Unleashing the Power of AI in MLB
Providing Better Predictions
Dykstra highlights the importance of data-powered predictive matchups. With AI, they can run scenarios faster and gain a better understanding of what's possible. From full-season predictions to batter tendencies, their models constantly churn out fresh information. For example, in 2023, their predictions were highly accurate. Understanding batter tendencies helps pitchers make more informed decisions. However, using data also has a cultural impact. Pitchers have their own sense of the game, and it's the team's job to empower them.They can toy with different matchups and run simulations to see how a game could unfold. This kind of insight is crucial for pitchers to enhance their performance. But it's not just about providing data; it's about using it in a way that respects the players' expertise.
The Rangers' data department has a wide range of models covering various aspects of the game, ensuring they are well-prepared for any situation.
Creating New Partnerships
Internal data talent is important, but successful MLB teams extend their working relationships. The Rangers collect data from different sources and use a combination of Apache Airflow and Astronomer's orchestration and observability platform to ensure timely insights for staff and players. Dykstra's department works closely with Astronomer to manage the Airflow implementation and handle the large amount of data.The dynamic nature of the game means data flow is not always consistent. If one piece of information takes longer, it can disrupt the entire process. Managing the supporting infrastructure requires continuous effort to look to the future.
It's not just about the professional level; they consider data from games across the country and the world. This global perspective helps them make more informed decisions and stay ahead in the competitive landscape.
Removing Manual Tasks
Baseball is a text-heavy industry, and the Rangers rely on scouts worldwide. Turning their written reports into useful data is a challenging task, but generative AI (Gen AI) comes to the rescue. There are many secret terms and codes used by scouts, making it difficult for one person to process all the information.Gen AI helps sort through these summaries, provide a dictionary to translate key phrases, and summarize the data. The Rangers are exploring different models, such as using the Llama LLM and collaborating with other technology partners like Databricks and Amazon. They are also looking at how to use retrieval-augmented generation to ingest the baseball rule book and provide useful information for staff and spectators.
For example, in healthcare and stadium rules, Gen AI can provide quick answers to common questions, making life easier for everyone involved.
Monitoring Other Factors
Player data is not the only source of competitive advantage. The Rangers also feed their models with external information like weather data. At Globe Life Field, with its retractable roof, weather conditions can vary significantly from other open stadiums.Every five minutes, they receive data from different fields, and understanding the weather dynamics is crucial for giving players feedback. For instance, they can tell a player that a certain hit would have been a home run in different weather conditions. The ability to provide immediate feedback helps players focus on the next game.
Astronomer's ability to meet data windows and deliver insights quickly after the game is essential for the team's success.
Building New Cultures
Industry experts stress the need to democratize data access. At the Rangers, this is clearly happening, especially with the manager, Bruce Bochy, embracing data-powered opportunities. A data analyst is embedded within the team to help coaches and players make the most of data.It's a continuous conversation about how to use data effectively. However, using data also comes with risks, and the Rangers must abide by MLB's strict rules and regulations. Success lies in understanding how data moves, where it comes from, and where it goes.
By following a clear path and communicating the data journey effectively, the Rangers are able to make the most of emerging technologies and gain a competitive edge.