Joo, Jessica [480460860]/ Lee, Sunny (Yusun) [480392033]/Son, Dahii Cynthia [480274038]/ Zhang, Emily (Ziru) [480344685] | Dr. Dai, Wen | Friday morning tutorial (Carslaw Comp Lab 705-706)

library(readr)
Gambling <- read_csv("Gambling.csv")

S = Gambling$State
C = Gambling$GamblingCategory
D = Gambling$GamblingType
E = Gambling$Expenditure

Gambling_vs_casinos <- read_csv("Gambling vs casinos.csv")

Report

1 Dataset

*Please refer to Attachment7 to view the data used and Attachment8 for citations.

1.1 Aim

The aim of this report is to investigate the profitability of various gambling businesses in Australia by analysing factors that may point to the likelihood of gambling operator success, as observed in the current market. For the purpose of this study, the success of a gambling business will be numerically defined as the gross winnings of the operator, hereinafter referred to as ‘expenditure’. The investigation will be conducted by exploring the relationships, if any, between expenditure and different variables(that may indicate likelihood of success).

1.2 Domain Knowledge

According to Roy Morgan, an Australian market research company, research shows that gaming machines are the most common form of gambling as it comprises 56.7 percent of all the dollars gambled in Australia which is more than double all forms of other betting being at 20.7 percent. This is due to the low level of complexity and easy access of poker machines within Australia. Statistics (Online pokies, 2014) show NSW owns the most of 100,500 machines, VIC 30,000, ACT 30,000, NT 2,195, QLD 50,000, SA 13,113, TAS 3,680 and WA with the least of 1,750. Australia is known to own 20% of gaming machines in the world and this attracts large amounts of tourists. The top three tourist attractions in Australia are all casinos. Crown Casino in Melbourne is the most popular with up to 10.9 million visitors in a year, followed by Jupiter Casino in Queensland attracting 10.6 million tourists. Sydney’s Star Casino is next with 9 million visitors (Australasian Gaming Council, 2016). According to the Australasian Gaming Council, gambling holds a significant portion to Australias tourism dollar. In gambling, there are two types, games of pure chance and games combining both chance and skill. These include, gaming machines and forms of lotto. With games of pure chance, the odds can not be changed by the player. A possibility of gaining money exists however long term play will result in a net loss for the player. In games combining both skill and chance, the player’s knowledge and judgment could have an impact on their chance of winning. These include, card games and betting offered by bookmakers such as on horses and other sports (Gambler’s Help, 2018).

1.3 Classification of Variables

Variables Classification and Analysis
State Classification: Qualitative(categorical), nominal, 8 categories.
Categories: Australian Capital Territory, New South Wales, Northern Territory, Queensland, South Australia, Tasmania, Victoria, and Western Australia.
Gambling Category Classification: Qualitative(categorical), nominal, 3 categories.
Categories: Categories: Racing, gaming, and sports betting.
Gambling Type Classification: Qualitative(categorical), nominal, 3+ categories.
Example categories: Casino, gaming machines, lotteries, pools, keno are gambling types under the gambling category of ‘Gaming’.
Expenditure Classification: Quantitative, discrete.
Expenditure, in millions of AUD, is the net amount lost by the wager, which is the gross winnings of the operator.

1.4 Stakeholders Assessment

Anyone interested or involved in the Australian gambling industry/economy will be a possible stakeholder in this data investigation, particularly those interested in the profitability or financial income of various gambling businesses. Of these stakeholders, there will mainly be two groups: the operating and the waging side. Current/potential investors, workers, and owners involved in operations would be interested in the data in order to maximise their profit/investment(e.g. ideal location/ type of gambling business to invest in/start a gambling business), or scope the competition. Recreational gamblers will also be interested in the data as they would want to maximise their chance of winning - which category/type of gambling activity will increase their chance of winning prize money. Other miscellaneous stakeholders may include the government(for changing or implementing policies and regulations) and businesses in tourism industries as gambling is heavily affected by tourist activities.

2 Data Story

ACTValues = c(207.82, 22.35, 0)
NSWValues = c(7802.987, 945.547, 162.866)
NTValues = c(243.797, 651.899, 285.7)
QLDValues = c(3383.943, 339.189, 21.664)
SAValues = c(1013.625, 1.341, 9.68)
TASValues = c(277.3585, 30.066, 3.152)
VICValues = c(4951.259, 542.735, 259.746)
WAValues = c(1223.843, 281.602, 71.786)
Table <- matrix(c(ACTValues, NSWValues, NTValues, QLDValues, SAValues, TASValues, VICValues, WAValues), ncol=3, byrow=TRUE)
Type <-c("Gaming", "Racing","SportsBetting")
colnames(Table) <- Type
rownames(Table) <- c("ACT", "NSW", "NT", "QLD", "SA", "TAS", "VIC", "WA")
table <- as.table(Table)
cols <- c("orangered","peachpuff3", "palegoldenrod", "palegreen","paleturquoise", "peachpuff","pink","plum")

barplot(table, legend = rownames(table), beside = TRUE, col = cols)
title(xlab = "Gambling Category")
title(ylab = "Expenditure (millions of dollars) ")
title(main = "Figure A. Expenditure per State in Each Category of Gambling")

2.1 Evidence-based Conclusion 1

R. Question Statistical Method Evidence-based Conclusion
Does the gambling industry’s profitability vary by location? Factors such as differing policies and regulations, tourism popularity, population size, gambling culture, economy, etc. may vary the success of the gambling industry in different states, even within Australia.
- Bar graph of total expenditure was produced measuring the profitability of the whole gambling industry in each state(Attachment1)
- Mean and sd of each state was produced to look at the specific average values of expenditure(Attachment2)
- Comparative boxplot of expenditure by state was produced to show more robust values(Attachment3)		 The bargraph shows that NSW is the state with the highest total expenditure at almost 9billion AUD. Significantly lower at 6 and 4 billion AUD, is Victoria and Queensland, respectively, and the other states fall below 2billion AUD. The mean and median show similar trends, pointing to NSW being the most profitable location for gambling operators. However, it was also observed, the higher the net income, the higher the spread, which may be due to variance between gambling types.

2.2 Evidence-based Conclusion 2

R. Question Statistical Method Evidence-based Conclusion
Is there a trend in profitability by gambling business type? Now that the gambling businesses have been analysed by location, it is necessary to look at expenditure by the various types within each state as high variance was indicated.
- A pie chart was used to visually compare the proportion of expenditure of each different category of gambling(Attachment 4)
- A bar graph was used to quantitatively compare the expenditure across different states, split into the different categories(Figure A above)		 As can be seen in the figures, the Gaming category of gambling has the greatest expenditure overall across Australia, followed by Racing, and then Sports Betting. Gaming may be the most popular form of gambling as it contains the greatest number of subcategories (9), most of which are found in popular casinos, whilst Sports Betting contains a much smaller amount of subcategories (4 compared to 9), and is not found in as many casinos. Figure A shows that across the individual states, there is a similar trend, suggesting that there is a proportionate relationship between location and expenditure, but points to no association between location and the popularity of different categories of gambling.

2.3 Evidence-based Conclusion 3

R. Question Statistical Method Evidence-based Conclusion
Is there a trend between the availability of gaming facilities and expenditure? As Gaming was determined to be the most profitable category, but also the most abundant, we looked at availability vs. expenditure. Although we naturally assume that a greater number of platforms will lead to a higher gambling expenditure, if this is not the case it would mean that there is a confounding factor.
- A bar graph and scatter plot was produced to show the availability of gambling platforms per state and whether a higher number of platforms was associated with higher expenditure using coefficient ‘r’(Attachment 5)
- Data for Casino expenditure in the Gaming Category across each state is gathered and compared to the availability of gambling facilities for each state (Casino gaming expense vs number of casinos) by plotting a scatter plot and calculating the correlation coefficient ‘r’(Attachment 6)		 The correlation coefficient was calculated to be around 0.50 for both scatter plots. This shows that there is no significant relationship between expenditure and gambling platform availability.
Other possible factors that might affect the relationship between the two variables include: level of tourism activity, population, ethnicity, income level and occupational status. A further investigation is required to determine what other factors may affect this relationship and this process may be complex as multiple factors need to be accounted for.

Summary

A general trend across the states suggests that businesses in the gaming category, especially in New South Wales, is collectively the most profitable faction of the Australian gambling industry. Further research is required to investigate which factors contribute to this trend. An investigation into the relationship between the availability of gambling platforms and expenditure (i.e. number of casinos in each state vs expenditure) did not lead to a conclusive result, also prompting the need for further research for any hidden confounding factors as is needed for observational studies.

Attachments

Attachment 1

Bar graph of Total Expenditure by State

E.ACT = E[S == "ACT"]
E.NSW = E[S == "NSW"]
E.NT = E[S == "NT"]
E.QLD = E[S == "QLD"]
E.SA = E[S == "SA"]
E.TAS = E[S == "TAS"]
E.VIC = E[S == "VIC"]
E.WA = E[S == "WA"]
yis = c("ACT", "NSW", "NT", "QLD", "SA", "TAS", "VIC", "WA")
E.S.S= c(sum(E.ACT), sum(E.NSW), sum(E.NT), sum(E.QLD), sum(E.SA), sum(E.TAS), sum(E.VIC), sum(E.WA))
barplot(E.S.S, xlab = "State", ylab = "Total Expenditure(in millions of AUD)", names.arg = (yis), col = "lightblue", main = "Total Expenditure by Location")

Attachment 2

Mean and SD of Expenditure in each State

c(mean(E, na=TRUE), sd(E, na=TRUE))
## [1] 258.3404 745.1589

All Expenditure

c(mean(E.ACT, na=TRUE), sd(E.ACT, na=TRUE))
## [1] 23.01700 51.54111

Australian Capital Territory

c(mean(E.NSW, na=TRUE), sd(E.NSW, na=TRUE))
## [1]  636.5286 1528.7561

New South Wales

c(mean(E.NT, na=TRUE), sd(E.NT, na=TRUE))
## [1] 118.1396 202.7839

Northern Territory

c(mean(E.QLD, na=TRUE), sd(E.QLD, na=TRUE))
## [1] 416.0884 700.3243

Queensland

c(mean(E.SA, na=TRUE), sd(E.SA, na=TRUE))
## [1]  93.14964 215.91214

South Australia

c(mean(E.TAS, na=TRUE), sd(E.TAS, na=TRUE))
## [1] 25.88137 38.80012

Tasmania

c(mean(E.VIC, na=TRUE), sd(E.VIC, na=TRUE))
## [1] 523.0673 873.1746

Victoria

c(mean(E.WA, na=TRUE), sd(E.WA, na=TRUE))
## [1] 143.3846 256.7691

Western Australia

Attachment 3

Comparative Boxplot of Expenditure by State

summary(Gambling)
##     State           GamblingCategory   GamblingType      
##  Length:88          Length:88          Length:88         
##  Class :character   Class :character   Class :character  
##  Mode  :character   Mode  :character   Mode  :character  
##                                                          
##                                                          
##                                                          
##   Expenditure      
##  Min.   :   0.005  
##  1st Qu.:   1.139  
##  Median :  20.459  
##  Mean   : 258.340  
##  3rd Qu.: 137.418  
##  Max.   :5744.291
boxplot(E ~ S, xlab="State", ylab = "Expenditure(in millions of AUD)", main="Expenditure by State",outline=FALSE)

boxplot(E ~ S, xlab="State", ylab = "Expenditure(in millions of AUD)", main="Expenditure by State (including outliers)")

Attachment 4

Piechart showing Gambling Category Distribution

TotalRacing = 2814.729
TotalGaming = 19104.6325
TotalSportsBetting = 814.594
 
colors = c("darkseagreen1", "lightblue2", "lightpink")
slices <- c(TotalRacing, TotalGaming, TotalSportsBetting)
lbls <- c("Racing", "Gaming", "SportsBetting")
pct <- round(slices/sum(slices)*100)
lbls <- paste(lbls, pct) # add percents to labels
lbls <- paste(lbls,"%",sep="") # ad % to labels
pie(slices,labels = lbls, col=colors, main="Gambling Category Distribution")

Attachment 5

Bargraph showing Gambling Variety per State

barplot(table(E,S), xlab = "State", ylab = "Number of Gambling Types", col = "lightblue", main = "Availability of Gambling Platforms per State")

Scatter Plot showing the Gambling Variety vs Expenditure

plot(Gambling_vs_casinos$`Number of Types`, Gambling_vs_casinos$Expenditure, xlab="Number of Gambling Types", ylab="Expenditure (millions of dollars)", main="Availability of Gambling Platforms vs. Expenditure")

Correlation Coefficient

cor(Gambling_vs_casinos$`Number of Types`, Gambling_vs_casinos$`Total Expenditure`)
## [1] 0.5540635

Attachment 6

Scatter Plot showing the Number of Casinos vs Expenditure in Gaming

plot(Gambling_vs_casinos$`Number of Casinos`, Gambling_vs_casinos$Expenditure, xlab="Number of Casinos", ylab="Expenditure in Gaming (millions of dollars)", main="Number of Casinos vs Expenditure in Gaming")

Correlation Coefficient

cor(Gambling_vs_casinos$`Number of Casinos`, Gambling_vs_casinos$Expenditure)
## [1] 0.4966358

Attachment 7

Dataset Used (before cleaning) *Please note: only a portion of the data is included as the dataset is large. Part of gambling dataset.

Attachment 8: Citations

Australasian Gaming Council 2016, A Guide to Australasia???s Gambling Industries 2015/16, prepared and published by the Australasian Gaming Council (AGC), Melbourne.

Pokies still the ???King??? of gambling in Australia 2018, viewed 28 May 2018, https://www.roymorgan.com/findings/7574-gambling-currency-report-pokies-december-2017-201804270759

Pokie Regulations in Australia 2014, viewed 28 May 2018, http://www.onlinepokies.com/state-regulations.htm

Victorian Responsible Gambling Foundation 2018, Types of Gambling, Gambler???s Help, Victoria, viewed 28 May 2018, https://gamblershelp.com.au/learn-about-gambling/types-gambling/