Data scientist, physicist, and fantasy football champion

Week 15 DEF Predictions

[Note: Sorry for the late publishing. I usually try to get these out by the waiver deadline on Tuesday night but extenuating circumstances delayed my posting. The short version is that I hurt my finger and so that explains both the late posting and some of the spelling errors. Please forgive both.]

Last week neither my models nor FantasyPros top experts did very well. Of my models, Models A, B, and F did the best, although it really was pretty ugly.

We’re coming into the home stretch. If you’re in the playoffs, congrats. If you’re playing in a secondary bracket, you probably need even more luck than I can give. This week I’m continuing to use the same six models and I don’t really want to introduce any new ones before the end of the season. Model F is the one to beat, though I’ll keep an eye on A and B as well.

Model A:

Model A parameters:

  • The team that’s playing
  • Who their opponent is
  • Their predicted score
  • Their opponent’s predicted score
  • Whether the team is home or away

Model A Notes:

  • Uses only data from 2016
  • Best model from weeks 8 and 9, 3rd best in week 10
  • Terrible in week 12 but great in Week 13

Model B:

Model B parameters:

  • The team that’s playing
  • Who their opponent is
  • Their opponent’s predicted score

Model B Notes:

  • Uses only data from 2016
  • Comparable results to Model A while reducing overfitting
  • Terrible in week 12

Model C:

Model C parameters:

  • The team that’s playing
  • Who their opponent is
  • Their predicted score
  • Their opponent’s predicted score
  • Whether the team is home or away

Model C Notes:

  • Uses data from 2015 and 2016
  • OK performance weeks 8 and 9, best model in week 10

Model D:

Model D parameters:

  • The team that’s playing
  • Who their opponent is
  • Their predicted score
  • Their opponent’s predicted score
  • Whether the team is home or away

Model D Notes:

  • Uses only data from the last 6 weeks

Model E:

Model E parameters:

  • The team that’s playing
  • Who their opponent is
  • Their predicted score
  • Their opponent’s predicted score
  • Whether the team is home or away
  • NEW: “Team:week” term (does the team have momentum?)
  • NEW: “Opponent:week” term (does the opponent have momentum?)
  • NEW: “week” term (basically a bookkeeping term if I want to use the other two)

Model E Notes:

  • Uses data for last 8 weeks
  • includes week + Team:week and Opponent:week terms for each team
  • New terms try to account for momentum

Model F:

Model F parameters:

  • The team that’s playing
  • Who their opponent is
  • Their predicted score
  • Their opponent’s predicted score
  • Whether the team is home or away

Model F Notes:

  • Uses data from 2015 and 2016
  • Best model in weeks 11*, 12, and 13
  • Ignores defense TDs but weights interceptions and fumbles to try to account for it

Week 15 model summary

Rank

Model.A

Model.B

Model.C

Model.D

Model.E

Model.F

1

KC

KC

KC

BUF

BUF

SEA

2

SEA

BUF

SEA

NYG

TB

BUF

3

MIN

MIN

HOU

SEA

ATL

HOU

4

BUF

MIA

MIA

OAK

CAR

KC

5

MIA

SEA

BUF

MIA

SEA

MIA

6

OAK

ATL

MIN

KC

OAK

OAK

7

BAL

OAK

WAS

HOU

DEN

BAL

8

PHI

WAS

OAK

GB

BAL

NE

9

ATL

PHI

BAL

ATL

DET

WAS

10

WAS

BAL

GB

NE

NYG

MIN

11

HOU

HOU

DEN

PIT

SF

PIT

12

GB

GB

ATL

TB

KC

GB

13

ARI

ARI

NE

MIN

GB

NYG

14

NO

NO

PHI

CAR

JAC

NO

15

NYG

DEN

PIT

LA

HOU

ATL

16

CAR

CAR

NYG

NO

CLE

DEN

17

DEN

NYG

NO

BAL

NE

PHI

18

NE

CHI

TB

ARI

MIA

ARI

19

PIT

PIT

CAR

DET

PIT

TB

20

CHI

NE

LA

CHI

PHI

DAL

21

DET

NYJ

ARI

PHI

DAL

DET

22

TB

DET

DAL

DEN

LA

CAR

23

CLE

DAL

JAC

SD

NO

NYJ

24

LA

CLE

IND

CLE

MIN

JAC

25

JAC

TB

DET

IND

CHI

TEN

26

DAL

SD

TEN

WAS

NYJ

LA

27

NYJ

JAC

CLE

JAC

TEN

CIN

28

TEN

LA

CHI

CIN

ARI

IND

29

IND

TEN

NYJ

NYJ

WAS

CLE

30

SD

IND

SD

SF

CIN

CHI

31

CIN

CIN

CIN

DAL

SD

SD

32

SF

SF

SF

TEN

IND

SF

Not a ton of surprises this week. At this point in the season most of these teams have shown what you can expect from them. There will always be weeks like last week in which MIN doesn’t even hit the top 15, but they still got a respectable 7 points and a handful of teams really pushed their way to the top.

This week I’m seeing BUF, KC, SEA, and MIA at the top of Models A and F. Not surprising really given that these are all playing solidly and going up against teams that have given up a lot of points this year (“wait, you’re predicting a high score for BUF against CLE and MIA againsy NYJ? Gasp!”).

Weirdly, Models A and F almost have MIN and HOU switched between 3rd and 10/11th place this week. Model A is pretty insistent that MIN belongs up there. Given that they have been far more consistent than HOU I would tend to agree. But HOU is up against Jacksonville anf Blake Bortles this week, so who knows. Plus, if there’s a proper amount of garbage time I wouldn’t be surprised if Bortles gets a handful of TDs and drops HOU’s score.

The rest of the top 10 shifts out as expected as well. OAK, NE, PIT, and BAL are all good defenses going against teams that give up middling points to defenses. WAS is dead to me after Week 13 (1 point against Arizona?). In their defense, Palmer has had some amazing weeks, but in my defense I really needed to win that week to avoid the secondary playoff bracket, formerly known as the losers’ bracket.

GB is an okay DEF with an okay matchup this week (CHI), but they’ve been so hit-and-miss that it’s tough to trust them. Same with NO.

ATL might be worth a play. Model A puts them higher than Model F which tells me they tend to turn more of their turnovers into TDs, but since those are a little more unpredictable I’m going to stay away. Then again, I can afford to be a little cavalier since I have Miami.

Good luck to everyone this week! Unless your name is Nick, in which case go to hell!

Week 14 K Evaluation

Week 14 DEF Evaluation