Data scientist, physicist, and fantasy football champion

Week 5 DEF Predictions

Model C is the best model that I have, but Yahoo’s experts (i.e., these guys here ) consistently beat me by just a little. Man, I really hate that. I’m going to have to refine my models when I get the time.

The new model did nothing good last week, but I’m not going to drop it yet. Let’s see some predictions for this week:

ModelTypeScoreOppScoreOppHomeAwayDataYearsPointSystem
ALMXXXX2017IOUH/Standard
CLMXXXX2015-2017IOUH/Standard
FLMXXXX2015-2017TruScor
GGibbs SamplerXXXX2015-2017TruScor
HLMXXXX2017TruScor
IMany LMsXXXX2015-2017IOUH/Standard

 

Model A:

.Model A-1.png


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 2017
Standard points
Did well in 2016, but didn’t stand out

Model C:

.Model C-1.png


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 2015-2017
Standard scoring
Did well in 2016, but didn’t stand out

Model F:

.Model F-1.png


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 2015-2017
Truscor ignores defense TDs but weights interceptions and fumbles to try to account for it
I really liked this one in 2016. It was marginally better than A and C

Model G:

.Model G-1.png


Model G 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 G Notes:

Uses data from 2015-2017
Truscor ignores defense TDs but weights interceptions and fumbles to try to account for it
Uses Bayesian statistics (Gibbs sampler) to create a model
Would have done very well in 2016 based on simulation

Model H:

.Model H-1.png


Model H 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 H Notes:

Uses only data from 2017
Uses TruScor points to reduce the effect of TDs

Model I:

.Model I-1.png


Model I parameters may include:

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 I Notes:

Data from 2015-2017
This model tries to predict all of the components of scoring separately (interceptions, sacks, etc.). Each component comes from a linear model containing only the significant terms (p < 0.05) from the list of possible terms above.

Week 5 model summary:

RankACFGHI
1DETPHIPHIPHIDETPHI
2SFKCPITPITINDPIT
3INDPITCLECLEBALMIN
4KCLARDETNYJKCDET
5DALDETNYJDETSFKC
6BALCLENYGNYGCLECLE
7HOUNYGLAROAKDALNYG
8CARMINKCKCTENLAR
9JACSFOAKLARHOUTEN
10TENTENTENTENMINCIN
11CLENYJSFMINCARSF
12PHIMIAMINSFNYJMIA
13BUFOAKCARCINPHIOAK
14NEHOUCININDNECAR
15PITSEAINDCARBUFNYJ
16CINARIBUFBUFPITBUF
17MINCARARIBALJACHOU
18NYJBUFHOUNEOAKIND
19ARICINBALHOUARILAC
20GBINDNEARIGBBAL
21LARBALLACLACLARARI
22TBLACMIAMIACINSEA
23OAKNEGBSEANYGNE
24SEAGBSEAGBMIADAL
25MIATBJACCHISEAGB
26NYGDALCHIJACTBCHI
27LACCHIDALDALLACJAC
28CHIJACTBTBCHITB


Conclusions


[Another week, another new model. It gives results similar to C despite a pretty different methodology. Encouraging, but likely not earth-shattering improvement]

With the exception of Models A and H, Philadelphia is the clear consensus choice for the week. PIT, CLE, DET, NYG, LAR, MIN, and KC are all top choices according to Model C. These are all teams with good matchups, though CLE and NYG haven’t done much this year to distinguish thelselves as top-scoring DSTs. They’ve also had somewhat tougher matchups so far, though, which is why my models still put them up there.

Yahoo’s experts have SEA, ARI, and NYJ all toward the top of their lists. Odds are you won’t be able to stream ARI or SEA. NYJ is worth a look against CLE; they did really well week 3 against Miami, but have struggled against better offenses. That said, the Browns can hardly be considered a better offense.

I’m giving Detroit a long look this week. They’ve done very well the first four weeks against both good and bad offenses, and we’ll see if they continue their streak against a mid-level Carolina offense.

Week 5 QB Predictions

Week 5 K Predictions