Data scientist, physicist, and fantasy football champion

Week 14 DEF Evaluation

Since we’re coming to the end of th eseason I don’t think I’m going to introduce any new models. Last week I used the same six models I’ve been using and for the last few weeks Models A, B, and F have done the best. What about Week 14:

Rank

Model.A

Model.B

Model.C

Model.D

Model.E

Model.F

FP.top.20

DWC

True.Order

True.Score

1

MIN

MIN

MIN

MIA

CIN

MIN

MIN

MIN

ATL

28

2

CAR

CAR

DEN

CIN

KC

CIN

CIN

CIN

GB

19

3

DEN

DEN

MIA

MIN

MIN

DEN

DEN

DET

CAR

18

4

MIA

SD

CAR

CAR

CLE

TEN

ARI

MIA

CHI

14

5

SD

SF

CIN

TB

OAK

MIA

DET

NE

NYG

13

6

SF

MIA

SD

DET

ARI

CAR

ATL

ATL

DAL

12

7

CIN

ATL

TEN

SD

NE

SD

NE

ARI

KC

11

8

ATL

CIN

DET

SEA

GB

WAS

SEA

SEA

TB

11

9

TB

TB

IND

DEN

NYG

IND

TEN

KC

TEN

11

10

ARI

ARI

ATL

IND

DET

SF

MIA

CAR

PHI

10

11

DET

PHI

SF

PIT

JAC

DET

KC

SF

MIA

10

12

WAS

DET

WAS

CLE

NYJ

ARI

SD

DEN

CIN

9

13

TEN

WAS

SEA

GB

SEA

PIT

PIT

IND

SF

8

14

SEA

TEN

ARI

TEN

MIA

SEA

WAS

WAS

WAS

8

15

PHI

KC

KC

NE

TEN

TB

NYJ

TEN

HOU

8

Model.A

Model.B

Model.C

Model.D

Model.E

Model.F

FP.top.20

DWC

79

77

88

85

84

77

84

92

Ugh, an ugly week for prediction. None of the models did very well this week, but Models A, B, and F did marginally better as usual. The problem this week was that of the actual top 5 (ATL, GB, CAR, CHI, NYG) only one (CAR) was ranked very highly. ATL was between 6 and 10 on a few of these lists, but none ranked GB, CHI, NYG or even DAL in the top 15. This wasn’t a great week overall for defense predictions.

Model F does consistently well. That’s still the model to beat. Go TruScor!

Week 15 DEF Predictions

Week 14 K Prediction