This study used PLS-SEM to analyse data (n = 613) collected via an online survey, in order to investigate potential relationships between gambling related cognitive fallacies, or miscognitions, and both participation in gambling and problematic gambling behaviour in the target population. Statistically significant relationships were found to exist in both cases, an unexpected finding was the negative correlation between Interpretive Bias and both dependent variables. One potential explanation for this may be the sense of personal agency and control engendered through regular video game playing. This study forms one of the first to apply existing measures for gambling related cognitions and apply it to a newly emergent population of video game gamblers, and it marks the start of attempts to understand whether cognitive frameworks which underpin gambling apply to emergent forms in the same way as established forms. This study demonstrates that the Gambling Related Cognitions Scale is an applicable measure in the sample population, while highlighting areas for further investigation. As such, it suggests that the context of video games may affect the way in which gambling is perceived, an issue of significance as the gambling industry attempts to woo younger consumers with more skill-based activities inspired by video games. Under the growing influence of new media gambling has not only been brought into the realm of web-based electronic commerce, but also into many other areas of online life, such as video games, social media networks, and electronic sports (King, Delfabbro, Kaptsis, & Zwaans, 2014; LopezGonzalez & Griffiths, 2016). The coalescence of video gaming and gambling is just one example of the wider trend of digital convergence (Jenkins, 2006), and has emerged as the popularity of esports has increased worldwide. Esports is popularly conceived of as organised, competitive video gaming (Hamari & Sjöblom, 2017); its rapid growth has been facilitated by the emergence of Internet Protocol Television (IPTV; Scholz, 2011) and online fan communities (Taylor, 2012). The convergence of gambling and gaming has resulted in the emergence of new forms of gambling, with activities connected to video games and esports available in many forms, both familiar and novel. Esports match betting and fantasy esports are virtually identical to established formats of sports betting (Tsai, 2015). However, esports have also afforded more novel manifestations of gambling practices such as Player-vs-Player (PvP) betting, opening loot crates/cases, virtual item lotteries, and the use of virtual items as wagers for casino and card games (Macey & Hamari, 2018).
Recent legal cases
While the prevalence of video game-related gambling (hereafter, VG gambling) has dramatically
increased in recent years, there is a distinct lack of academic research in the area, what work there
is has focussed on Social Network Casino games (SNC) and free play areas offered by dedicated
gambling sites
Whilst VG gambling utilises existing forms, such as sportsbook betting, and contexts, being
predominantly accessed via the internet, there are several ways in which it may potentially differ
from established practices. In regard to forms of VG gambling, there are several which are not
found elsewhere, these include PvP betting and “crash” betting. Furthermore, participation in these
and other activities is often facilitated via the use of virtual items which can only be obtained by
owning and playing video games, these items can take a number of forms, the most common of
which are “skins”
As such, the role of cognitive biases in the newly-emergent population of video game gamblers urgently requires researching. If cognitions related to established forms of gambling are also found to be function in the same way for newly-emergent forms, then established approaches to treatment, for example Cognitive Behavioural Therapy (CBT), may also benefit this new population. On the other hand, if differences exist between cognitive frameworks underpinning established and emergent gambling activities, it is important that we understand how the context of VG gambling may affect the way in which gambling is perceived.
Currently, the younger generations are abandoning established forms of gambling, such as at
casinos
This study, therefore, will use the Gambling Related Cognitions Scale (GRCS; Raylu & Oei, 2004)
to investigate the presence of cognitive biases associated with gambling in a population of regular
video game players who also consume esports. Furthermore, this research aims to understand the
ways in which cognitive biases interact with the consumption of gambling activities related to video
games, and the prevalence of problematic and potentially problematic gambling behaviours by
using the Problematic Gambling Severity Index
RQ1: How do gambling related cognitions relate to the consumption of video game-related gambling activities in a sample of esports spectators? RQ2: How do gambling related cognitions relate to problematic, and potentially problematic, gambling behaviours in a sample of esports spectators?
An involvement model
Established research has shown that, in the context of established gambling activities, miscognitions are associated with increased consumption of gambling (Ciccarelli et al., 2016) and with the development of problematic gambling behaviour in established forms of gambling, whether land-based or online. Therefore, the model employed the five subscales of the GRCS as independent variables, with consumption of video game-related gambling and problematic gambling assessment constituting the dependent variables. Age, average household income, and employment status were used as control variables.
Although the particular forms of VG gambling, and the environments which surround it, may differ from established practices, at heart they retain the same elements of calculated risk, uncertain outcome, and an exchange of “wealth” (Griffiths, 2018). Therefore, it is expected that the association between cognitive biases and consumption of gambling will also hold true for the novel gambling activities related to video games (H1) and with increased rates of problematic gambling in the context of video game-related gambling (H2).
Accordingly, all individual constructs of the GRCS are expected to show positive associations with consumption of VG gambling, in varying degrees (H1a). Of the 5 GRCS constructs, Gamblingrelated Expectations (GE) is expected to demonstrate the strongest individual associations with increased consumption of video game-related gambling activities (H1b) as it is theorised that the anticipation of positive gambling experiences would be associated with increased consumption. Similarly, all individual constructs are expected to show positive associations with PGSI score, in varying degrees (H2a). Finally, Inability to Stop Gambling (IS) is expected to have the strongest individual associations with increased rates of problematic gambling (H2b) as it has been associated with lack of control and, consequently, continued participation (Raylu & Oei, 2004). The GRCS constructs used as independent variables are all part of a previously validated measure, this study replicates the approach of other research which has validated the GRCS in reference to specific populations and established forms of gambling. As such, this study does not examine the potential interactions or mediators with other variables. 2.
Data was gathered via an online survey, distributed via social media and online discussion forums associated with esports. The final data set (n=613) constitutes 25.6% of total responses received. Participants were eligible to take part in the survey if they had played video games in the previous 12 months, and had also gambled or watched esports within the same timeframe. All valid responses were given the option of participating in a prize draw to win one of five gift vouchers worth $50.
The final dataset was skewed heavily both young (27% under 18, 49% 18 to 25) and male (91%).
These characteristics are not unexpected, echoing previous research on both esports enthusiasts
The Gambling Related Cognitions Scale (GRCS) is an established and validated tool, the scale consists of a total of 23 items which reflect five distinct sub-scales: Illusion of Control (IC); Interpretive Bias (IB); Gambling-related Expectancies (GE); Inability to Stop Gambling (IS); and, Predictive Control (PC). Responses to each item were collected using a seven-point Likert scale, with options ranging from 1 “strongly disagree” to 7 “strongly agree”. Cronbach’s alpha for the present study was .9, with each of the subscales being: PC (α = .726); IB (α = .525); IC (α = .748); GE (α = .813); and, IS (α = .857).
The Problem Gambling Severity Index (PGSI) is a widely used means of assessing problematic,
and potentially problematic, behaviours related to gambling. The PGSI has been designed to be
applicable to any form of gambling, and in any context, as such it is a useful tool to investigate
newly-emergent forms of gambling for which specific tools have not yet been developed. The
short-form version (nine items) has been shown to be as effective a tool as the longer 27-item
version
Analysis was conducted using Partial Least Squares Structural Equation Modelling (PLS-SEM) as
it has been shown to be the most appropriate method when assessing models which include latent,
formative, or reflective constructs. Furthermore, it is an approach which best suits predictive
studies. All results were obtained using SmartPLS 3 for Windows. Table 1 shows the effects
between each of the variables in the model (not including control variables):
The model demonstrates clear support for H1, with four of five GRCS constructs showing
statistically significant relationships with VG Gambling Consumption: Gambling Expectations (β
= .158, p = 0.011); Interpretive Bias (β = .219, p < 0.001); Illusion of Control (β = -.213, p = 0.001);
Inability to Stop Gambling (β = .173, p = 0.007). In total, the model accounted for 21.2% of
variance in the consumption of gambling activities connected to video games. Surprisingly, Illusion
of Control was found to have a negative correlation with video game-related gambling, thereby
rejecting H1a. H1b was also rejected, as the largest correlation for VG Gambling Consumption
was with Interpretive Bias rather than Gambling Expectations (β = .219, and β = .158, respectively).
Although only three of five GRCS constructs were found to have statistically significant
relationships with PGSI Score, the total variance explained was 31.4%, an effect considered large
That cognitive biases were shown to have clear associations with increased participation in gambling connected to video games and esports, and with problematic gambling behaviours, was in line with expectations and echoes previous research in other contexts (Ciccarelli et al., 2016). It seems likely, therefore, that existing approaches which target the mistaken beliefs, such as Cognitive Behavioural Therapy, would be suitable candidates for addressing problematic gambling behaviours of video game gamblers.
The sub-scale GRCS-GE was expected to correlate most strongly with participation in gambling related to video games, however, GRCS-IB was found to have the strongest, positive association with consumption (β = .225), rather than GRCS-GE (β = 0.144). It may be that by re-framing outcomes of gambling, the cognitive biases which constitute Interpretive Bias serve to effect greater influence on future consumption of gambling activities than Gambling-related Expectancies, a group of miscognitions characterised by desire. As such, it may be that the ways in which participants process and rationalise events would prove to be a productive area for further study.
A particularly surprising finding was that one of the subscales (GRCS-IC) had a statistically significant, negative correlation with both increased consumption of video game-related gambling and PGSI score. The Illusion of Control refers to a set of miscognitions in which good fortune is thought to be secured via some kind of ritualistic behaviour or appeal to higher powers. Examples of such behaviour are a belief in prayer, the collection of “lucky” objects and so on. It may be interpreted as suggesting that this particular set of miscognitions actually serves to prevent gambling. However, another perspective is that, as the data set consisted of regular video game players, this particular population possessed more powerful feelings of agency and control. As a result, they rejected the passivity implicit in the beliefs encapsulated in GRCS-IC, further research is required in order to ascertain which of the possible interpretations, if any, are correct. It may even be that GRCS-IC would require re-framing in order to be relevant to this particular population of gamblers.
Indeed, the fact that only three of the five GRCS constructs were found to have statistically
significant relationships with PGSI score in this sample is at odds with previous research which
has found the constructs to be statistically significant predictors of problematic gambling in
established gambling contexts
Data was collected via a publicly available link, and as such the standard criticisms related to
selfselected samples apply. However, the specific characteristics of the target population meant that
the process of recruiting a suitable number of participants through established methods of
probability sampling, such as calling telephone numbers from a registered directory, would be
unfeasible. A further point to consider is that the anonymity of online surveys facilitates potentially
sensitive topics, including gambling and addiction, to be addressed while at the same time reducing
the potential for responses to be guided by the desire for social acceptance (Griffiths, 2010).
Finally, online surveys have been found to produce data as consistent and valuable as that produced
by practices such as using university students, or paying participants
The results of this work suggest that miscognitions related to gambling function in the same manner for the newly emergent behaviours of esports and video game-related gambling as they do for established forms of gambling. Clear evidence was found of positive correlations between miscognitions and both consumption of gambling, and of problematic gambling. With this in mind, approaches such as Cognitive Behavioural Therapy are likely to be as effective methods for addressing problematic behaviour in video game-related gambling as they are for traditional types of gambling. However, the results also show that the constructs that make up the GRCS are not as universally applicable in respect to emergent gambling activities as for established activities, with only three of five demonstrating statistically significant relationships.
This work has also highlighted several areas which would benefit from further investigation, for example the roles played by specific miscognitions and their effects on problematic behaviour in emergent forms of gambling based on video games. Furthermore, this work identified a potential issue with the GRCS in regard to the population of video game gamblers; further work is required in order to assess whether the subscale Illusion of Control is applicable to this particular group. The discrepancies between established and emergent gambling activities are likely to become more meaningful as casino operators, and the gambling industry as a whole, seek to attract younger generations by offering new forms of skill-based gambling derived from, and inspired by, video games. Gainsbury, S. M., Abarbanel, B., & Blaszczynski, A. (2017). Game on: comparison of demographic profiles, consumption behaviors, and gambling site selection criteria of esports and sports bettors. Gaming Law Review, 21(8), 575-587.
Gainsbury, S. M., Russell, A., Hing, N., Wood, R., Lubman, D. I., & Blaszczynski, A. (2014). The prevalence and determinants of problem gambling in Australia: Assessing the impact of interactive gambling and new technologies. Psychology of Addictive Behaviors, 28(3), 769.
Gainsbury, S. M., Russell, A., Wood, R., Hing, N., & Blaszczynski, A. (2015). How risky is Internet gambling? A comparison of subgroups of Internet gamblers based on problem gambling status. New media & society, 17(6), 861-879.
Griffiths, M. D. (2010). The use of online methodologies in data collection for gambling and gaming addictions. International Journal of Mental Health and Addiction, 8(1), 8-20.
Griffiths, M. D. (2018). Is the buying of loot boxes in video games a form of gambling or gaming? Gaming Law Review.
Hamari, J., & Sjöblom, M. (2017). What is eSports and why do people watch it? Internet research, 27(2), 211-232.
Holden, J. T., & Ehrlich, S. C. (2017). esports, skins betting, and wire fraud vulnerability. Gaming Law Review, 21(8), 566-574.