How RosterOps Works

Behind the User-Interface, RosterOps is analyzing a massive dataset for patterns of success, to enable quick insights. An overview of how this is achieved is below:

DATA

  • RosterOps considers over 3.5 million statistics* of past player and roster performance
  • Data is rationalized by division, conference and position allowing like-to-like comparison

DATA SCIENCE

  • Machine learning algorithms recognize patterns of success** for teams which made a deep tournament run, and those which did not
  • Analysis of the entire historical dataset affords time and efficiency savings
  • Meaningful factors such as efficiency, experience and team balance are emphasized within the algorithm

RESULTS

  • Confidence through analytics to compliment your coaching style and validate your instincts with a data-backed perspective
  • Quickly highlight the most impactful players in your roster
  • Easily identifying hidden strengths and potential vulnerabilities of roster variations

* Five historical seasons considered, over 20,000 players each season, and well more than 35 key statistics supporting the player score model = >3.5MM datapoints
** Teams which made the Sweet Sixteen in the NCAA Tournament are considered to make a deep tournament run