top of page
  • Writer's picture@WizeOwlSports

Using Real Plus/Minus-Based Power Ratings to Bet NBA


On Friday, we unveiled forecasts from our Real Plus/Minus (RPM)-based NBA power ratings framework. This approach intends to predict how each team will perform over the coming seven days by aggregating numerical assessments of individual player contributions to their team's overall margin of victory per 100 (offensive and defensive) possessions, considering both the quality of teammates, opposing players, coaches and other factors.

In an evaluation of advanced statistics such as Wins Shares per 48 Minutes, Wins Produced per 48 Minutes, Player Efficiency Rating, Ridge-regression Adjusted Plus/Minus, and Adavnced Statistical Plus/Minus, FiveThirtyEight Statistician, Neil Paine, conlcuded that RPM's predecessor, Ridge-regression adjusted plus minus (RAPM), is consistently the most predictive advanced statistic available to the public. Our findings are consistent with this assertion, as our adaptation has been a consistently winning strategy since January 19th, amassing a record of 107-82-3.

FiveThirtyEight, in consultation with RPM developers, Steve Ilardi and Jeremais Ingleman, publishes offensive and defensive RPM ratings for each of the 30 NBA teams each week. WizeOwlSports adds a standard (optimized) home court advantage factor for each scheduled contest and compares opposing teams' RPMs to derive line estimates. We also quantitatively assess the impact of injuries to key players since the last RPM reset, and adjust the weekly figures where appropriate (for this task we use the season-long RPMs from ESPN).

Out the gate, this approach delivered a 3-1 record on the four games we recommended. This methodology followed Friday's debut by splitting the two highlighted games Saturday. The Brooklyn Nets delivered a sweat-free win as they breezed past the Pacers (from the position of 8-point underdogs). However, the Portland Trail Blazers lost by 10 to the Memphis Grizzlies as 3.5 point dogs. Yesterday's misstep is what we consider a 'good loss' though -- as it likely owes to a virtually unpredictable event. Portland's LaMarcus Aldridge aggravated his ailing thumb, and did not play in the second half; Memphis only assumed control of the game after Aldridge retreated to the locker room.

We have long been fans of the RPM methodology, and are quite pleased to be able to implement a process designed around this metric that we believe in philosophically, and that wins.

Happy betting!!

bottom of page