Predicting Tetris Performance Using Early Keystrokes

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review

Abstract

In this study, we predict the different levels of performance in a Nintendo Entertainment System (NES) Tetris session based on the score and the number of matches played by the players. Using the first 45 seconds of gameplay, a Random Forest Classifier was trained on the five keys used in the game obtaining a ROC-AUC score of 0.80. Further analysis revealed that the number of down keys (forced drop) and the number of left keys (left translation) are the most relevant keys in this task, showing that by merely including the data from these two keys our Random Forest Classifier reached a ROC-AUC score of 0.83. We conclude that the keylogger data during the early phases of a game session can be successfully used to predict performance in longer sessions of Tetris.

Original languageEnglish
Title of host publicationProceedings of the 18th International Conference on the Foundations of Digital Games, FDG 2023
EditorsPhil Lopes, Filipe Luz, Antonios Liapis, Henrik Engstrom
PublisherACM
Pages1-4
Number of pages4
ISBN (Electronic)9781450398565
ISBN (Print)978-1-4503-9855-8
DOIs
Publication statusPublished - 12 Apr 2023
EventFDG 2023: Foundations of Digital Games 2023 Lisbon Portugal - Lisbon, Portugal
Duration: 12 Apr 202314 Apr 2023

Publication series

NameACM International Conference Proceeding Series

Conference

ConferenceFDG 2023: Foundations of Digital Games 2023 Lisbon Portugal
Country/TerritoryPortugal
CityLisbon
Period12/04/2314/04/23

Keywords

  • Expertise
  • Machine Learning
  • Performance
  • Peripherals
  • Tetris
  • Video games

Fingerprint

Dive into the research topics of 'Predicting Tetris Performance Using Early Keystrokes'. Together they form a unique fingerprint.

Cite this