It’s the Most Wonderful Time of the Year
Is that Andy Williams singing “It’s the most wonderful time of the year?”
It’s here! The 2021 NFL season is upon us, finally! In an offseason that featured Aaron Rodgers cementing his place as the league’s most emo QB, Dak Prescott getting paid, oh, and Tom Brady is still playing football, I’m just glad we have a solid 5 months of entertainment ahead!
For new readers, this blog is a chance for me to share my sports betting process with you. Hopefully you can take what resonates with you and apply it to your decision-making arenas. For this post, I plan to outline how I seed my Week 1 NFL power ratings.
Many serious sports bettors will say that the offseason is where the money is made. Tweaking the algorithm (or “algo” as I call it) during the season or changing your decision-making process under duress are not recipes for success. Correctly adjusting team ratings between weeks, especially in the early season where the sample size is small, can be tricky too. Therefore, the more accurate your pre-season assessment of every team, the higher your likelihood of winning.
Here’s my process for determining the Week 1 power ratings.
Pre-Market Ratings
First, I start with an evaluation of each team’s performance from last season. There are many ways to do this but I like to first take a simple average of each team’s power rating over the course of the prior season. Other tactics might be to use FiveThirtyEight’s Elo Ratings as a starting point. Second, I calculate the Pythagorean expected win total and regress each team’s power rating toward their expected wins. Third, I assess the key offseason personnel moves and adjust accordingly. Fourth, I carefully evaluate the pre-season performances and tune the ratings. These first four steps are what I call the “pre-market” ratings.
Market Adjustment
We should always be asking ourselves, “What am I missing?”. While the market isn’t a perfect indicator of a team’s strength, it does provide a lot of information that we can use to sharpen our Week 1 ratings. If the market line for a game is way different than my number, I have learned to pause and consider where I might be blind.
In this step, I take my pre-marking ratings and compare them to the market-derived rating. Admittedly, I have yet to create a formula to translate market prices into implied team ratings. But if I were to create one, I’d start with the season win totals futures prices and work from there. Fortunately, I’ve found a few sources online that have calculated the market-based ratings. In applying the market adjustment, I like to slightly bias the adjustment towards my initial pre-market ratings.
Model Composite Adjustment
Finally, I source power ratings from a motley assortment of reputable analytics services like ESPN and Football Outsiders. I combine all of the models to form a composite rating. Using this composite rating, I tweak my market-adjusted rating depending on the magnitude of disagreement between the models.
I take one more human look at the results to understand where I may have been off in my initial ratings and that’s it!
Now that we feel confident in our Week 1 power ratings, there are a few best practices I’ve learned over the years that I’d like to share:
If you need to make a midseason change, move slowly. Be vigilant against overreacting.
Break the season into chunks. I like to reevaluate each team after every 4-5 weeks.
Respect the market, don’t follow the market. Remember, the market peddles perception not reality.
Good luck!