The 2020 NFL season is right in front of us, and with the hunt for wins, so begins each player's quest to be crowned as the NFL's most valuable player.
While it's technically true that every single player in the NFL is eligible for MVP honors, the race for the title is really only open to the league's 32 signal-callers. And this is why Kevin Cole and I developed a methodology to estimate MVP odds for each team's starting quarterback.
Throughout the upcoming season, we will be updating these odds every week in a Race to MVP column. But as no games were played before Week 1, we will use the very first article of this series to explain the methodology behind updating the odds based on new, weekly data.
We will also take another look at the odds going into the season, which come with some minor changes. This is due to roster changes (Cam Newton's addition to the New England Patriots having the largest effect) that affect our win-total simulations along with some minor tweaks to the model that simulates each quarterback's efficiency during the regular season.
As a refresher, here is the basic methodology:
We trained a model to predict the winner of the MVP award based on only three regular-season parameters after the conclusion of Week 17:
1. The number of wins and how it ranks across the league
2. Rank in the respective division
3. Expected points added per play and how it ranks across the league
Since the model works with the regular-season stats in hindsight, we need to simulate the remaining games of the season to have results required to feed the model. We invoke our weekly win-total simulations as well as a Bayesian Updating method to simulate each quarterback's EPA per play in 2020.
This yields the following odds going into Week 1. Naturally, the column that describes the change from last week doesn't carry any information at this time, but it will be interesting to watch this and compare it to how the MVP betting odds change once the season actually kicks off.
DOMINATE FANTASY FOOTBALL & BETTING WITH AI-POWERED DATA & TOOLS TRUSTED BY ALL 32
Already have a subscription? Log In