MASBOS: Mixture of Advanced Statistics and Betting Odd Scores

The Mixture of Advanced Statistics and Betting Odd Scores (MASBOS), is an all-in-one metric that perspectives from other Advanced Statistics (such as Box Plus-Minus, Estimated Plus-Minus, RAPTOR, and DARKO) to estimate a player's overall value while accounting for Betting Odds from the house (using Pinnacle).

The purpose of our system is to notice whether or not, using the perspectives of multiple Advanced Statistics, that capture information such as a player's impact, on/off ratings, offensive load, defensive versatility, etc, helps with modeling a player's overall status.

The problem with other traditional All-in-one metrics such as Player Efficiency Rating (PER), or Box Plus-Mins (BPM), is that they tend to overrate and underrate various aspects of the game. PER & BPM both do not account for Team Seed. PER & BPM do not account for the fact that previous MVPs do have an effect on a player's odds of winning (Voter Fatigue). We hope that by incorporating Betting Odds, and impact ratings, we can create an All-in-one statistic, that may not perfectly model a player's offense or defensive metrics but does a good enough job modeling everything about a player.

To prevent overfitting, we only took a few parameters, that we thought would best fit the overall message of our MASBOS statistic. We took betting odds and combined them with DARKO (The most popular metric used in NBA Box Office Teams due to its forecasting nature), Luck-adjusted player Estimate using a Box prior to Regularized ON-off (LEBRON) which measures Impact, Estimated Plus-Minus (EPM), another impact metric tends to favorite Defensive Ability and On-Off over LEBRON, and finally ,PER (Player Efficiency Rating), as it does a great job modeling a player's efficiency (their True Shooting, Impact Per Minute, how many numbers they put it adjusted for their time, etc).

  • Incorporates player efficiency, plus-minus, and win shares
  • Factors in real-time betting odds for more accurate predictions
  • Provides a single, easy-to-understand score for player impact
  • Updates dynamically throughout the season

When looking at MASBOS, the score is interpreted as follows:

  • A MASBOS of4.0+ is considered a good player (a player who overall contributes to winning and is a positive asset, or Teams with a Positive Win Record)
  • A MASBOS of6.0+ is considered an All-Star level player (think Zion Williamson, a player who has a great impact on a team on the winning side, or Dark Horse NBA Teams)
  • A MASBOS of 9.0+ is considered Amazing (near MVP-level, 2nd seeds, etc)
  • A MASBSOS of 10.0 indicates a High Likelihood of Finishing on Top (Whether it be MVP, the best team, etc)