Developing a Nation-Wide Padel Rating System: A Data-Driven Approach
summary
A new padel rating replaces tennis tooling with a data-first system trained on 300,000 matches, compares Elo variants and a game‑index model, and uses simulations to tune convergence and stability. It adds a gender correction factor for asymmetrical pairings, removes the artificial ceiling by raising the upper boundary to 11, doubles maximum step size and other K‑like adjustments, and reports the isolated change makes predictions about 9.6% more accurate than the old system.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 3:01What is Padel
- 6:55Old rating problems
- 8:42Asymmetrical matches
- 12:55Boundary issues
- 14:22Rating adjustment speed
- 17:30New model design
- 20:45New system performance
- 22:31Final results
- 23:13Audience questions and answers
related talks
45mThe Technical Journey to Define the True Value of a Football Goal
33mUsing Data Science to Better Evaluate American Football Players
35mBuilding a Player-Focused Data Team
38mDeep Learning plays Handball
09mSmall Data in Sports: Little Differences that Mean Big Outcomes
20mIEEE CoG 2019 - Industry Day - Marko Knezevic (Nordeus)
28mPatterns of an AI lineup card
49mProper Scoring Rules for Text and Super-alignment