SMT Data Challenge

The newly added data challenge is sponsored by SportsMedia Technology (SMT), an industry leader in sports data collection and visualization.

Winners

  • Graduate
    • To Shift or Not to Shift? Using Player & Ball Motion Data to Build a Highly Flexible Defensive Positioning Algorithm in Baseball
  • Undergraduate
    • Should I Stay or Should I Go
    • Billy Fryer, North Carolina State University

Other Finalists

  • Graduate
    • Evaluating Minor League Outfielder Fly Ball Success using Player Tracking Data
    • Using Routine Plays to Assess Infielder Arm Ability
  • Undergraduate

Problem Statement

For this data challenge, your goal is to analyze an aspect of player movement (e.g. baserunning, movement while fielding, backing up a play) for minor league baseball players. SMT has provided in-game player and ball location data for multiple teams over multiple seasons. This spatiotemporal information can fuel a thoughtful analysis to answer questions that are difficult or impossible to answer with manually-collected data or subjective observation. Since this challenge provides the opportunity to work with previously-unavailable player tracking data, your analysis should involve player motion; this includes any topic that uses player location data over time. Below are a few example topics.

  • Baserunning
    • How could you evaluate a batter's chance to advance to second base on a ball in play? Or, for a runner on first base, the chance to advance from first-to-third on a single or first-to-home on a double?
    • What circumstances are most likely to induce pickoff throws? Given a situation involving pickoff throws, what baserunner behavior is most likely to result in a stolen base? A successful pickoff?
    • How would you evaluate a baserunner's ability to read a ball in play?
  • Fielding
    • When a ball in play reaches the outfield, what aspects of fielding are most important in preventing a baserunner from advancing, or a batter from taking extra bases?
    • How would you evaluate a player's fielding ability in the context of judgment and risk-taking? For example, how would you compare a player who attempts and fails to make a difficult play (possibly leading to an error) to a player that does not attempt to make a play?
    • How would you evaluate a fielder's ability to appropriately read a ball in play?
    • What attributes are most important to fielding and assisting on infield groundouts? Double plays?
  • General
    • How could you estimate expected runs throughout the course of a play based on player and ball locations, and how would you use that to evaluate player baserunning and fielding?
    • Which baserunning and fielding abilities are most predictive or most consistent from one season to the next?

We emphasize that this list is not exhaustive, and participants should feel free to study an aspect of player movement that interests them.

Submission requirements

Please submit

  • A short paper on your study in PDF format (max: 3000 words)
  • A GitHub repo link containing code files and .csv files with results

Submissions are due by Monday August 1, 11:59pm Eastern Daylight Time.

Judging

A panel of judges from across academia and the sports industry will judge your submissions based on the following:

  • How original is the analysis?
  • How applicable is the analysis?
  • How appropriate were the methods used?
  • How well did you communicate your findings? This includes both written text and visualizations. How did the use of facts, data-supported narratives, anecdotes etc. buttress your storytelling?

Winners will be notified in early September.

Live coding sessions

Times and dates will be provided to registered participants.

Registration

(This is independent of the symposium registration.) To register, please fill the Google Form. The registration has been closed.

If you have any questions, please contact Dr. Meredith Wills m.wills@smt.com.