Gemini 1.10.0
March 13, 2023
New Feature Release Notes
Feature Name: Pre-Built AI
- Users can now access foundational modeling projects built by the GSA team to filter, sort, and explore problems in their sport
- Access through left-hand sidebar of the application, then click on any prebuilt project. Details of the underlying model(s) and the data behind it are currently available upon request and will soon be on display in the application interface.
- First Project: Major League Soccer Players Predicted Value v Actual Salaries
- Purpose: Predict on-field player valuation and market value in the MLS
- View a list of MLS players and their predicted market value performance (GAAR, goals added above replacement) compared to current salaries as derived with a Random Forest Extra Trees model utilizing player characteristics and performance from the three most recent seasons.
- Step-by-Step Instructions - Step 3 and 4 of the UserGuide
- Data used to create this prediction was publicly available. This project will be iterated over and improved upon with new data sources in the coming months.