Skip to content

Using deep learning to determine chances of winning a match of League of Legends in champion select

Notifications You must be signed in to change notification settings

dwang733/lol-winrate

Repository files navigation

lol-winrate

In this project, we use a neural network to determine the chances of winning a match of League of Legends using only data from champ select and the summoners' previous matches.

Most data was collected through Riot's official API, with overall champion stats (like winrate, ban rate, etc.) collected through champion.gg's API and basic champion stats collected through CommunityDragon.

Using the fast.ai library, we achieved an accuracy of around 73%.

About

Using deep learning to determine chances of winning a match of League of Legends in champion select

Topics

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published