WPI (Worcester Polytechnic Institute)

Computer Science Department
------------------------------------------

Analysis of Data Gathered from League of Legends and the Riot Games API

Artian Kica, Andrew La Manna, Lindsay O'Donnell, Tom Paolillo

Advisor: Mark Claypool

Major Qualifying Project MQP-MLC-LL15
Computer Science, WPI
Terms ABC 2015-2016

Abstract

Understanding the impact of game balance changes on character performance is important for game designers and players. For this study, we analyzed the effect of patches on champion win rates, pick rates, and ban rates from League of Legends, a popular online game created by Riot Games. This was achieved by categorizing patches by change type. These changes were correlated to game statistics sampled by pulling statistics from 128 champions and roughly 465,000 games from the Riot Games API. In order to be more conveniently accessible to the developers and playerbase, this information is displayed publically on a WPI hosted website on a per character basis. Analysis of the data shows that the rates of a champion have weak correlation with one another and that rates vary depending on the role of the champion being analyzed. The data also shows that Riot Games tends to increase the strength of champions more often than they decrease the strength of champions and that these increases and decreases have a weak correlation with win rate respectively.


Download

See also:


[WPI Homepage] [CS Homepage] [Mark Claypool's Homepage]