Predictions 2007 Week 14 and Accuracy Report

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 Summary of Accuracy
LAST WEEK
Linear regression, opponent adjusted: 11 out of 16
Logistic regression, opponent adjusted: 11 out of 16
Linear regression, unadjusted: 12 out of 16
Logistic regression, unadjusted: 11 out of 16

YEAR SO FAR
Linear regression, opponent adjusted: 67.5%
Logistic regression, opponent adjusted: 66.25%
Linear regression, unadjusted: 65%
Logistic regression, unadjusted: 63.125%

Full accuracy numbers after the jump...

Predictions based on opponent-adjusted stats
Game	Predicted Final Score Margin	P(Home Team Wins)
MIA @ BUF	3.9046	64.1947
PIT @ NE	5.1843	65.4391
CLE @ NYJ	1.0127	52.2435
IND @ BAL	-3.3048	40.6289
STL @ CIN	5.9873	68.3691
TB @ HOU	-2.2242	40.6991
CAR @ JAX	6.1614	69.8635
SD @ TEN	2.8242	59.6101
KC @ DEN	4.9643	66.3614
NYG @ PHI	4.0124	64.3319
CHI @ WAS	4.9728	67.6438
DAL @ DET	-4.2615	35.0455
OAK @ GB	10.0840	79.2978
NO @ ATL	3.3919	61.3326
MIN @ SF	-5.0540	31.7059
ARI @ SEA	4.8023	64.3473

Predictions based on unadjusted stats
Game	Predicted Final Score Margin	P(Home Team Wins)
MIA @ BUF	4.1386	64.7451
PIT @ NE	3.9195	62.1613
CLE @ NYJ	1.0829	53.3056
IND @ BAL	-2.3997	44.1654
STL @ CIN	4.6875	65.6812
TB @ HOU	-1.7954	42.7898
CAR @ JAX	5.3252	67.6244
SD @ TEN	3.4384	60.5596
KC @ DEN	5.8664	67.8183
NYG @ PHI	3.1892	61.1819
CHI @ WAS	5.3094	68.1301
DAL @ DET	-4.1566	34.4295
OAK @ GB	8.9132	75.5337
NO @ ATL	3.1149	61.4284
MIN @ SF	-5.5798	29.9452
ARI @ SEA	4.5893	64.9595

OPPONENT-ADJUSTED
Linear regression
Win prediction accuracy: 67.5000%
Mean absolute error: 11.2238 points
Correlation with result: 0.3778
% of games predicted as home team wins: 72.5000%
Logistic regression
Win prediction accuracy: 66.2500%
Correlation with result: 0.3701
% of games predicted as home team wins: 75.0000%
Win prediction accuracy when 20%≤P(home team wins)<30%: 100.0000
Win prediction accuracy when 30%≤P(home team wins)<40%: 78.5714
Win prediction accuracy when 40%≤P(home team wins)<50%: 69.5652
Win prediction accuracy when 50%≤P(home team wins)<60%: 50.0000
Win prediction accuracy when 60%≤P(home team wins)<70%: 71.7949
Win prediction accuracy when 70%≤P(home team wins)<80%: 64.5161
Win prediction accuracy when 80%≤P(home team wins)<90%: 87.5000


UNADJUSTED
Linear regression
Win prediction accuracy: 65.0000%
Mean absolute error: 11.1960 points
Correlation with result: 0.3747
% of games predicted as home team wins: 71.2500%
Logistic regression
Win prediction accuracy: 63.1250%
Correlation with result: 0.3659
% of games predicted as home team wins: 71.8750%
Win prediction accuracy when 10%≤P(home team wins)<20%: 100.0000
Win prediction accuracy when 20%≤P(home team wins)<30%: 66.6667
Win prediction accuracy when 30%≤P(home team wins)<40%: 66.6667
Win prediction accuracy when 40%≤P(home team wins)<50%: 62.5000
Win prediction accuracy when 50%≤P(home team wins)<60%: 45.7143
Win prediction accuracy when 60%≤P(home team wins)<70%: 63.1579
Win prediction accuracy when 70%≤P(home team wins)<80%: 76.6667
Win prediction accuracy when 80%≤P(home team wins)<90%: 63.6364
Win prediction accuracy when 90%≤P(home team wins)<100%: 100.0000