Tuesday, October 9, 2007

Predictions 2007 Week 6

The predictions are based on yards per play on rushing and passing plays (sacks and sack yards included), sack rates, third down conversion rates, penalty yards per game, and fumble and interception rates. The models are trained on the 1996-2006 seasons. Linear regression predicts the final score margin (home team points - away team points). Logistic regression estimates the probability that the home team will win.

Quick recap of last week's predictions: The logistic regression based on adjusted stats was 12-2 last week. For weeks 4-5, the logistic regression predictions have gone 20-8 (71.45%).

Based on opponent-adjusted stats

GamePredicted Final Score MarginP(Home Team Wins)
PHI @ NYJ-4.617033.7208
STL @ BAL4.090863.7903
MIA @ CLE2.271855.0079
HOU @ JAX2.850859.0307
CIN @ KC3.079261.0696
OAK @ SD8.100870.9335
NE @ DAL1.040054.5029
MIN @ CHI-2.201540.3186
WAS @ GB-0.403148.9549
NYG @ ATL-1.179345.3254
TEN @ TB4.632667.4386
CAR @ ARI5.605968.3911
NO @ SEA10.134982.4083

Based on unadjusted stats

GamePredicted Final Score MarginP(Home Team Wins)
PHI @ NYJ-3.873935.8694
STL @ BAL6.520772.8809
MIA @ CLE3.583658.0139
HOU @ JAX3.454961.3604
CIN @ KC4.051465.2268
OAK @ SD7.426468.7658
NE @ DAL2.858859.1881
MIN @ CHI-4.116632.9111
WAS @ GB-1.289146.2764
NYG @ ATL0.156049.5781
TEN @ TB4.995170.1540
CAR @ ARI3.441263.8702
NO @ SEA11.377285.1483

With interconference home teams being so strong, I'd take Dallas over New England, at least to beat the spread. Same goes for TEN@TB and STL@BAL. You might also want to stay away from PHI@NYJ based on that and the Eagles' erratic performance. Maybe St. Louis will win and the Jets will lose. Or you could just trust the stats.

Watch out for away team upsets in HOU@JAX and OAK@SD given how 55% of intradivision games have been won by the away team so far (check out the Homefield Advantage article I posted today). Based on this, I'm also pretty confident in the Minnesota pick. Plus, the Bears suck.

No comments: