Hackathon

Next batter up. Compete for $10,000 in total prizes and a chance to present on stage at Women in Sports Data 2024

Overview

The Women in Sports Data Hackathon is a project-based data competition open to women and non-binary individuals who have never worked full-time in professional sports, offering the opportunity to work with exclusive tracking data sets typically unavailable to the general public and tackle real-world problems with guidance from industry professionals. Previously we've played with soccer and basketball data. This year, we're throwing up the bat signal...

⚾️ 2024 Theme: Baseball Bat Tracking

Bombs away: it's time to swing for the fences. We love a good dinger. This year, we're investigating the fundamentals of hitting using bat tracking data. Is all bat speed created equal? How might the position of the bat on contact affect bat speed (since bat speed is measured at contact). How do players make the most of the bat speed they do create?

Your challenge is to submit a project geared toward making a recommendation for a baseball front office or coaching staff using provided bat tracking data, in the form of functioning application prototype or an analysis notebook. This year's competition features anonymized bat tracking data from Phillies Minor League affiliate games. Participants are also encouraged to use MLB's recently released subset of public bat tracking data, as well as any other publicly available data set, such as biography data, to enhance your projects.

The top three submissions will receive an invitation to present their work at the 2024 Women in Sports Data Symposium at Citizens Bank Park on September 7, including travel and two nights stay in Philadelphia, where they will compete for $10,000 total in cash prizes.

Participants must register for the competition as individuals, but may compete either alone or in teams of up to four (4) people. After registering for the hackathon, you will then be invited to join a Slack community, where you can meet fellow contestants and mentors, form teams, and engage in related hackathon programming.

Timeline

May 20—June 16

Registration & Team Formation
June 17–July 28

Build Period
July 29–August 11

Judging
August 12

Finalists Notified
September 7

Finalists ➡️ Philadelphia

Project Guidelines

Bat tracking data consists of swings, sampled at a rate of 30 frames per second, resulting in hundreds of frames. For each swing over time, we sample location data for the bat head and handle, as well as the ball, at the frame level. In addition to location data, each swing has a frame labeled with the contact event, and every swing is associated with metadata about the entire play (e.g. pitching: pitch speed, pitch type; ball: exit velocity, launch angle.)

Here are some questions and project prompts to spark your creative process:

What makes a swing "good"?
What physical factors impact a player's bat speed?
How can we measure contact quality with bat and ball positions?
Can we learn anything about the pitcher based on the swings that hitters take?

Example project prompts include, but are not limited to:

Develop metrics to assess swing quality and how they relate to performance (e.g. exit velocity, launch angle, swing and miss percentage)
Create an interactive visual for a coach or analyst to compare two swings (including swing characteristics and evaluation)
Engineering challenges with tracking data: develop an ETL pipeline mapped to publicly available data, build an API interface

Requirements

The only hard requirement for project submissions is that you must use the provided tracking data set.

Projects should take the form of:

A functional minimum viable product (MVP) deployed as an interactive website (e.g. React, Svelte, Dash, Shiny), AND/OR
A data science project or notebook with targeted, actionable recommendations based on analysis (e.g. Jupyter notebook, R Markdown AND slide deck, PDF)

Be creative! In addition to being technically innovative, projects should consider the decision-making needs of a non-technical front office or coaching audience, especially in the final presentation.

Participants must register for the competition as individuals, but may compete either alone or in teams of up to four (4) people. Previous finalists have typically submitted projects as individuals or teams of two.