The Power Rank uses data and analytics to make accurate predictions for football and March Madness. I developed these methods based on my PhD in applied math from Stanford.
Get a sample of my best football predictions
If you sign up for my free email newsletter, you get a sample of my best football predictions usually saved for paying members of the site. Each week, I also provide analysis on each of the games.
Here is some unsolicited feedback on the 2020 version of the newsletter.
A remarkable newsletter.— Michael H.
This newsletter is a different take on football action. Curious and thoughtful.— Gino G.
To get these services, sign up for my free email newsletter. Enter your best email address and click on “Sign up now!”
NFL, Conference Championship
These predictions are based on my team rankings that take margin of victory in games and adjusts for strength of schedule. There is also a preseason component that gets less weight with each week.
Members of The Power Rank have access to more accurate ensemble predictions. To learn more, click here.
1. Buffalo at Kansas City.
Kansas City (2) will beat Buffalo (3) by 1.6 at home. Buffalo has a 45% chance of beating Kansas City.
2. Tampa Bay at Green Bay.
Green Bay (4) will beat Tampa Bay (5) by 1.5 at home. Tampa Bay has a 45% chance of beating Green Bay.
College Football Championship Game
In 2020, the college football team rankings are based on games from the 2019 and 2020 season, with the current season getting twice the weight. A home field of 1.2 points is used.
This prediction below based on points is off and favors the wrong team. Members of The Power Rank have access to more accurate predictions based on more data sources. To learn more, click here.
1. Ohio State versus Alabama at a neutral site.
Ohio State (1) will beat Alabama (2) by 1.5 at a neutral site. Alabama has a 46% chance of beating Ohio State.
Major League Baseball
There will be no more baseball predictions for 2020. There’s a bug with my updating system, and I don’t have the time during football season to fix it. My apologies.
The matchup shows the projected FIP for each starting pitcher according to ZiPS.
The integer in parentheses gives the team’s rank in The Power Rank.
Since Aug 18, 2020, the team with the higher win probability has won 327 of 591 games for a win percentage of 55.3%.
The team favored by the markets has won 340.0 of 591 games for a win percentage of 57.5%.
European Club Soccer
These predictions are based on expected goals (xG) from past matches. This raw data is obtained from FBRef. I adjust for strength of schedule based on a least squares algorithm, which is equivalent to the Simple Rating System.
After these schedule adjustments, I have offensive and defensive ratings for each team. These numbers imply goal rates for each team in a match.
I assume a Poisson model and calculate the probability for a win, loss and draw. I’m assuming a home advantage of 0.12 goals based on matches with no fans.
To learn more about how the efficiency prediction works, check out my ultimate guide to predictive college basketball analytics.