Data scientist, physicist, and fantasy football champion

Week 10 DEF Predictions

Model F fell a little behind Yahoo’s experts again last week, but all the models are similar. You can basically use any of them except Model A, which is just an awful model apparently.

I’m still streaming defenses this year, as I haven’t found one I can use from week to week. I used NO twice in a row, which I was a little surprised about, but it paid off.

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 10 model summary:
 

Conclusions

All of my models have the same exact top-6 (though sometimes in slightly different order): LAR, DET, CAR, PIT, SEA, and NE. Most of those will be taken already, but let’s look at these and some other top defenses that might be available:

  • PIT (at IND): Huge favorites against the offense that gives up the most points to opposing defenses. The only reason they might be available is because they were on by last week. PIT is also a legitimately good offense. Go get them.
  • NE (at DEN): Huge favorites against the offense that gives up the most points to opposing defenses. The only reason they might be available is because they were on by last week. NE is also a legitimately… OK defense. Crap, almost got to copy-paste that last one. NE hasn’t been great, but they’re getting better and you don’t have to be against DEN this year.
  • CAR (v. MIA): Jay Cutler hasn’t been sacked much this year, but he’s almost guaranteed at least one interception.
  • TEN (v. CIN): good for an interception, a few sacks, and a low score against a struggling Cincinatti offense.
  • CHI (v. GB): Brett Hundley has been pretty terrible, and Chicago has been scoring TDs which, while difficult to predict, makes them an exciting option for upside.

CAUTION:

  • TB (v NYJ): Even though they’re toward the top of my rankings, I don’t really like TB this week. I don’t think their offense will be able to posess the ball enough to control the game with Winston and Evans out, and I think NYJ will have lots of opportunites to score. I’m going to steer clear this week in a that seems like a trap against a surprisingly OK Jets offense.

Prioritize those top-six, then grab those streamers. Then win your league. Then thank me on Twitter (@FFDataStream).

Week 10 K Predictions

Week 9 DEF Predictions