Tuesday, November 12, 2013

2013 - week 11 - GRIDs - Efficiency and QB rating differential

This week I decided to add both GRID as I've noticed that they both have revelance these past few weeks.

2013 - week 11 - GRID - Efficiency

2013 - week 11 - GRID - QB rating differential

Here's what I've noticed in my tracking of predicting wins SU. The QB Rating Differential seems to be better at a rate of about 2-4 games per week. For the majority of picks they will agree, but where they differ is where QB Rating Differential shines as it continues to out pick the Efficiency GRID.

The Efficiency GRID has captured the demise of ATL better in my opinion. Though they are both accumulative of the season, the Efficiency GRID has shown a greater decline and thus ranks ATL below TB, instead of above TB as the QB Rating Differential has done.

You'll also note the great rise that PHI and is now considerably ahead of DAL in the QBR GRID vs the equal billing that the Efficiency GRID has them at the moment.

IND has also taken a great fall w/o their #1 WR Wayne in the QB rating differential.

In general those teams that have a good QB are better to follow w/QB Rating Differential and those teams that depend on their D, their ST and a run game are much more predictable with the Efficiency GRID. The problem lies when you have to choose a team in a game and they are more predictable w/a different GRID. It then becomes a toss of the coin, or if their is a distraction (drama), I'll use that as a deciding factor. I've even gone back to the basics of too many injuries to overcome for a team to use as a deciding factor.

Below is a simple text output of predicting outcomes for week 11 games:

QBRating Differential
road roadO-D home homeO-D predict DIF
IND 7.975 TEN 13.96 TEN 5.985 ?
NYJ -32.525 BUF 2.68 BUF 35.205
ATL -23.425 TB -22.6 TB 0.825
DET 12.88 PIT 7.05 DET -5.83 ?
WAS -20.98 PHI -26.6 WAS -5.62
SD 2.28 MIA 15.225 MIA 12.945 ?
BAL -25.36 CHI 13.18 CHI 38.54
CLE -17.8 CIN 38.25 CIN 56.05
OAK -27.8 HOU -4.275 HOU 23.525
ARI -16.575 JAC -49.575 ARI -33
KC 30.2 DEN 43.1 DEN 12.9
MIN -25.875 SEA 35.725 SEA 61.6
SF 12.125 NO 59.7 NO 47.575
GB -12.85 NYG -6.175 NYG 6.675 ?
NE -12.075 CAR 23.825 CAR 35.9

road RO-Deff home HO-Deff predict DIF
IND 17% TEN 2% IND -14.77% ?
NYJ -7% BUF -5% BUF 2.13%
ATL -15% TB -12% TB 3.42%
DET -10% PIT -1% PIT 8.67% ?
WAS -9% PHI -18% WAS -9.36%
SD -1% MIA -8% SD -7.40% ?
BAL -4% CHI -0% CHI 4.25%
CLE -8% CIN 10% CIN 17.62%
OAK -13% HOU -6% HOU 6.32%
ARI -18% JAC -30% ARI -11.93%
KC 14% DEN 22% DEN 8.22%
MIN -13% SEA 21% SEA 34.21%
SF 8% NO 27% NO 18.97%
GB 9% NYG -1% GB -10.12% ?
NE 4% CAR 21% CAR 17.11%


As you note above their are four games which the GRIDs differ in predicting the outcome, we'll see which GRID is right and wrong.

One other thing, the text output is partially based on the GRID, the GRIDs themselves show the seasons accumulative effects. At this point of the season, I modified the text output to only measure a road team and its road record vs a home team and its home record, thus the numbers in the text are not the same as you see in the GRID. I do this because their are teams that have a greater or lesser tendency to produce on the road vs at home. Most of the time the GRID will predict the same outcome as the simple text output, but their are times when they will differ.

Teams that are 'Better on the road' (as measured w/QB rating differential):

No comments: