NFL Week 1 model projections
This is part of a recurring series on VSiN.com in which VSiN host Jonathan Von Tobel tracks the progress of a model he created for the NFL season
***Top NFL Resources***
*It's the VSiN Black Friday Special. Take advantage of the largest savings of the year by upgrading to VSiN Pro. For a limited time, you can secure Pro access until May 1st for only $60.*
- NFL Expert Picks
- NFL Betting Hub
- NFL 2024 Week-By-Week Schedule
- Parlay Calculator
- NFL Betting Splits
- NFL Betting Odds
“We ain’t got tomorrow. We got now. We ain’t got next. We got now. We ain’t coming no more. We here.”
The now immortal words of Colorado head coach Deion Sanders ring true for Jon Von Model, as the NFL season is finally here. After a summer of YouTube videos, books, tweaks and adjustments it’s time to test what Jon Von Model can do.
But, before we get to what our final Week 1 projections are, let’s talk about what transpired over the last week, because we do have a small problem.
The day I published the first entry in this journal ended up being the same day that Aaron Schatz, NFL analyst and creator of the DVOA metric, officially announced that he would be joining FTN Network as their Chief Analytics Officer. But, Jon Von Model does not use DVOA, so why does it matter that the source for the metric changed?
Well the model does use two metrics from Football Outsiders: net drive success rate (DSR) and adjusted line yards allowed (ALYA). And Football Outsiders happened to shut its website down on the same day Schatz announced his new gig at FTN Network.
That leaves me in a small pickle.
RBSDM does have a DSR of their own, but for whatever reason regression testing with that metric instead of Football Outsiders’ version of the statistic brings the correlation down. There is an old article from 2018 that Chase Stuart wrote at Football Perspective that goes over calculating Football Outsiders’ version of DSR, so we might be doing some good old fashioned math each week to get the number, but that still leaves a hole when it comes to ALYA, which I cannot find anymore.
For the next week I will be running as many different metrics as I can to see if there is another metric which I can use to replace, but for this week we at least have the metric to use.
Which brings us to the adjustment made since we last … spoke?
I noted last week that I did not like the projections Jon Von Model had for the game between the Bills and Jets on Monday night, because it was clearly viewing New York as the Jets from last season. The quarterback-specific metrics I used for the model were tied to Aaron Rodgers from last season, but some of the team-specific offensive metrics were the Jets from last season.
In order to get a more accurate number on that game I gave New York the offensive statistics from Green Bay last season. I do not think it’s the perfect solution, but the returns were a projection that was slightly more in line with where the market is for that game, as opposed to a clearly inaccurate projection.
Last week, prior to the adjustments, Jon Von Model projected a 7.06-point win for the Bills. You can see in the latest projections that it is now down to a 5.53-point win for Buffalo. The market has pushed the Bills up to a consensus 2.5-point favorite since the last entry here, so what was once nearly a touchdown edge in a game has been cut down to a field goal.
That adjustment on New York – along with the evolving market for Week 1 – means we see a change in the five biggest edges for the first week of action. Last week the five best bets from Jon Von Model were the Packers, Chiefs, Eagles, Jaguars and Bills, but since the edge for Buffalo is now only 3.03 points it is now the Dolphins (+3.12) who are the fifth-largest edge according to the model.
Mitch Moss and I did sign up for the SuperContest last week, so when contest lines are posted I will run the numbers to see if there are any significant changes but for now it seems our contest card is set for Week 1.
There are just two days until kick-off of the NFL season, and I am nervous and excited at the same time to see what this model can do. This experiment can go many ways, and while I do believe the journey will be rough I can’t help but have some hope that Jon Von Model will do well.
There will be some negative feedback along the way – as there always is – but as Deion Sanders says:
“This ain’t got nothing to do with the naysayers. The unbelievers. The haters. The doubters. This is about us.”
