Making better football bets with PhD level analytics
You want to win at sports betting. It’s 2023, and legalization in the United States has opened up massive opportunities. In many states, betting is as simple as opening a smart phone app.
However, it’s hard to win at football. NFL spreads are widely considered the hardest market to beat because of massive liquidity. College football is not that much easier.
You’re fighting an uphill battle against sports books. Since they require $110 to win $100, you can’t just win more than half of your bets. Instead, you need a win rate of at least 52.4% in the sharpest of markets.
In this uphill battle, you can’t just wait for lukewarm water to separate into hot and cold. Predictive analytics is a crucial component of winning football betting, and The Power Rank provides top flight numbers.
As someone who studies football for a living, it’s extremely valuable to have other people in the industry who you can trust and bounce ideas off of. Ed’s analysis is one of the first I go to on a week-to-week basis, and should be one of yours as well.
— Dr. Eric Eager, Head of Research and Development at PFF
I first heard about the The Power Rank on one of Gill Alexander’s podcasts. Gill said Ed was one of the few people he trusted that made money in the analytics business besides Rufus Peabody.
Ed releases great member numbers for college football and the NFL. For under $100 annually this is the best bargain in the industry in my opinion.
— Member of The Power Rank from Georgia.
To become a member of The Power Rank, click here.
From academia to football analytics
Hi, my name is Ed Feng. I’m the founder of The Power Rank, a company that believes predictive analytics is a crucial component of winning sports betting.
While I’ve been a crazy fan my entire life, I never expected to work in the sports world. After getting my PhD from Stanford in chemical engineering, I was one of many in search of an academic job.
That changed in 2008. I was reading the paper on Google’s PageRank algorithm and realized all of its similarities with my research. This inspired me to develop a new method for ranking teams.
I worked out some NFL rankings and sent an email to my friends. Enough of them were intrigued for me to explore other sports. It became an obsession, and I found a way to make The Power Rank my job.
In March of 2012, SB Nation made a video of my March Madness analytics. In January of 2013, I accurately predicted Alabama would beat Notre Dame on the cover of Sports Illustrated. They called me their analytics expert.
Since then, my numbers and content have appeared in a wide range of media outlets.
In addition, the ideas in my Amazon best selling book How to Win Your NCAA Tournament Pool have helped many people win their March Madness pool.
To become a member of The Power Rank and get all my best analytics, click here.
How to make football predictions
I want to tell you about the membership product that gives you access to my best football predictions. But first, let’s take a look under the hood at the engine that drives these analytics.
The foundation of my football analytics is accurate adjustments for strength of schedule. This started when I developed a ranking algorithm based on my PhD research at Stanford.
Let’s look at how my algorithm ranks NFL teams on offense and defense based on success rate, a powerful efficiency metric. In my methods, a play is a success if it obtains 50% of the necessary yards on 1st down, 70% on 2nd down, and 100% on 3rd and 4th down.
- The algorithm gets the play by play data from every NFL game. While this includes up to 272 games for an NFL regular season, it can be thousands in college football.
- It creates a network in which the offense and defense of each team are nodes and games are edges that connect the nodes. This was inspired by Google’s PageRank method, which brought order and insight to the complex world of web search.
- From this network, the method develops a set of equations to be solved. For the NFL, there are 64 equations (two for each team) with 64 unknown variables.
- The algorithm solves for the 64 variables simultaneously. This is the key to accurately accounting for strength of schedule. Many other ranking systems do not do this.
- Finally, the variable for each unit (offense or defense) is transformed into a rating that gives an expected success rate against an average team.
Every morning after an NFL or college game, my MacBook Air crunches the numbers. The resulting rankings allow me to make predictions for point spreads and totals.
Analytics experts such as Bill Connelly of ESPN trust my work.
Accuracy of the predictions
Members of The Power Rank have access to my best computer predictions for the NFL, college football and college basketball.
NFL spreads
For the 2020 through 2022 seasons, here is how the NFL spread predictions have performed against the closing market.
- 52.4% (310-282 with 10 pushes) when the prediction differs from the market by one or more points.
- 58.7% (138-97 with 3 pushes) when the prediction differs from the market by three or more points.
There is probably some good fortune in the small sample size with a threshold of three points. From 2018 through 2022, the predictions have been 56.9% (214-162 with 3 pushes) when the prediction differs from the closing market by three or more points. However, the results suggest the model is valuable tool against the closing market.
For this set of games, the root mean squared (rms) error of the predictions from the game results has been 12.74 points. The rms error for the closing market has been 12.58 points. The Power Rank member predictions have been within 1.3% of the market.
The NFL predictions have picked the game winner straight up in 65.6% of games (547-287). This accuracy is better than the closing markets at 64.6% (539-295).
College football spreads
In the 2020 COVID season, I developed a new college football model that I’ve found to be more accurate than my prior model. From the 2020 through 2022 seasons, here is how the college football spread predictions have performed against the opening market.
- 51.9% (791-767 with 31 pushes) when the prediction differs from the market by one or more points.
- 52.1% (369-339 with 12 pushes) when the prediction differs from the market by three or more points.
- 53.4% (134-117 with 5 pushes) when the prediction differs from the marketing by five or more points.
For this set of games, the root mean squared (rms) error of the predictions from the game results has been 15.80 points. The rms error for the opening market has been 15.69 points. The Power Rank predictions have been within 0.7% of the opening markets.
The predictions are not has good against the closing market: 50.3% against the spread (776-767 with 31 pushes when the prediction differs from the market by one or more points). The rms error has been within 1.6% of the closing markets.
The college football predictions are 71.5% (1476-592) in picking the winners of games straight up over the past three seasons. The closing markets have a win rate of 72.5% (1500-568).
College football totals
In 2021, I tweaked the college football totals model based on an accuracy analysis of the components. These predictions usually start around week 5 in the season.
From 2021 to 2022, here is the performance of the totals model compared to the opening market.
- 55.5% (471-378 with 4 pushes) when the prediction differs from the market by one or more points.
- 59.0% (270-188 with 3 pushes) when the prediction differs from the market by three or more points.
For this set of games, the root mean squared (rms) error of the predictions from the game results has been 16.10 points. The rms error for the opening market has been 16.20 points. The Power Rank predictions have been 0.7% better than the opening markets.
Against the closing market, the predictions have been 50.3% (776-767 with 31 pushes) with a threshold of a point, and the accuracy gets worse with a larger threshold. The rms error of the predictions have been within 1.7% of the closing market.
For college football, it is important to bet spreads and totals early in the week. In 2022, I started posting these predictions on Sunday to take advantage of these markets.
When I started betting football seriously, I quickly found that data and analytics is essential to being a consistent winner.
What I like so much about The Power Rank is that it not only provides excellent rankings and predictions, but also adjusted numbers for “real” football stats like Yards Per Play and Success Rate.
Having “real” football stats adjusted for strength of schedule helps me recognize and exploit inefficiencies in the betting markets.
Jordan Jones, member of The Power Rank
College basketball spreads
Let’s look at results for The Power Rank’s college basketball predictions for the two seasons starting in November, 2021 and ending in April, 2023. Since I post these predictions the day before the games are played, let’s first look at performance against the opening market.
- 53.2% against the spread (2729-2405 with 99 pushes in games in which the prediction differs from the market by one point or more).
- A rms error of 11.14 points, which is the same as the opening market.
By the rms error, the predictions are not quite as good as the opening market in November. However, the predictions have their best performance in December and January and are still strong the remainder of the season.
The college basketball predictions are not quite as accurate against the closing market.
- 50.7% against the spread (2863-2787 with 108 pushes in games in which the prediction differs from the market by one point or more).
- A rms error within 0.73% of the closing market.
Be smart
I do not recommend making bets based only on The Power Rank’s predictions. While these numbers serve as an objective baseline, winning football betting also involves understanding other factors.
- Market movement. Looking at how a point spread moves before the market closes gives you information on the opinion of sharp sports bettors. This critical information should inform future bets.
- Context of teams. There is always some error in the predictive analytics, and understanding the ceiling and floor of a team can make for better predictions. For example, a team with a Hall of Fame QB like Tom Brady has a high ceiling for the pass offense, so less than stellar numbers might underestimate future performance.
- Injuries and situations. An injury to the quarterback almost always matters, and a team coming off a bye week has an edge over a team that played the past week.
In my weekly analysis for members during football season, I highlight these factors.
A membership to The Power Rank does not offer picks on football games. Instead, members get spread and total predictions on games and other analytics to help you make better decisions.
I spent a long time searching the web for analytics rather than picks. I finally found it with The Power Rank.
I signed up for the power rating and point spread analysis for football and am quite pleased with the offerings.
I need to have consistent line anticipation/prediction through objective reasoning to begin my weekly evaluation. Your site offers such reasoning on an excellent basis.
You provide an excellent foundation for determining the correct line. This assists me greatly with establishing my betting line for the week, and I’m able to adjust my line for non-objective factors knowing that objective factors have been taken into account.
I would recommend a membership to anyone in search of strong analytical reasoning and objective line determination.
— Jeff, Las Vegas, Nevada.
Become a member of The Power Rank
When you come in and bet me, believe me, I’m going to move the hell out of those numbers.
Chris Andrews, director of the South Point sports book on The Football Analytics Show
As a member of The Power Rank, you get the following:
- Predictions for the spread in every NFL game.
- Predictions for the spread in every college football game and totals starting week 5. These predictions are posted on Sunday to take advantage of a beatable early market.
- Weekly member analysis during football season that includes bets on games and props.
- Predictions for the spread in every college basketball game.
- Data files with the NFL and college football predictions. You can upload these files to a spreadsheet to perform your own analysis.
- My full bracket advice for the NCAA tournament.
- A Member Content Library with evergreen material on home/rest advantage in football, the wisdom of crowds and more.
The March Madness bracket advice has helped many members win their pool.
Thanks to The Power Rank, I won my pool in 2019. I had Virginia winning.
Guess how many games I’ve watched this season? Literally zero. I only used the cheat sheet you provide through my membership to The Power Rank.
I had guys in my pool that have watched college basketball for decades, and I beat them all.
— Mark Carvalho, member of The Power Rank
All football predictions are posted by Tuesday at noon Eastern at the very latest during the football season. In 2022, the college football predictions are posted on Sundays, and an email gets sent to members.
It starts with an algorithm I’ve developed that does for sports analytics what Google’s PageRank did for web search: brings order and insight into a messy system. This algorithm leads to NFL predictive analytics that have gone 52.4% against the spread from 2020-22.
A membership to The Power Rank is neither a handicapping service that provides picks or a get rich quick program. If you want either of these things, please save us both time and do not become a member.
Instead, you might be interested in the process behind making football bets. With the importance of predictive analytics, you might want to outsource these calculations to an expert for less than $5 a week.
The Power Rank’s predictive analytics help you make better decisions on bets. They might convince you to make a bet here or talk you out of a bet there.
Not every decision will work out in your favor. However, let’s suppose a few more decisions work for you than against you. Even with a $25 bet size, this makes a membership well worth it.
To become a member of The Power Rank for $99 per year and get instant access to the PhD level predictive analytics and premium content, click on the orange “Add To Cart” button.
This membership renews every year. You can cancel anytime.
30 day, no hassle, money back guarantee
Please take up to 30 days to evaluate any of the predictions and analytics of a membership to The Power Rank. If you’re not satisfied for any reason, I will refund your money.
No questions asked, no hard feelings.
Simply reply to the email you’ll get when you purchase. I will personally take care of the rest.