We just put the 2018 season/2019 post-season behind us, but I’m ready to take a look forward to the 2019 season. Specifically, the 2019 Season Win Totals.
To level-set, I will detail my methodology and the approach to deriving my Season Win Total projections, and then we will take a look at one division per article. We look at each team’s number, as well as where I’d expect Las Vegas to set the season win number if they did so today.
The two most important things to keep in mind in this exercise are, 1) these numbers are entirely derived off of the full 2018 season data and my power ratings and 2) these are being published well before the NFL actually releases the full schedule. Right now we know each team’s opponents — home and away — we just don’t know the order or the dates of the games.
How I create my power ratings
Foundationally, my power ratings are a blend of efficiency metrics of each team’s performance in four main categories, weighted by their actual ratios that they employ and face on Offense and Defense. I factor in Offensive Pass Efficiency, Offensive Run Efficiency, and Defensive Pass and Run Efficiency. I then adjust the rating by a strength of schedule factor and then depending on the week of the season, I employ a weighted blend of each team’s final offseason power rating (which moves only slightly from where we are today). The Defensive Yard Per Play (YPP) number is subtracted from the Offensive YPP number, and then the entire league data set is normalized, and set on a power rating scale that ranges from the very highest to the very lowest I’ve measured in the last 10 years of creating my own ratings.
How power ratings are translated into game lines
The next step in this process was to take my power ratings and develop a line — a spread — on every single game for the 2019 season. A single game line is merely the product of the Home Team’s power rating plus the Home Field Advantage (HFA), minus the visiting team’s power rating. I use 2.8 as the standard HFA, per an academic study that looked at 10 years worth of results (2,560 games) to derive that number. There are a few Home Fields that will always be rated higher (3.0 or 3.5 for this exercise), but the vast majority of teams and games were assigned an HFA of 2.8.
With all 256 games now individually lined, I converted the spread to a win probability (this is a great guide on how every whole and half number converts to a win probability) and then derived the sum of each game’s probability to create a single numerical value that is the team’s projected Season Win Total.
How a team’s opponents play into the Season Win Total
I feel it’s important to remember that a season win total is unique to that team, and that team alone. Nobody else plays that team’s schedule. Meaning, no other team will have the same 16 opponents, with the same 16 power ratings across from them over the course of 17 weeks. This means that two evenly power-rated teams could come out with pretty divergent season win total numbers based upon the quality of opponents. For example, I have New Orleans and New England power rated nearly identically (5.49 and 5.50 respectively), yet the two teams are nearly a game apart in the final win total projection. Why? Well, simply, the AFC East. More specifically, the AFC East relative to the NFC South. New Orleans faces better opponents (higher power ratings), which results in lower lines relative to New England in their Divisional games, which mathematically equates to lower win probabilities in each respective game. Add each of those up, and the final number diverges more than half-a-game from each other.
With this all as a backdrop, let’s see how well you can forecast! The highest Season Win Total projection came in at 11.7 wins. The lowest at 3.5. Can you guess those two teams? Remember, 2018 data across the 2019 schedule.
Two of the three highest aggregate division win totals are in the AFC, with the third being the NFC South. Can you guess the two AFC divisions? I’ll be back soon with an in-depth look at our first division, the Patriots’ bye weeks — I mean, the AFC East.