Multi-Parameterised Matchmaking: A Framework

Anders Harboell Christiansen, Emil Gensby, Bo Friis Nielsen

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

The competitive scene in online video games is becoming more and more prominent and player satisfaction is of key importance when it comes to a good user experience and a successful game. As such it is important to have efficient skill rating and matchmaking systems in order to provide a proper match experience. We propose a mathematical framework for the analysis of matchmaking systems. The mathematical model addresses the estimated skill or rating, calculation of winning probabilities based on the estimated skill, and the updating of the estimated skill upon completion of a game. We will briefly apply the framework to the ELO skill rating system. Next we will use the framework to analyse the robustness of the TrueSkill algorithm and discuss some of the findings. We have used simulated data to test the robustness of the TrueSkill algorithm. All of the data processing has been done in Python using our own code, built-in functions and Python packages. The code has primarily been used to make the simulations of matches and customise updating functions.
Original languageEnglish
Title of host publicationProceedings of 2018 IEEE Conference on Computational Intelligence and Games
Number of pages4
PublisherIEEE
Publication date2018
ISBN (Print)9781538643594
DOIs
Publication statusPublished - 2018
Event2018 IEEE Conference on Computational Intelligence and Games - School of Business and Economics of Maastricht University, Maastrict, Netherlands
Duration: 14 Aug 201817 Aug 2018

Conference

Conference2018 IEEE Conference on Computational Intelligence and Games
LocationSchool of Business and Economics of Maastricht University
CountryNetherlands
CityMaastrict
Period14/08/201817/08/2018
Series2018 Ieee Conference on Computational Intelligence and Games (cig)
ISSN2325-4289

Keywords

  • Matchmaking
  • Mathematical model
  • Skill rating
  • Framework
  • Player satisfaction
  • Online video games

Fingerprint Dive into the research topics of 'Multi-Parameterised Matchmaking: A Framework'. Together they form a unique fingerprint.

Cite this