Rapid Skill Capture in a First-Person Shooter

TL;DR

This paper introduces a method to quickly assess player skill in a first-person shooter using minimal input data, enabling faster game adaptation than traditional performance-based methods.

Contribution

It presents a novel analysis of skill metrics from brief input samples and demonstrates their effectiveness for rapid skill estimation in gaming.

Findings

Skill metrics can be predicted from seconds of input data

Rapid skill estimation outperforms traditional averaging methods

A new dataset of game logs from 40+ players is provided

Abstract

Various aspects of computer game design, including adaptive elements of game levels, characteristics of 'bot' behavior, and player matching in multiplayer games, would ideally be sensitive to a player's skill level. Yet, while difficulty and player learning have been explored in the context of games, there has been little work analyzing skill per se, and how it pertains to a player's input. To this end, we present a data set of 476 game logs from over 40 players of a first-person shooter game (Red Eclipse) as a basis of a case study. We then analyze different metrics of skill and show that some of these can be predicted using only a few seconds of keyboard and mouse input. We argue that the techniques used here are useful for adapting games to match players' skill levels rapidly, perhaps more rapidly than solutions based on performance averaging such as TrueSkill.