IGN data ; analysis of the scores
Jonathan Bouchet
October 20, 2016
In this notebook I wiil look at how IGN’s scores changed through time/platform/genre ## data file
df<-read.csv('/Users/jonathanbouchet/Desktop/WORK/PROJECT/gaming/IGN/ign.csv',sep=',')
#fix one entry having release_year = 1970
df[df$release_year<1980,]## X score_phrase title
## 517 516 Great The Walking Dead: The Game -- Episode 1: A New Day
## url platform
## 517 /games/the-walking-dead-season-1-episode-1/xbox-360-135866 Xbox 360
## score genre editors_choice release_year release_month release_day
## 517 8.5 Adventure N 1970 1 1df[df$release_year<1980,'release_year']<-2012
df[df$release_year<1980,'release_month']<-4
df[df$release_year<1980,'release_day']<-27
library(ggplot2)
library(gridExtra)
library(dplyr)##
## Attaching package: 'dplyr'## The following object is masked from 'package:gridExtra':
##
## combine## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionlibrary(circlize)#platform
sony<-c('PlayStation','PlayStation 2','PlayStation 3','PlayStation 4' ,'PlayStation Portable','PlayStation Vita')
micro<-c('Xbox','Xbox 360','Xbox One')
bigN<-c('Game Boy','Game Boy Advance','Game Boy Color' ,'GameCube', 'NES', 'New Nintendo 3DS' ,'Nintendo 3DS','Nintendo 64', 'Nintendo 64DD','Nintendo DS', 'Nintendo DSi','Super NES','Wii','Wii U')
sega<-c('Dreamcast','Dreamcast VMU','Genesis', 'Master System','Saturn' ,'Sega 32X','Sega CD')
bandai<-c('WonderSwan','WonderSwan Color')
nokia<-c('N-Gage')
neogeo<-c('NeoGeo','NeoGeo Pocket Color')
nec<-c('TurboGrafx-16','TurboGrafx-CD')
atari<-c('Atari 5200','Atari 2600','Lynx')
apple<-c('iPad','iPhone','iPod','Macintosh')
win<-c('Pocket PC','PC','Windows Phone','Windows Surface')
and<-c('Android','Ouya')
lin<-c('SteamOS','linux')
#genre
action<-c('Action','Action, Adventure','Action, Compilation','Action, Platformer','Action, Puzzle','Action, RPG','Action, Simulation','Action, Editor','Action, Strategy')
adventure<-c('Adventure','Adventure, Adventure','Adventure, Compilation','Adventure, Episodic','Adventure, Platformer','Adventure, RPG','Adventure, Adult')
sport<-c('Baseball','Sports','Sports, Action','Sports, Baseball','Sports, Compilation','Sports, Racing','Sports, Simulation','Sports, Editor','Sports, Fighting','Sports, Golf','Sports, Other','Sports, Party')
fighting<-c('Fighting','Fighting, Action','Fighting, Adventure','Fighting, Compilation','Fighting, RPG','Fighting, Simulation')
platform<-c('Platformer','Platformer, Action','Platformer, Adventure')
racing<-c('Racing','Racing, Action','Racing, Shooter','Racing, Simulation','Racing, Editor','Racing, Compilation')
rpg<-c('RPG','RPG, Editor','RPG, Simulation','RPG, Action','RPG, Compilation')
sim<-c('Simulation','Simulation, Adventure')
music<-c('Music','Music, Action','Music, Compilation','Music, Editor','Music, Adventure','Music, RPG')
puzzle<-c('Puzzle','Puzzle, Action','Puzzle, Compilation','Puzzle, Adventure','Puzzle, Platformer','Puzzle, RPG','Puzzle, Word Game','Puzzle, Compilation')
strategy<-c('Strategy','Strategy, RPG','Strategy, Compilation','Strategy, Simulation')
edu<-c('Educational','Educational, Action','Educational, Puzzle','Educational, Trivia','Educational, Adventure','Educational, Card','Educational, Productivity','Educational, Simulation')
flight<-c('Flight','Flight, Action','Flight, Racing','Flight, Simulation')
shooter<-c('Shooter','Shooter, Platformer','Shooter, RPG','Shooter, Adventure','Shooter, First-Person')
party<-c('Party')
card<-c('Card','Card, Battle','Card, Compilation','Card, RPG')
pinball<-c('Pinball','Pinball, Compilation')
compil<-c('Compilation','Compilation, RPG')
hunting<-c('Hunting','Hunting, Action','Hunting, Simulation')
wrestling<-c('Wrestling','Wrestling, Simulation')
prod<-c('Productivity','Productivity, Action')I create new columns for these 2 new features
newManufacturer<-function(x){
if (x %in% sony == TRUE) {return('SONY')}
else if(x %in% micro == TRUE) {return('MICROSOFT')}
else if(x %in% bigN == TRUE) {return('NINTENDO')}
else if(x %in% sega == TRUE) {return('SEGA')}
else if(x %in% bandai == TRUE) {return('BANDAI')}
else if(x %in% nokia == TRUE) {return('NOKIA')}
else if(x %in% neogeo == TRUE) {return('NEOGEO')}
else if(x %in% nec == TRUE) {return('NEC')}
else if(x %in% atari == TRUE) {return('ATARI')}
else if(x %in% apple == TRUE) {return('APPLE')}
else if(x %in% win == TRUE) {return('WINDOWS')}
else if(x %in% and == TRUE) {return('ANDROID')}
else if(x %in% lin == TRUE) {return('LINUX')}
else{return('OTHER')}
}
genreBetter<-function(x){
if(x %in% action == TRUE) {return('ACTION')}
else if(x %in% adventure == TRUE) {return('ADVENTURE')}
else if(x %in% sport == TRUE) {return('SPORTS')}
else if(x %in% fighting == TRUE) {return('FIGHTING')}
else if(x %in% platform == TRUE) {return('PLATFORM')}
else if(x %in% racing == TRUE) {return('RACING')}
else if(x %in% rpg == TRUE) {return('RPG')}
else if(x %in% sim == TRUE) {return('SIMULATION')}
else if(x %in% music == TRUE) {return('MUSICAL')}
else if(x %in% puzzle == TRUE) {return('PUZZLE')}
else if(x %in% strategy == TRUE) {return('STRATEGY')}
else if(x %in% edu == TRUE) {return('EDUCATIONAL')}
else if(x %in% flight == TRUE) {return('FLIGHT')}
else if(x %in% shooter == TRUE) {return('SHOOTER')}
else if(x %in% party == TRUE) {return('PARTY')}
else if(x %in% card == TRUE) {return('CARD')}
else if(x %in% pinball == TRUE) {return('PINBALL')}
else if(x %in% compil == TRUE) {return('COMPIL')}
else if(x %in% prod == TRUE) {return('PRODUCTIVITY')}
else if(x %in% wrestling == TRUE) {return('WRESTLING')}
else if(x %in% hunting == TRUE) {return('HUNTING')}
else{return('OTHER')}
}df$newGenre<-sapply(df$genre, genreBetter)
df$newPlatform<-sapply(df$platform, newManufacturer)General plots
IGN has 3 notations :- a numeric score (float) : from 0 to 10
-
a alphanumeric notation :
sort(unique(df$score_phrase))## [1] Amazing Awful Bad Disaster Good ## [6] Great Masterpiece Mediocre Okay Painful ## [11] Unbearable ## 11 Levels: Amazing Awful Bad Disaster Good Great Masterpiece ... Unbearable - an editor’s choice (Y/N)
scoreNoteVsPhrase<-ggplot(data=df,aes((score))) + geom_histogram(aes(fill=factor(score_phrase)),bins=83) + xlab('score') + theme(legend.position=c(.2, .6))+ theme(legend.title=element_blank())
scoreNoteVsChoice<-ggplot(data=df,aes((score))) + geom_histogram(aes(fill=factor(editors_choice)),bins=83) + xlab('score') + theme(legend.position=c(.2, .7)) + theme(legend.title=element_blank())
score_phraseVsChoice<-ggplot(data=df,aes((score_phrase))) + geom_bar(aes(fill=factor(editors_choice))) + theme(legend.position=c(.2, .85)) + xlab('score -word-') + theme(legend.title=element_blank())
print(scoreNoteVsPhrase)
print(scoreNoteVsChoice)
print(score_phraseVsChoice)
#grid.arrange(scoreNoteVsPhrase,scoreNoteVsChoice,score_phraseVsChoice,ncol=3)Scores vs. year and platform and genre
Did IGN change its notation through the years or are there platforms and/or genres that got better scores ?
vs platform
short answer : not really
ggplot(data=subset(df,newPlatform %in% c('SONY','MICROSOFT','NINTENDO')),aes(x=factor(release_year),y=score)) + geom_boxplot(aes(fill=(newPlatform))) + theme(legend.position=c(.1, .2)) + xlab('Year') + ylab('score')+ theme(legend.title=element_blank()) 
It looks like Microsoft’s consoles get better score than Sony’s or Nintendo’s, but not true for the last ~ 4-5 years. It looks also there is a wave pattern with a dip around year 2008 (2008 is also know as the golden year, whch saw the highrer number of titles, all consoles)
vs genre
To select the best genres through all platforms :
sort(table(df[,c('newGenre')]),decreasing=TRUE)##
## ACTION SPORTS SHOOTER RACING ADVENTURE
## 5019 2184 1640 1470 1200
## STRATEGY PUZZLE RPG PLATFORM FIGHTING
## 1150 1040 995 842 647
## SIMULATION OTHER MUSICAL FLIGHT CARD
## 568 450 422 189 174
## PARTY WRESTLING HUNTING PINBALL EDUCATIONAL
## 141 135 115 78 69
## COMPIL PRODUCTIVITY
## 56 41ggplot(data=subset(df,newGenre %in% c('ACTION','SPORTS','SHOOTER','RACING')),aes(x=factor(release_year),y=score)) + geom_boxplot(aes(fill=(newGenre))) + theme(legend.position=c(.1, .2)) + xlab('Year') + ylab('score')+ theme(legend.title=element_blank()) 
There is a small dependence all the genre’s score vs time : prior 1998-1999, scores were much lower than what they are now.
Comments