PandaScore bolsters League of Legends betting with PandaSkill

PandaScore, the Paris-based esports data and odds provider, has launched PandaSkill, a machine learning-powered skill rating system designed for League of Legends.

This isn’t just another stat tracker; PandaSkill promises to revolutionise how fans, bettors and analysts measure player performance.

What sets PandaSkill apart? Traditional systems like Elo or TrueSkill often miss the mark in esports, where team dynamics, regional silos, and role specialisation muddy the waters.

PandaSkill evaluates each player’s in-game performance by role, using a model that compares individuals on a global scale. No more apples-to-oranges comparisons between a star support and a top-lane carry; every role is judged on its own merits, and every player is measured against the world’s best.

Instead of focusing on match wins or losses, PandaSkill updates ratings based on individual performance metrics.

This means a player can shine, even in defeat, and see their rating rise. The system uses the Bayesian OpenSkill framework, crunching data from five years of professional matches to ensure fairness and accuracy.

As a result, fans and bettors can track skill across regions, thanks to a dual-rating system that blends local and global benchmarks.

But PandaSkill isn’t just for the number crunchers at PandaScore. It’s open source, ready to be embedded in apps, websites, or esports publications. Bettors get sharper odds, operators gain transparency, and fans have a new tool to fuel debates and guide wagers.

Oliver Niner, PandaScore’s Head of B2B, explained: “PandaSkill takes performance analysis to the next level, helping inform public discussion, culture building, and broadcast commentary, as well as helping players make smarter bets while also allowing us to improve the products and services we offer to operators.”