Paper Session 6: Broader Reflections CHI PLAY'19, October 22–25, 2019, Barcelona, Spain

Who Purchases and Why? Explaining Motivations for

In-game Purchasing in the Online Survival Game Fortnite
Jie Cai Donghee Yvette Wohn Guo Freeman
New Jersey Institute of New Jersey Institute of Clemson University
Technology Technology Clemson, USA
Newark, USA Newark, USA [email protected]
[email protected] [email protected]

ABSTRACT reciprocation from other players (e.g., when one needs an item
As a popular and free-to-play multiplayer online survival game, or favor from another player to proceed) [15, 19]. Some games
Fortnite not only offers novel game mechanics but also allows also provide purely cosmetic items for purchasing, such as
players to purchase special in-game items using real money. changing the clothes that one’s avatar wears [24].
Based on an online survey of 215 Fortnite players, this paper
Though the free-to-play business model heavily relies on sales
investigates in-game behaviors that can be used to explain who
of virtual goods [2], it is obvious that not all players spend
is more likely to spend money in Fortnite. Specifically, play-
real-life money in games. In addition, little is known about the
ers’ motivations to spend real-life money were categorized;
decision-making process of in-game purchasing– what are the
how the amount of spending was correlated with different
differences between spenders and non-spenders? And among
motivations and behavioral factors was also analyzed. In con-
those who do spend money, why do some spend more than
trast to prior virtual goods literature that associates in-game
others?
purchasing with social and task motivations, we found that
motivations to make in-game purchases in Fortnite were less In this paper, we use Fortnite as a case to explore the above-
about assimilating with others and more about looking visu- mentioned concerns. We choose Fortnite for two reasons.
ally unique from others. We conclude by discussing how the First, it is considered one of the most successful games to
survival genre of battle royale and other game affordances may generate revenue from selling in-game items [16]. Second,
explain these findings. in contrast to many games’ success that focus on instrumen-
tal in-game items, Fortnite’s success builds on only offering
Author Keywords cosmetic and aesthetic in-game items (e.g., changing the ap-
Fortnite; in-game purchasing; social pressure; individuality; pearance/clothing of one’s avatar) that do not affect the game
self-expression; free-to-play; virtual goods. mechanics. In particular, we focus on three research questions:
• RQ1: What factors are associated with the likelihood of
CCS Concepts spending money in Fortnite?
•Human-centered computing → User studies; Empirical
studies in HCI; • RQ2: What are the reasons for players to buy in-game
items in Fortnite?
INTRODUCTION
• RQ3: How do these factors and reasons influence the actual
The business model of publishing a game for free and then amount of money players spend in the game?
making revenue from selling in-game items for real-life money
(known as free-to-play) has been quite prevalent in Asian Our answers to these questions not only reveal the underlying
massively multiplayer role-playing online games (MMORPG) psychology of in-game purchasing behaviors so as to design
since the early 2000s. In the past decade, it has also gained more desirable games, but also inform the future model of
popularity in Western games, starting with "casual" games or video game publishing.
social network games [19, 23, 24]. In many free-to-play games,
in-game items are often instrumental and can directly benefit BACKGROUND
players’ gameplay– to enhance a certain skill, receive extra An Overview: Fortnite
time to play the game, or unlock features that require social Fortnite is a third-person multiplayer online survival game
(known as the battle royale game genre) released by Epic
Permission to make digital or hard copies of all or part of this work for personal or Games in July 2017. In this game, up to 100 players can
classroom use is granted without fee provided that copies are not made or distributed
for profit or commercial advantage and that copies bear this notice and the full citation conduct player-versus-player battle. Players have the options
on the first page. Copyrights for components of this work owned by others than ACM to play alone, as a duo, or a small team (usually three or
must be honored. Abstracting with credit is permitted. To copy otherwise, or republish,
to post on servers or to redistribute to lists, requires prior specific permission and/or a
four players). Once the game starts, they must compete with
fee. Request permissions from [email protected]. others to collect weapons, items, and resources for survival
CHI PLAY ’19, October 22–25, 2019, Barcelona, Spain. and combat with one another. One of the core game mechanics
Copyright © 2019 Association of Computing Machinery.
ACM ISBN 978-1-4503-6688-5/19/10 ...$15.00. is to collect resources to build platforms and walls to either get
http://dx.doi.org/10.1145/3311350.3347196

391
Paper Session 6: Broader Reflections CHI PLAY'19, October 22–25, 2019, Barcelona, Spain

a vantage point or hide from the enemy [9]. The last player, Belk’s 2013 work has especially framed digital consumption
duo, or team alive would be the winner [13]. as self-expression [1]. He pointed out that avatars are now act-
ing as the extended self, which can affect our offline behavior
Like many popular multiplayer online games, Fortnite also and sense of self; he also highlighted that traces of our digi-
uses the free-to-play business model. It is free to download tal consumption are acting as cues to personal and aggregate
and install the game. After creating an account, a player will sense of past, leading to the more actively managed and co-
be randomly assigned with an avatar skin during each game constructed self [1]. In addition, a recent study of League of
session. The avatar skin includes a standard backpack, emote Legends suggests that identification with the virtual group and
(a movement that the avatar makes in the game that allows self-presentation affects players’ purchase behavior [8]. Habit
users to express themselves), pickaxe (allowing users to mine plays an essential role in spending real-life money in games
and break materials or cause damage against enemies), and
as well. For example, studies of Second Life have shown that
glider (allowing users to fly while either jumping off the battle
habit positively affects participants’ actual purchase behavior
bus or using the launch pad) at the beginning of the game.
in that virtual world [4].
Fortnite has more than 150 million players [13] and hit 318
million in revenue in May 2018 [14]. Most of its revenue As previous studies have shown, players’ purchase behaviors
is from players’ small purchases of its in-game currency– V- may greatly vary across different gaming contexts, genres, and
Bucks, which can be used to buy virtual goods such as emotes, platforms. Representing a new game genre, Fortnite not only
avatar skins, and accessories in the game. 1000 V-Bucks costs offers novel game mechanics (PvP multiplayer survival battle)
USD$9.99. Therefore, the amount of V-Bucks that players but also focuses on cosmetic only in-game items. Therefore,
own reflects the amount of real money they spend in the game. who purchases and why in this game may be different from
other games that have already been studied, such as Second
It should also be noted that while in-game purchasing is cer- Life, Habbo Hotel, and World of Warcraft (e.g., see the meta-
tainly not new, Fortnite is unique in that its purchasable items analysis by [7]), leading to the importance and necessity to
are purely cosmetic and do not give players any tactical advan- study in-game purchasing in this unique context.
tage [20], as we mentioned at the beginning of this paper. If
anything, a flamboyant costume may make the player easier METHODOLOGY
to detect from afar. In this sense, this game does not provide a
disadvantage to those who are not spending real money, which Participant Recruitment
is different from many mainstream multiplayer online games. We collected data using a 15-minute anonymous online survey
with both closed and open-ended questions. Examples of
Factors Associated with Actual Purchase Behavior
open-ended questions include "Why do you give money to a
Fortnite streamer?” "What’s the purpose of subscribing to
In-game items in Fortnite can be conceptualized as a type
a streamer?” and so forth. However, since this study was
of virtual good. Virtual goods refer to the subset of the vir-
part of a larger project on Fortnite players’ different patterns
tual asset that can be mass-produced, bought, and sold live
of playing, spending in-game, viewing Fortnite live streams,
conventional consumer products, including items, characters,
and spending money on live streams, results of open-ended
and currencies of massively multiplayer games. Very often,
questions were not reported in this paper.
three attributes of such virtual goods would drive players to
purchase: functional, hedonic, and social [11]. To recruit participants, we used multiple online and offline plat-
forms, including Mechanical Turk, Facebook groups, Reddit,
A comprehensive literature review about why people buy vir-
Twitter, and collegiate eSports clubs through direct contact.
tual goods reveals that research focuses more on purchase
Participants were first asked whether they had played Fort-
intention than on purchase behavior [6]. In addition, among
nite in the past month and what their Fortnite IDs were (to
the independent variables that can be applied to predict ac-
make sure they indeed were Fortnite players). Only English-
tual purchase behavior, both social and individual factors play
speaking adults (18 or above) who have played Fortnite in the
significant roles. For example, social factors such as the ex-
past month were allowed to take the survey. As a result, we
change of virtual goods (given and received) in online social
collected a total of 247 responses. After eliminating people
games increases the likelihood of spending real-life money;
who did not write legible responses to open-ended questions,
and giving virtual goods is the strongest factor associated with
215 responses were used for analysis.
actual spending [24]. Similarly, Yee’s [25] research on player
motivations in multiplayer online games has identified ten
Survey Measures
player perceptions of spending real-life money on virtual as-
All the survey items of measures were measured with a 5-point
sets, including three social factors (socializing, relationship,
Likert Scale rating from 1 (Strongly disagree) to 5 (Strongly
and teamwork) [10]. Other research also shows that players
agree). For the question regarding reasons to purchase in-
with intensive interaction in the game and longer total play-
game items in Fortnite ("why do you buy in-game items in
time are more likely to make the in-game purchase in the
future [18], and that social interaction is positively associated Fortnite?"), all potential answers (e.g., "I want to look cool" "I
with how much money players spend on in-game content [5]. want to be funny" "To coordinate my looks with teammates")
were developed based on pilot interviews with Fortnite players.
Yet less research focuses on the individual perspective. Re- The question regarding habit ("When I start to play Fortnite, it
search has shown that people buy virtual items to present and is something...") was adopted from prior literature with minor
customize a favorite image of themselves [6, 21]. Russell alterations to fit the context of this research and measured by

392
Paper Session 6: Broader Reflections CHI PLAY'19, October 22–25, 2019, Barcelona, Spain

the self-reported habit index [22]. 12 potential answers were have played the game. We also included demographic charac-
included, such as "That belongs to my daily/weekly/monthly teristics such as age and income (Table 2 left). The dependent
routine," "That’s typical ‘me’," "That I do regularly," "I do variable was if the person had spent any money (answer Yes,
frequently," "That I do consistently." coded as 1) or no money (answer No, coded as 0).
We also asked questions about play frequency to better under- The Omnibus Test showed that the binomial regression ( χ 2
stand players’ gaming behavior. Example questions include = 31.22, df = 5, p < .001) was a significant model; moreover,
"How long have you been playing Fortnite?" Potential an- the Hosmer and Lemeshow Goodness-of-Fit Test (χ 2 = 4.01,
swers were organized as "Under one month" "1-3 months" df = 8, p = .86) indicated that the model fitted the data well.
"4-6 months" "7-9 months" and "more than 9 months." An- It predicted overall 76.1% of in-game purchasing correctly
other example was "How many hours do you spend playing (Nagelkerke R2 = .21). As can be seen in Table 1, people who
Fortnite per week?" and the participants had to type in a spe- have been playing longer were more likely to spend money (β
cific number. = .36, odds ratio = 1.43), and younger people (β = -.09, odds
ratio = .91) were more likely to spend money in the game.
Two dependent variables related to money spending were mea-
sured. One was the likelihood of purchasing in-game items
Factor Analysis of Why Purchase in Game
("Do you buy in-game items in Fortnite?"). The answers were
treated as the categorical variable ("No" or "Yes"). The other To explore RQ2, we included the survey question "why do
was the amount of purchased V-bucks, an indicator of real- you buy in-game items in Fortnite?" and listed 21 possible an-
life money players spent in the game ("How many V-Bucks swers. All possible answers were measured in a 5-point Likert
have you bought in the game?"). To answer this question, scale ranging from strongly disagree to strongly agree. We ap-
participants could enter the actual number of V-Bucks they plied a principal component analysis with the Varimax rotation
purchased. method and the eigenvalue greater than one. As a result, we
identified two factors with six items (i.e., social pressure and
Demographic and financial information such as age, race, and individuality; see Table 1) and removed other non-significant
monthly income were also collected. Since income was in a items such as "Because it is limited edition," " I want to look
wide range and heavily skewed, we transformed income by attractive," and " The item is fun (to me)." Social pressure
using the log function for further analysis. (M= 2.98, SD= 1.11, α= 0.76) measured players’ perception
towards being influenced by friends, and individuality (M=
RESULTS 3.97, SD= 0.92, α= 0.78) measured participants’ desire to be
unique. These two factors explained 70% of the total variance.
Descriptive Results
Our participants were aged 17 to 70 (M = 28.67, SD = 7.77), Items Loadings
with 74.4% male and 25.6% female. Most participants self
identified as White (70.2%), followed by Asian (18.1%) and To look unique .84 .11
black (12.1%). I want to look different from other people .82 .24
I want to express my individuality .80 .10
Some participants played Fortnite on multi-platforms, but the Because my friends bought it .02 .88
primary platform was PC (47.9%), followed by PlayStation Because it’s popular .15 .83
4 (20.9%), and Xbox One (18.1%). Around 30% of our par- I want to show others I am a serious player .40 .70
ticipants have been playing Fortnite for four to six months, Table 1. Factor analysis of why users buy in-game items.
followed by for one to three months (27%) and more than
nine months (21.4%). Most participants played alone-against
strangers (69.1%) or in a team with friends (58.1%) (check all
Modeling the Actual Purchase Behavior
that apply). Participants played 0 to 35 hours per week (M =
6.52, SD = 6.56, Median = 4) and reported having an average To identify the factors that contributed to the actual money
score of 3.38 of habit, which is in the range between neutral spending on V-bucks (RQ3), we applied a Negative Binomial
and agree on a five-point scale (SD = 0.96, α = 0.91). Logistic Regression. This model is often used when the depen-
dent variable is a count variable with a non-normal distribution.
Among the 215 responses, 60 participants bought in-game In our case, our dependent variable was the actual amount of
items, accounting for 27.9%. 56 of 60 participants spent V- V-bucks that the users purchased, and it was heavily skewed
bucks on outfits, followed by emotes (N=37), gliders (N=26), (Skewness = 3.65).
and harvesting tools (N=25) (check all that apply). The
amount of purchased V-bucks widely ranged from 0 to 100,000 Independent variables included all the behavioral factors that
(M=7,405, SD=17,254, Median=1,000). were in the logistic model and the two motivational factors
identified in the last section (i.e., social pressure and individu-
ality) (Table 2).
Modeling the Likelihood of In-Game Purchasing
To identify the factors that contributed to the likelihood of The Omnibus Test (Likelihood ratio χ 2 = 62.32, df= 9, p< .001)
purchasing (any) in-game items (RQ1), we applied a Bino- showed that the model was significant. In Table 2, individual
mial Logistic Regression. Independent variables included motivation (β = .46) was positively related to the amount
behavioral factors associated with gameplay, such as habit, of purchased V-Bucks, yet the association between social
frequency of play, hours spent playing, and how long they motivation and the actual purchase amount was negative (β =

393
Paper Session 6: Broader Reflections CHI PLAY'19, October 22–25, 2019, Barcelona, Spain

Model Binomial Regression Negative Binomial Regression

Variables Beta Wald Exp (B) Beta Wald

Constant .28 1.19 1.32 8.62*** 63.94

Habit .28 1.66 1.32 -.42* 4.13
Age -.09** 9.89 .91 -.02 .56
Income (log) -.25 1.01 .78 -.25 .89
Hours play .03 1.27 1.03 .06* 4.73
Length play .36* 6.04 1.43 .72*** 16.14
Individuality .46* 5.30
Social Pressure -.72*** 16.57
Model Chi-square [df] 31.22 [5] p<.001 62.32 [7] p<.001
Dependent variables: Buy in-game items No/Yes; actual number of V-Bucks purchasing, *p<.05 ** p<.01 *** p<0.001
Table 2. Modeling likelihood of in-game purchasing (N=197) and the number of V-Bucks purchasing (Included 54) among players

-.72). In addition, the two frequency questions, namely, hours individuality was positively while social pressure was nega-
play (β = .06) and length play (β = .72), were positively related tively related to actual money spending. In this sense, players
to the actual purchase amount. However, habit (β = -.42) was who want to differentiate them from their friends and other
negatively associated with the amount of V-Bucks that players players will spend more money on in-game items.
purchased.
The Digital Consumption (Spending) Motivated by Being
DISCUSSION Unique
Characteristics of In-Game Purchases in Fortnite
In general, our findings are consistent with results reported
One significant finding of our study is the negative association in previous literature that players who buy aesthetic items
between age and the likelihood of purchasing in-game items, intend to improve the avatar appearance rather than the game-
which indicates that younger players are more likely to buy play performance [12] because these adornments highlight the
in-game items. In this sense, features or promotions to target individuality and uniqueness [4].
younger players should be highlighted in designing free-to- Our findings also point to some aspects that may have been
play games. overlooked in other studies, especially regarding digital con-
In addition, while hours of play was not a significant predictor sumption (spending) and self-expression. For example, Belk
of the likelihood of in-game purchasing, the total length of seems to emphasize the social aspect of co-construction of
staying with the game was. This finding suggests that it is the self, explaining the shared digital possessions and aggregate
attachment to the game rather than the time spending in the self [1]. However, we found that digital consumption (spend-
ing) is motivated by being unique/separated from others rather
game that affects players’ willingness to pay. However, hours
than wanting to belong to a community/attach to others. In
of play was still positively related to actual purchase behavior,
this sense, digital consumption is not for aggregating or co-
which indicates that players who spend more time will also
spend more money. constructing the self but for establishing a more distinctive
self.
Somewhat surprising, income was an insignificant predictor
for both models. Therefore, the willingness to pay and how Therefore, due to Fortnite’s unique game mechanics (a survival
game) and social dynamics (different type of collaboration me-
much players would pay are determined not by their financial
chanics and tendency to play with strangers), players’ avatars
capability but by the game itself.
act as not only the extended self but also a more distinctive self
Regarding habit, the negative correlation between habit to than the offline self. Their digital consumption represents how
play and the actual spending behavior indicates that players the individual/aesthetic dimension of the online self some-
who do not play regularly/consistently spend more money. how overweights its social/instrumental dimension. Since
A possible explanation is that players who spend more in contextualization is an important factor of studying the virtual
the game are compulsive shoppers themselves and care more economy [17], we thus believe that our work also points to
about the cosmetic value rather than the instrumental value of the importance of taking the specific online context and social
virtual goods. dynamics into account when discussing digital consumption
as self-expression.
The factor analysis also showed that players bought in-game
items because they wanted to either look unique and stand out In addition, social pressure, different from other social factors
from the groups or to catch up on popular trends with other such as social interaction [5, 24] and social influence [3, 4]
players. It appears that the design of virtual goods should that have been discussed in previous literature, is rarely ex-
take both social and individual attributes into account. Yet in plored in the gaming context, perhaps the first time applying
Fortnite, individual attributes should be more critical because to Fortnite. Interestingly, the negative association between

394
Paper Session 6: Broader Reflections CHI PLAY'19, October 22–25, 2019, Barcelona, Spain

social pressure and actual money spending is contradictory to [5] Juho Hamari, Kati Alha, Simo Järvelä, J. Matias
previous research on social factors. One possible explanation Kivikangas, Jonna Koivisto, and Janne Paavilainen.
is that, different from other games, Fortnite as a survival game 2017. Why do players buy in-game content? An
requires limited collaboration (2 or 4 people as a group) but empirical study on concrete purchase motivations.
more killing since most participants reported that they played Computers in Human Behavior (0 2017). DOI:
with strangers. http://dx.doi.org/10.1016/j.chb.2016.11.045

Limitation and Further Research [6] Juho Hamari and Lauri Keronen. 2016. Why do people
This study has several limitations that point to further research. buy virtual goods? a literature review. Proceedings of
First, since this is a quantitative study, we can identify the neg- the Annual Hawaii International Conference on System
ative relationship between social pressure and money spending. Sciences 2016-March (2016), 1358–1367. DOI:
Yet we are only able to provide a possible explanation based http://dx.doi.org/10.1109/HICSS.2016.171
on the assumption that the team dynamic in Fortnite is differ- [7] Juho Hamari and Lauri Keronen. 2017. Why do people
ent from other games. Further research that involves in-depth buy virtual goods: A meta-analysis. Computers in
qualitative analysis is needed to validate this explanation. Sec- Human Behavior 71 (2017), 59–69. DOI:
ond, more studies can be conducted to investigate the team http://dx.doi.org/10.1016/j.chb.2017.01.042
dynamics in this special game genre (i.e., multiplayer PvP
survival combat) so as to explain the contradictory findings of [8] Bastian Kordyaka and Sidney Hribersek. 2019. Crafting
social factors. Third, this research only focuses one platform Identity in League of Legends - Purchases as a Tool to
(i.e., Fortnite); future cross-platform studies can compare find- Achieve Desired Impressions. In Proceedings of the
ings from this study to those from other popular platforms 52nd Hawaii International Conference on System
such as Second Life or World of Warcraft. Sciences, Vol. 6. 1506–1515.
https://hdl.handle.net/10125/59591
CONCLUSIONS
[9] Sarah LeBoeuf. 2018. What is ’Fortnite’?: A look at the
Using Fortnite as a case, we have identified several factors video game that has become a phenomenon. (2018).
related to the likelihood of in-game purchasing. We also have https://www.nbcnews.com/tech/tech-news/
introduced social pressure (a social factor rarely explored in what-fortnite-look-video-game-has-become-phenomenon-n887706
the in-game context) and individuality (a clear individual fac-
tor) to explain why people purchase in-game items. We found [10] Vili Lehdonvirta. 2005. Real-Money Trade of Virtual
that less social pressure and more individuality would cause Assets: Ten Different User Perceptions. (2005). DOI:
more spending, which highlighted digital consumption as not http://dx.doi.org/10.2139/ssrn.1351772
for aggregating or co-constructing the self but for establishing [11] Vili Lehdonvirta. 2009. Virtual item sales as a revenue
a more distinctive self. We conclude that Fortnite’s unique model: Identifying attributes that drive purchase
game mechanics (a survival game) and social dynamics (differ- decisions. Electronic Commerce Research 9, 1-2 (2009),
ent type of collaboration mechanics and tendency to play with 97–113. DOI:
strangers) are possible explanations for these findings. We also http://dx.doi.org/10.1007/s10660-009-9028-2
point to future research directions including more qualitative
and cross-platform studies to validate such explanations. [12] Rodney Lim and Ee Yaw Seng. 2010. Virtual Goods in
Social Games: An Exploratory Study of Factors that
REFERENCES Drive Purchase of In-Game Items. In The 9th
[1] Russell W Belk. 2013. Extended Self in a Digital World. International Conference on e-Business. 26–32.
Journal of Consumer Research 40, 3 (2013), 477–500. https://pdfs.semanticscholar.org/2760/
DOI:http://dx.doi.org/10.1086/671052 6c8fe3c23c18dac3cf7f7bb61d6a4cd3daaf.pdf
[2] Robert Flunger, Andreas Mladenow, and Christine [13] David Mainenti. 2018. Fortnite: Addictive Berry
Strauss. 2017. The free-to-play business model. In Foraging with Algorithms. Information Science (2018).
Proceedings of the 19th International Conference on https://medium.com/@kylepob61392/fortnite-battle-
Information Integration and Web-based Applications &
Services - iiWAS ’17. ACM Press, New York, New York, [14] Rani Molla. 2018. Fortnite generated a record $318
USA, 373–379. DOI: million in revenue in May - Recode. (2018).
http://dx.doi.org/10.1145/3151759.3151802 https://www.recode.net/2018/6/26/17502072/
fortnite-revenue-game-growth-318-million
[3] Yue Guo and Stuart Barnes. 2007. Why people buy
virtual items in virtual worlds with real money. ACM [15] David B. Nieborg. 2015. Crushing Candy: The
SIGMIS Database 38, 4 (10 2007), 69. DOI: Free-to-Play Game in Its Connective Commodity Form.
http://dx.doi.org/10.1145/1314234.1314247 Social Media and Society 1, 2 (2015), 1–12. DOI:
http://dx.doi.org/10.1177/2056305115621932
[4] Yue Guo and Stuart Barnes. 2011. Purchase behavior in
virtual worlds: An empirical investigation in Second [16] Hamza Shaban. 2019. FortniteâĂŹs free-to-play model
Life. Information and Management 48, 7 (2011), threatens major game makers. (2019).
303–312. DOI:
http://dx.doi.org/10.1016/j.im.2011.07.004

395
Paper Session 6: Broader Reflections CHI PLAY'19, October 22–25, 2019, Barcelona, Spain

[17] Paurav Shukla and Judy Drennan. 2018. Interactive Discourses_of_Microtransactions_Acceptance_and_

effects of individual- and group-level variables on Rejection_in_Mainstream_Video_Games
virtual purchase behavior in online communities.
[21] Heikki Tyni, Olli Sotamaa, and Saara Toivonen. 2011.
Information and Management 55, 5 (2018), 598–607.
Howdy pardner!: On free-to-play, sociability and
DOI:http://dx.doi.org/10.1016/j.im.2018.01.001
rhythm design in FrontierVille. In Proceedings of the
[18] Rafet Sifa, Sankt Augustin, Fabian Hadiji, Julian Runge, 15th International Academic MindTrek Conference:
Anders Drachen, Kristian Kersting, Christian Envisioning Future Media Environments. 22–29.
Bauckhage, Fraunhofer Iais, and Sankt Augustin. 2015. http://dl.acm.org/citation.cfm?id=2181042
Predicting Purchase Decisions in Mobile Free-to-Play
[22] Bas Verplanken and Sheina Orbell. 2003. Reflections on
Games. In Proceedings, The Eleventh AAAI Conference
Past Behavior: A Self-Report Index of Habit Strength.
on Artificial Intelligence and Interactive Digital
(2003). DOI:
Entertainment (AIIDE-15). 79–85.
http://dx.doi.org/10.1111/j.1559-1816.2003.tb01951.x
http://www.aaai.org/ocs/index.php/AIIDE/AIIDE15/
paper/viewFile/11544/11359 [23] D.Y. Wohn and E.-K. Na. 2012. Virtual item purchasing
patterns in a social game: Differences between high and
[19] Jieun Sung, Torger Bjornrud, Yu-hao Lee, and D. Yvette
low spenders. In ACM International Conference
Wohn. 2010. Social network games: Audience traits. In
Proceeding Series. 431–432. DOI:
Proceedings of the 28th of the International Conference
http://dx.doi.org/10.1145/2132176.2132237
Extended Abstracts on Human Factors in Computing
Systems - CHI EA ’10. ACM Press, New York, 3649–54. [24] Donghee Yvette Wohn. 2014. Spending Real Money:
DOI:http://dx.doi.org/10.1145/1753846.1754033 Purchasing Patterns of Virtual Goods in an Online
Social Game. In Proceedings of the SIGCHI Conference
[20] Jan Švelch. 2017. Playing with and against
on Human Factors in Computing Systems. 3359–3368.
Microtrnasctions: The Discourses of Microtransactions’
DOI:http://dx.doi.org/10.1145/2556288.2557074
Acceptance and Rejection in Mainstream Video Games.
In The Evolution and Social Impact of Video Game [25] Nick Yee. 2006. The demographics, motivations, and
Economics, Casey B. Hart (Ed.). Lexinton Books, derived experiences of users of massively multi-user
Maryland, Chapter Seven, 101–120. online graphical environments. In Presence:
https://www.researchgate.net/publication/328981126_ Teleoperators and Virtual Environments, Vol. 15.
Playing_with_and_against_Microtransactions_The_ 309–329. DOI:
http://dx.doi.org/10.1162/pres.15.3.309

396