 rating on a neutral field—are an average of these two individual scores.
The Elo method rewards teams for performing better than expected and punishes teams for playing worse than expected. For instance, a team that is 15-point favorites and only wins by 10 will see a decline in their rating.
Click on table headers to sort.
Read More »
The Week 10 NFL ratings are below. Ratings are calculated based on an Elo backbone using both margin of victory as well as just the result of the team’s game. Final ratings, represented on a points scale—a 21 rating is three points better than an 18 rating on a neutral field—are an average of these two individual scores.
The Elo method rewards teams for performing better than expected and punishes teams for playing worse than expected. For instance, a team that is 15-point favorites and only wins by 10 will see a decline in their rating.
The weekly weightings have been changed for each sets of ratings in order to keep the same amount of spread throughout the league each week. As such, the Jump column has been replaced by last week’s rank in the table below.
Click on table headers to sort.
Read More »
Oakland’s No. 1 cornerback, Nnamdi Asomugha, is regarded by most as one of the top corners in all of football. When this applies to fantasy football, most experts maintain that the No. 1 receivers lining up against him each week should have tempered results. The truth is, however, that the top wideout doesn’t always line up opposite the same defender on every play—if Arizona is facing Oakland, Larry Fitzgerald might line up across Nnamdi about 80 percent of the time.
In a quick study, I decided to look up the Nnamdi Effect for the first nine games of this season. Do No. 1 receivers perform worse against Nnamdi compared to every other week, and how much do No. 2 receivers benefit (if at all) from this effect?
Read More »
The Week 10 NFL ratings are below. Ratings are calculated based on an Elo backbone using both margin of victory as well as just the result of the team’s game. Final ratings, represented on a points scale—a 21 rating is three points better than an 18 rating on a neutral field—are an average of these two individual scores.
The Elo method rewards teams for performing better than expected (such as the Rams, who were 25-point underdogs but lost by just five) and punishes teams for playing worse than expected (such as the Saints). For instance, the Colts were one-point favorites by these ratings and won by one point. Thus, their rating this week only jumped by 2.62 points, mostly due to their four-point increase in W-L rating.
Click on table headers to sort.
Read More »
