Online Program

Return to main conference page
Friday, May 31
Engaging Students in Statistics & Data Science
Fri, May 31, 3:40 PM - 5:15 PM
Grand Ballroom I

STEAMS Approach on Playing Video Games (305123)

Charles Chen, Applied Materials 
*Mason Chen, Stanford OHS 

Keywords: ROI, Regression, Cluster, JMP

This STEAMS project is to demonstrate how to play Video Game through STEAMS approach and learning science and statistics. Kids are playing video games too long and parents do not want kids to play video games since most video games are not developing critical thinking. Hill Climb Car Racing game was chosen not based on commercial rating but on potential of applying statistical data-driven methodology. The author took STEAMS approach: Science (Physics, Mechanics), Technology (Car Upgrading), Engineering (Failure Mode), Artificial Intelligence (Clustering), Math (Geometry, Trigonometry) and Statistics (Mixture DOE Optimization). Based on the engineering failure mode analysis and scientific understanding, author can develop a systematic car upgrading system through statistical modeling to optimize car performance. Simple linear regression was conducted to quantify the ROI return (car travelling distance) of investment (car grading cost). The regression model accuracy has been improved from original 66% (random playing mode) to 92% (systematic playing mode). The ROI slope has been improved from 147.2 to 512.4 meter/upgrade unit. The clustering analysis grouped the similar field stages with common challenges and science which has helped upgrade car to support multiple stages. The BEST car of each stage has matched well with literature research. It’s a very successful STEAMS project on playing video games.