An interesting article was posted by Benjamin Cronin regarding using win totals to predict week one winners. I thought I would take his information a step further and create a league wide power ratings chart that can be maintained throughout the entire season.
Market efficiency is a term in sports betting which relates to how accurate the prices in a market are. The more efficient a market is, the harder it is to beat as a bettor. A market can be more or less efficient based on how much of all available information has already incorporated into the price.
While you can never be certain a market is 100% efficient, NFL season win totals are among the most efficient markets bettors will encounter during a calendar year of betting. This high efficiency is because bookmakers have had north of three months to adjust their prices based on team information and react to how bettors have bet themselves.
For most bettors, the minute the NFL season kicks off, team win total markets disappear in their minds. Ignoring win totals once the season starts is a foolish practice. Because of the high efficiency in the team win total markets, bettors can use this to their advantage during the season by creating accurate power ratings based off the numbers.
Power ratings are something that all bookmakers use to evaluate teams and quickly come up with a starting line for weekly matchups. Each team begins the season with a rating and it is adjusted up or down each week based on team performance. Instead of coming up with a brand new opinion for each team each week, bookmakers collectively build on their past opinions throughout the season.
Power ratings are something that all bettors should create for themselves. By creating power rankings, bettor can focus on making their rankings more accurate as the season goes on. It is easier to be accurate in grading what a team has done rather than being accurate in predicting what they will do. When bettors have their power ratings, they can compare the rating prices vs the market price and make bets without emotion playing a factor.
Creating power ratings for a league can seem a daunting task. Building confidence in power ratings can also take a long time to earn, especially if the season starts off on a losing note. However, as I mentioned above, bookmakers have already done all the hard work for bettors.
Here is a quick method to use the highly efficient win total markets to create power ratings.
First, open an excel spreadsheet and list all 32 teams in the league with their current win total price.
Second, add in the Pythagorean Expectation for all teams based on their 2015 performance. I have included the values below to save time calculating.
In short, the Pythagorean Expectation is a calculation uses to generate an expected win total for each team based on their points scored and points allowed. Teams that won more games than their expectation can be expected to regress this season. Teams that won less games than their expectation can be expected to improve this season. Factoring the Pythagorean Expectation difference into the already efficient win total market makes the power ratings even more accurate.
Third, the final step is simply adding the two columns together and create the expected wins column for 2016 which you will use for creating point spreads. Then create a fourth column to round each figure to the nearest half win.
Now that each team has their adjusted win total determined, the ratings can be used to generate point spreads. A general rule to follow is that for each half win a team is better than their opponent, they should be favored by 1 point on a neutral field.
The reason for this is that a moneyline for a typical 3 point favourite in the NFL is -145. The implied probability of a -145 moneyline is 59%. If an average NFL team is expected to win 8 games over the course of the season, then a team favoured by 3 points in each game would be expected to win 59% of their 16 games, or 9.5 games. If you subtract 9.5 wins from a team valued 3 points better than the league from an team considered average at 8 wins, you get a difference of 1.5 wins, or 0.5 wins per 1 point.
Here is what the point spread looks like for all 16 games this week using these power ratings.
There are six teams in week one, Denver, Tennessee, NY Jets, Jacksonville, Seattle and New England which have a difference of more than field goal compared to the current point spread. These will be considered “Top Bets” and will be tracked each week throughout the season.