Tractable Victory-Probability Objective for Rotisserie Fantasy Sports Optimization
In Rotisserie fantasy leagues the natural objective, a team's probability of winning the season-long standings, is intractable because it sums over all orderings of teams across every scoring category. This work derives a tractable, differentiable approximation of that victory probability under a simplified model in which category outcomes are Normally distributed, yielding an objective that implicitly rewards a balanced roster. The approximation supplies the format-specific objective function that lets the dynamic H0 drafting algorithm, previously limited to head-to-head play, be applied to Rotisserie player selection.
Optimizing for Rotisserie Fantasy Basketball Zach Rosenof January 2025 Abstract Previous work on
In Rotisserie fantasy sports leagues, teams accumulate category totals over a season, are ranked within each category for points, and the team with the most total category points wins. The natural ob…