Scribd
Upload
47 views2 pages

Gaming Skill & Intelligence Study

This document discusses data pre-processing for a study exploring the relationship between video game expertise and fluid intelligence. It describes how rankings like ELO ratings, MMR ratings, and tiers/divisions in League of Legends correspond to player skill levels based on win/loss performance. It also notes that distributions were inspected for outliers and outliers were removed according to Tukey's method before analysis, and provides example histograms of rankings before and after outlier rejection for several games.

Uploaded by

Miguel Díaz
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
47 views2 pages

Gaming Skill & Intelligence Study

This document discusses data pre-processing for a study exploring the relationship between video game expertise and fluid intelligence. It describes how rankings like ELO ratings, MMR ratings, and tiers/divisions in League of Legends correspond to player skill levels based on win/loss performance. It also notes that distributions were inspected for outliers and outliers were removed according to Tukey's method before analysis, and provides example histograms of rankings before and after outlier rejection for several games.

Uploaded by

Miguel Díaz
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 2

Exploring

the relationship between video game expertise and fluid Intelligence


A.V. Kokkinakis, P. Cowling, A., Drachen, A.R. Wade

Supplementary Material – 1
Data pre-processing

ELO/MMR ratings
After 10 placement games LoL players are assigned a Tier and a Division based on their win/loss
performance. They subsequently move up and down on that Division and move between Tiers based
on their win ratio. Tiers and Divisions therefore correspond to MMR ranges. In addition, there may
be non-linearities in MMR within and between Tiers and data in our relatively small sample do not
pass standard tests for normality. For these reasons, we show parametric statistics here only for
visualisation purposes and all correlations are computed using Spearman’s rho. We note that the
results from parametric and non-parametric analyses are almost identical.

Study 1
Rank Nonparametric correlations
We treat our data as parametric for visualisation purposes. However, League Tiers and Divisions are
not normally distributed and thus we opted for non-parametric calculations throughout our analysis.
We note that the results from parametric and non-parametric analyses are almost identical.

Study 2
All distributions were inspected for outliers and we then used Tukey’s outlier technique (k=2.2) to
identify other candidate points. Here we provide illustrations of the distributions before and after
outlier rejection.

Note all these terms in the x-axis (ELO, Matchmaking Rating, Combat_PVP) are interchangeable
(different naming conventions depending on the company) and they use win-ratio as a primary
determinant of a player’s “skill-level”.

League of Legends

S1 Fig 1 Histograms of League of Legends MMRs before (Left) and after (right) outlier rejection


Exploring the relationship between video game expertise and fluid Intelligence
A.V. Kokkinakis, P. Cowling, A., Drachen, A.R. Wade

Battlefield 3

S1 Fig 2 Histograms of Battlefield3 ELOs before (Left) and after (right) outlier rejection

Destiny

S1 Fig 3 Histograms of Destiny PVPs before (Left) and after (right) outlier rejection

DOTA 2

S1 Fig 4 Histograms of DOTA2 MMRs. No outliers were detected in this dataset

You might also like