Super Bowl LIV Forestcast

The below chart details the probability each team will win Super Bowl LIV, and the edge one has over a major offshore book.

This model analyzes a myriad of advanced statistics over the past decade of NFL playoffs, determines which statistics are most likely to forecast winners, and leverages those statistics in a Random Forest Model.

kelly% indicates the percentage of one’s bankroll one would wager if pursing a maximally aggressive betting strategy and only selecting a single wager from the below table. Note that in the event a bettor makes simultaneous wagers, the kelly fraction for each wager would need to be reduced.

teambovada oddsmodel oddsedgekelly%
SEA3.4%20.9%506.1%18.1%
HOU2.4%12.1%396.1%9.9%
NE9.1%22.6%148.6%14.9%
SF20.0%40.0%100.0%25.0%
BAL31.3%33.9%8.5%3.9%
GB9.1%7.5%-17.5%
TEN2.2%1.0%-54.0%
BUF2.0%0.7%-64.3%
MIN2.4%0.7%-71.3%
NO14.3%1.6%-88.8%
PHI3.4%0.2%-94.2%
KC21.1%0.8%-96.2%

There are 10 total statistical features used by this model to make its evaluation. They are:

  • The current odds (start of the post season) of winning the Super Bowl
  • Certain DVOA statistics, courtesy of FootballOutsiders. The specific DVOA statistics used are Defense DVOA (including the recency-weighted version and Special Teams DVOA (including the recency-weighted version).
  • Each team’s net points.
  • Each team’s margin of victory.
  • Each team’s strength of schedule throughout the season.
  • Each team’s SRS and Defense SRS.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Powered by WordPress.com.

Up ↑

%d bloggers like this: