Good news! My projections are back for the fantasy playoffs!
Since last we met, I've been sitting back, and evaluating the various statistical models. It turns out that all positions are doing about as well as expected. For instance, here are the weekly prediction errors for quarterbacks:
Week | Avg. Error | HILO Right? |
---|---|---|
1 | 5.9 | 66.7% |
2 | 6.1 | 70.8% |
3 | 6.4 | 87.5% |
4 | 4.5 | 75.0% |
5 | 6.7 | 66.7% |
6 | 6.1 | 70.8% |
7 | 5.6 | 45.8% |
8 | 6.4 | 50.0% |
9 | 8.5 | 62.5% |
10 | 6.3 | 58.3% |
11 | 5.3 | 66.7% |
12 | 5.6 | 83.3% |
13 | 6.4 | 45.8% |
Avg |
6.1 |
65.4% |
On average, the model comes within about six points of accuracy for quarterbacks in the top 24 of a given week, and correctly predicts whether or not a quarterback is going to outscore his average about 65 percent of the time. That's easily statistically signficantly different from 50 percent, which is what you would expect from random chance. In short, things are going well.
(And in case you're wondering, that epic failuire in Week 9 was due to Nick Foles' record-tying, seven-touchdown performance. The model predicted 16.2 points, and he scored 49.7. Alas.)
Here are the model's projections for the top 24 starting quarterbacks in Week 14:
QB | Tm | Avg | Matchup% | Matchup%LO | Matchup%HI | Points | PointsLO | PointsHI | Direction? |
---|---|---|---|---|---|---|---|---|---|
Peyton Manning | DEN | 29.7 | 13.8% | 8.9% | 17.2% | 26.3 | 18.4 | 34.2 | LOWER |
Matthew Stafford | DET | 25.0 | 12.1% | 9.9% | 13.7% | 23.6 | 16.9 | 30.4 | LOWER |
Drew Brees | NO | 25.4 | 8.8% | 7.5% | 9.8% | 23.0 | 16.5 | 29.4 | LOWER |
Nick Foles | PHI | 19.5 | 17.4% | 15.8% | 18.5% | 22.6 | 16.2 | 29.0 | HIGHER |
Philip Rivers | SD | 22.2 | 10.7% | 9.2% | 11.8% | 22.1 | 15.8 | 28.4 | LOWER |
Tom Brady | NE | 19.6 | 12.1% | 7.0% | 15.7% | 21.3 | 14.4 | 28.1 | HIGHER |
Tony Romo | DAL | 20.9 | 8.5% | 7.7% | 9.0% | 21.0 | 15.0 | 26.9 | HIGHER |
Cam Newton | CAR | 23.0 | 1.9% | -1.0% | 4.0% | 20.4 | 14.1 | 26.7 | LOWER |
Andy Dalton | CIN | 20.0 | 5.7% | 2.9% | 7.7% | 20.0 | 13.8 | 26.2 | LOWER |
Alex Smith | KC | 19.5 | 5.3% | 4.1% | 6.2% | 19.7 | 13.9 | 25.6 | HIGHER |
Russell Wilson | SEA | 22.1 | -4.0% | -5.8% | -2.7% | 18.9 | 13.1 | 24.8 | LOWER |
Matt Ryan | ATL | 20.3 | -0.5% | -2.2% | 0.8% | 18.9 | 13.1 | 24.7 | LOWER |
Josh McCown | CHI | 16.2 | 7.9% | 6.8% | 8.7% | 18.9 | 13.2 | 24.5 | HIGHER |
Ben Roethlisberger | PIT | 20.5 | -1.4% | -4.8% | 1.1% | 18.8 | 12.8 | 24.9 | LOWER |
Carson Palmer | ARI | 18.1 | 3.5% | 0.6% | 5.5% | 18.8 | 12.8 | 24.7 | HIGHER |
Robert Griffin III | WAS | 20.6 | -3.2% | -6.1% | -1.1% | 18.5 | 12.6 | 24.4 | LOWER |
Joe Flacco | BAL | 17.1 | 2.7% | -2.0% | 5.9% | 18.2 | 12.1 | 24.3 | HIGHER |
Colin Kaepernick | SF | 18.9 | -1.3% | -2.7% | -0.3% | 18.2 | 12.6 | 23.8 | LOWER |
Eli Manning | NYG | 16.4 | 4.0% | 0.6% | 6.4% | 18.2 | 12.3 | 24.1 | HIGHER |
Andrew Luck | IND | 20.6 | -10.1% | -16.9% | -5.3% | 17.3 | 11.1 | 23.5 | LOWER |
Mike Glennon | TB | 14.9 | 1.9% | 0.9% | 2.7% | 17.2 | 11.8 | 22.6 | HIGHER |
Ryan Fitzpatrick | TEN | 17.6 | -5.1% | -20.0% | 4.1% | 17.1 | 10.0 | 24.1 | LOWER |
Ryan Tannehill | MIA | 19.2 | -11.4% | -17.4% | -7.1% | 16.7 | 10.7 | 22.6 | LOWER |
EJ Manuel | BUF | 16.1 | -4.5% | -6.7% | -2.8% | 16.6 | 11.1 | 22.1 | HIGHER |
NICK FOLES
Everyone has officially (or unofficially) jumped on the bandwagon. And from a statistical perspective, Foles figures to benefit from playing in a close game that features a ton of scoring (over-under = 53.5 points).
carson palmer
Palmer's been on a bit of a hot streak of late (aka a "serviceable" streak), but this week Arizona is a six-point favorite over St. Louis, and the Rams pass defense merely ranks in the middle of the pack according to Football Outsiders.
alex smith
Since jump street in 2013, Smith has been the sneaky fantasy start. He scored 25 points against Denver on Sunday, and figures to follow that up with feasting on Washington's 27th-ranked pass defense.
andrew luck
So you may have noticed that the Colts' pass offense is non-existent since Reggie Wayne got hurt. As a seven-point underdog against the Bengals pass defense this week, things don't figure to get much better for Luck and company.
RUSSELL WILSON AND COLIN KAEPERNICK
This game is just about a toss up in terms of the spread, and the 41.5-point over-under suggests the betting public doesn't see a high-scoring game either. If possible, you're better off starting someone else.