Sunday, September 16, 2007

Win Probabilities and Expected Wins - 2007 Week 2

The following table lists the probabilities that the home team would win each game given the rushing averages, passing averages (includes sack yards lost and sacks as pass attempts), sack rates, third down conversion rates, penalty yards, and interception and fumble rates (by number of completions and rushing plays) for both teams. A logistic regression model trained on 1996-2006 data computes the probabilities. The final score margin is Home team points - Away team points. The sum of these probabilities of victory form the expected win totals in the second table to be added after the Monday night game. (Edit: Second table added.)



















Home TeamAway TeamP(Home Team Won) (%)Final Score Margin
CLECIN86.30766
NYGGB12.3475-22
CHIKC22.084410
MIADAL12.0833-17
DENOAK99.96703
DETMIN83.96563
NESD98.120224
PHIWAS48.4337-8
CARHOU10.4949-13
TENIND55.4136-2
JAXATL41.66666
PITBUF98.093223
STLSF81.8986-1
TBNO99.976317
ARISEA52.21313
BALNYJ48.15397


The following games "should have been" won by the team that didn't actually win:
KC @ CHI - I don't take punt returns into consideration anymore.
IND @ TEN - 55/45 chance and the Titans lost by 2 points. I find the Colts' 6.8 yards per pass more encouraging for future performance, however.
ATL @ JAX - A 60/40 split in favor of Atlanta. Again, Jacksonville's pass efficiency (8.6 yards/pass) is more encouraging for future performance. I am disturbed at how close the first two games have been for Jacksonville, however.
SF @ STL - It came down to a 50+-yard field goal that was short by about 6 inches, but it shouldn't have to begin with. The Rams' were much more efficient on offense. Alex Smith averaged 4.6 yards per attempt against what is considered a poor defense. Beware of regression with the 49ers.
NYJ @ BAL - 52/48 split. Kellen Clemens was the more efficient passer.


















































TeamActual WinsExpected WinsProjected Win Total
AFC East
BUF0.00.04750.3803
MIA0.00.67475.3979
NE2.01.978015.8242
NYJ0.00.52164.1731
AFC North
BAL1.00.49923.9934
CIN1.01.11938.9543
CLE1.00.87006.9598
PIT2.01.974015.7922
AFC South
HOU2.01.892615.1412
IND2.01.441311.5303
JAX1.01.334510.6759
TEN1.00.63635.0905
AFC West
DEN2.01.971215.7696
KC0.00.78166.2525
OAK0.00.05490.4396
SD1.00.98887.9101
NFC East
DAL2.01.846114.7685
NYG0.00.15661.2526
PHI0.00.96597.7270
WAS2.00.96187.6940
NFC North
CHI1.00.25092.0071
DET2.01.785014.2803
GB2.01.395011.1599
MIN1.01.02398.1912
NFC South
ATL0.00.71985.7582
CAR1.00.97567.8046
NO0.00.00480.0385
TB1.01.08308.6641
NFC West
ARI1.00.76796.1430
STL0.00.94847.5868
SF2.00.93537.4822
SEA1.01.394611.1569

2 comments:

Phil Birnbaum said...

Interesting! These are retrospective after-the-game probabilities, right?

Have you posted the formula somewhere? It would be interesting to compare to others, like "The Wages of Wins" Win Score.

Derek said...

Correct. I have not posted the exact formula, but I could give you the regression coefficients. E-mail me if you want them.