Blog Archives

Tech’s appeal: Is there a relationship between addiction to video games and slot machines?

Back in 1987, I began my PhD on slot machine addiction, and one thing that I began to notice as I spent the first few hours of (100s of hours) doing observational research in amusement arcades that there were many similarities between arcade slot machines and arcade video game machines. It wasn’t until 1991 that I finally did a comparative analysis of slot machine gambling and video game playing and published my observations in the Journal of Adolescence. In the intervening years I have published many papers examining the commonalities and similarities between these two behaviours and it wouldn’t surprise me if I am still writing about these issue in many years to come.

My initial insights into the existence of video game addiction arose out of the research I had been doing on slot machine addiction. Both slot machines and video game machines may be considered under the generic label of “amusement machines”. The main difference between the playing of video games and the playing of slot machines are that arcade video games are typically played to accumulate as many points as possible whereas slot machines are played (i.e., gambled upon) to accumulate money. In my 1991 paper, I (somewhat paradoxically) claimed that playing an arcade video game could be considered as a non-financial form of gambling.

Both types of machine require insertion of a coin to play, although the playing time on a slot machine is usually much less than on a video game machine if starting with the same amount of money. This is because on video games the outcome is almost solely due to skill, whereas on slot machines the outcome is much more likely to be a product of chance. However, the general playing philosophy of both slot machine players and video game players is to stay on the machine for as long as possible using the least amount of money. I have also argued that regular slot machine players play with money rather than for it, and that winning money is a means to an end (i.e., to stay on the machine as long as possible). This is exactly what arcade video game players do too.

Besides the generic labeling, their geographical juxtaposition, and the philosophy for playing, it could be argued that on both a psychological and behavioural level, slot machine gambling and video game playing share many similarities (e.g., similar demographic differences such as age and gender breakdown, similar reinforcement schedules, similar potential for “near miss” opportunities, similar structural characteristics involving the use of light and sound effects, similarities in skill perception, similarities in the effects of excessive play, etc.). The most probable reason the two forms have rarely been seen as conceptually similar is because video game playing does not involve the winning of money (or something of financial value) and therefore cannot be classed as a form of gambling.

However, the next generation of slot machines is starting to use video game graphics and technology. While many of these relate to traditional gambling games (e.g., roulette, poker, blackjack, etc.) there are plans for developing video gambling games in which people would win money based on their game scores. This obviously gives an idea of the direction that slot machines and the gaming industry are heading.

Furthermore, there are a growing number of researchers who suggest that video games share some common ground with slot machines including the potential for dependency. On 1995, Dr Sue Fisher and myself edited a special issue of the Journal of Gambling Studies and wrote a paper examining trends in slot machine gambling. We pointed out that arcade video games and slot machines shared some important structural characteristics, these being:

  • The requirement of response to stimuli that are predictable and governed by the software loop.
  • The requirement of total concentration and hand–eye coordination.
  • Rapid span of play negotiable to some extent by the skill of the player (more marked in video games).
  • The provision of aural and visual rewards for a winning move (e.g., flashing lights, electronic jingles).
  • The provision of an incremental reward for a winning move (points or money) that reinforces “correct” behaviour.
  • Digitally displayed scores of “correct behaviour” (in the form of points or money accumulated).
  • The opportunity for peer group attention and approval through com- petition.

As with excessive slot machine playing, excessive video game playing partly comes about by the partial reinforcement effect. This is a critical psychological ingredient of video game addiction whereby the reinforcement is intermittent – that is, people keep responding in the absence of reinforcement hoping that another reward is just around the corner. Knowledge about the partial reinforcement effect gives the video game designer an edge in designing appealing games. Magnitude of reinforcement is also important. Large rewards lead to fast responding and greater resistance to extinction – in short to more “addiction.” Instant reinforcement is also satisfying.

Video games rely on multiple reinforcements (i.e., what I call the “kitchen sink” approach) in that different features might be differently rewarding to different people. Success on video games comes from a variety of sources and the reinforcement might be intrinsic (e.g., improving a personal high score, beating a friend’s high score, putting a name on the “hall of fame,” mastering the machine) or extrinsic (e.g., peer admiration). As early as the 1980s, Dr. Thomas Malone reported that video game engagement is positively correlated to (i) a presence or absence of goals, (ii) the availability of automatic computer scores, (iii) the presence of audio effects, (iv) the random quality of the games, and (v) the degree to which rapid reaction times enhance game scores.

In 2007, Dr Jonathan Parke (Salford University, UK) and I developed a new taxonomy of structural characteristics related to gambling, listing all the known structural characteristics that have been shown to influence gambling behaviour in some way. All the 60+ structural characteristics were grouped into one of six types of characteristic:

  • Speed and frequency characteristics: Factors relating to the frequency, duration and expediency of the game or reward.
  • Playability characteristics: Factors that make gambling fun, interaction and/or engaging.
  • Payment characteristics: Factors that relate to how one pays to gamble
  • Reward characteristics: Factors relating to how one receives financial rewards or winnings.
  • Educational characteristics: Factors that educate, protect, or provide information to players.
  • Ambient characteristics: Factors that may influence the immediate situation of the game or may contribute to other factors already mentioned (e.g., the use of colour and sound).

Using this typology, Dr Parke and I argued that future research and policy initiatives may be to focus on regulating structural factors relating to payment (spending) and player awareness/education and focus less on structural factors relating to playability (which may also include reward, ambient, and speed characteristics). In this way, slot machines can continue to be fun, exciting, and play inducing, but with the eventual aim of minimizing harm.

It wasn’t until 2010 that I – along with Dr Daniel King and Dr Paul Delfabbro (both at the University of Adelaide, Australia) – developed a separate taxonomy of structural characteristics related to video game playing (published in the International Journal of Mental health and Addiction). We used some earlier empirical work that I had done with Dr Richard Wood (GamRes Ltd, Canada) back in 2004 and published in the journal CyberPsychology and Behavior. We devised a list of structural features by (a) playing a variety of video games, (b) examining and comparing known gambling structural characteristics, (c) discussing these features with players of video games, and (d) examining relevant research in the area of video game design. Our framework included the following characteristics:

  • Sound, including sound effects, speaking characters and background music.
  • Graphics, including high-quality realistic or cartoon-style graphics and full motion video (FMV).
  • Background and setting, including whether the game is based on a story, film, or television program, and the use of realistic or fantasy settings.
  • Duration of game, referring to how long the game usually takes to complete.
  • Rate of play, referring to how quickly the player “absorbs” or “gets into” the game. & Advancement rate, referring to how quickly the game play advances.
  • Use of humour in the game.
  • Control options, referring to what the player can control in the game (including sound, graphics, and skill settings, choice of control methods, and physical feedback). &
  • Game dynamics, including exploring new areas, elements of surprise, fulfilling a quest, skill development, AI interactions, collecting things, avoiding things, surviving against the odds, shooting, different ending options, different modes of transport, solving puzzles, beating times, cheats/Easter eggs, solving time limited problems, building environments, mapping, and linear/non-linear game format.
  • Winning and losing features, referring to the potential to gain or lose points, finding bonuses, having to start level again, and ability to save regularly.
  • Character development, referring to character development over time and character customization options.
  • Brand assurance, referring to brand loyalty and/or celebrity endorsement.
  • Multiplayer features, referring to various multi-player options, communication methods, building alliances, and beating other players.

Using this paper, and the gambling structural characteristics taxonomy, we developed our new video game structural characteristics taxonomy comprising five types of feature. These were: (a)

  • Social features (i.e., social aspects of video game playing)
  • Manipulation and control features (i.e., the role of user input in influencing in-game outcomes)
  • Narrative and identity features (e.g., the role of character creation and interactive storytelling)
  • Reward and punishment features (i.e., the ways in which players win and lose in video games)
  • Presentation features (e.g., the visual and auditory presentation of video games).

Since developing the taxonomy, we have started to test it out empirically. Dr. King, Dr. Delfabbro and myself recently published a study investigating our structural characteristic taxonomy among 421 video game players (aged between 14 and 57 years). Our results showed that the reward and punishment features, such as earning points, finding rare game items, and fast loading times, were rated among the most enjoyable and important aspects of video game playing. There was some evidence that certain structural characteristics were stronger predictors of problematic involvement in video games than factors such as gender, age, and time spent playing. This included the use of adult content in the game, earning points, getting 100% in the game, and mastering the game. Our latest research supports the notion that some structural characteristics in video games may play a significant role in influencing problem video game playing behaviour.

Dr Mark Griffiths, Professor of Gambling Studies, International Gaming Research Unit, Nottingham Trent University, Nottingham, UK

Further reading

Fisher, S.E., & Griffiths, M.D. (1995). Current trends in slot machine gambling: Research and policy issues. Journal of Gambling Studies, 11, 239-247.

Griffiths, M.D. (1991). The observational analysis of adolescent gambling in UK amusement arcades. Journal of Community and Applied Social Psychology, 1, 309-320.

Griffiths, M.D. (1991). Amusement machine playing in childhood and adolescence: A comparative analysis of video games and fruit machines. Journal of Adolescence, 14, 53-73.

Griffiths, M.D. (1995). Adolescent gambling. London: Routledge.

Griffiths, M.D. (2002). Gambling and Gaming Addictions in Adolescence. Leicester: British Psychological Society/Blackwells.

Griffiths, M.D. (2005). The relationship between gambling and videogame playing: A response to Johansson and Gotestam. Psychological Reports, 96, 644-646.

Parke, J. & Griffiths, M.D. (2007). The role of structural characteristics in gambling.  In G. Smith, D. Hodgins & R. Williams (Eds.), Research and Measurement Issues in Gambling Studies. pp.211-243. New York: Elsevier.

Griffiths, M.D. (2011).  A typology of UK slot machine gamblers: A longitudinal observational and interview study. International Journal of Mental Health and Addiction, 9, 606-626.

King, D.L., Delfabbro, P.H., Derevensky, J. & Griffiths, M.D. (2012). The classification of video games with gambling themes and content: An Australian perspective. International Gambling Studies, in press.

King, D.L., Delfabbro, P.H. & Griffiths, M.D. (2010). Video game structural characteristics: A new psychological taxonomy. International Journal of Mental Health and Addiction, 8, 90-106.

King, D.L., Delfabbro, P.H. & Griffiths, M.D. (2010). The convergence of gambling and digital media: Implications for gambling in young people. Journal of Gambling Studies, 26, 175-187.

King, D.L., Delfabbro, P.H. & Griffiths, M.D. (2011). The role of structural characteristics in problematic video game play: An empirical study. International Journal of Mental Health and Addiction, 9, 320-333.

Malone, T.W. (1981). Toward a theory of intrinsically motivating instruction. Cognitive Science, 4, 333–369.

What is the relationship between accessibility and problem gambling?

Environmental factors additional to gambling exposure are known to have an impact on problem gambling. Some are part of, or closely associated with, the physical and social contexts in which gambling occurs and play a role in increasing or decreasing exposure. Others, while more peripheral, include a number of major risk factors for problem gambling. A number of researchers in the gambling studies field have noted that empirical investigation of relationships between proposed risk factors and outcomes requires accurate and reliable measurement coupled with methodologically robust studies in which exposure levels are varied while other factors that may affect outcomes are held constant or controlled for statistically. If this is not achieved, the findings and conclusions may be invalid and/or misleading.

With most drug-based addictions, different parameters of exposure are typically examined including dose, potency and duration. In the gambling situation, it is much more difficult to quantify social and behavioural exposures. Furthermore, practical and ethical considerations place constraints on experimental investigation. Gambling research is at a relatively early stage of development and it is only recently that public health approaches have been incorporated. In the future, it is likely that more complex measures of gambling exposure will be used. This could include the availability of, and expenditure on, different forms of gambling, the dispersal of and degree of accessibility to these forms, the time they have been available and extent to which harm minimisation strategies have been prescribed and implemented.

Back in 1999, the Australian Productivity Commission (APC) developed a multidimensional framework to assess exposure. It highlighted nine specific dimensions comprising: (i) number of opportunities to gamble, (ii) number of venues, (iii) location of venues, (iv) opportunities to gamble per venue, (v) opening hours, (vi) conditions of entry, (vii) ease of use of gambling form, (viii) initial outlay required, and (ix) social accessibility. Using these criteria the APC conducted several analyses to examine the relationships between accessibility and gambling using state-level electronic gaming machine (EGM) density and expenditure data as well as data drawn from a national Australian gambling survey of gambling prevalence. The results suggested that high levels of problem gambling with gambling machines was correlated to their density relative to the population. In one analysis, the pathological gambling prevalence rate for different Australian States was plotted against the number of gaming machines per 1000 adults in each State. In another analysis, the number of gaming machines per 1000 adults was plotted against the estimated amount spent per capita on gaming machines. Both analyses showed positive relationships suggesting that (at a State level) a greater density of gaming machines per capita was associated with both higher per capita expenditure and higher problem gambling prevalence rates.

However, it should be noted that although several other studies have shown that a higher density of gaming machines in the population correlated to higher rates of problem gambling, this does not, in itself, show that the number of machines in a specific venue has any impact on levels of problem gambling. The number of machines in these studies was related to a large number of venues, and consequently the number of gaming machines in this context does not tell us much about the impact of the number of gambling opportunities in one or a few centralized venues. Furthermore, it could perhaps be speculated that far fewer games in a venue could conceivably encourage a problem gambler to stay on one particular machine for fear of having to wait for another machine to become vacant.

A more complex quantitative procedure was proposed by a group of researchers affiliated to Harvard University led by Professor Howard Shaffer in the US. They generated a ‘standardised exposure gradient’ that assessed gambling exposures within a particular region. This index includes the: (i) dose (i.e., number of gaming venues and people working in the gambling industry), (ii) potency (i.e., the number of different major gambling modalities), and (iii) duration (i.e., the time casinos have been legalized). Although limited, the accuracy could be enhanced by the integration of further information, (e.g., the extent of illegal gambling, access to gambling in adjoining jurisdictions, gaming venue attendance, and advertising). Whether or not exposure indexed by these types of measure has an impact is strongly influenced by the form of gambling involved.

In a review of situational factors that affect gambling behaviour, Professor Max Abbott of Auckland University in New Zealand, concluded that although increased availability of and exposure to gambling activities have contributed to increases in problem gambling, it was highly probable that other situational factors including venue characteristics, social context, access to cash or credit, availability of alcohol, and industry marketing and advertising also have an influence.

Dr Rachel Volberg, Director of Gemini Research in the US, also reached a similar conclusion suggesting there is a correlation between increased availability of gambling opportunities and problem gambling. However, she then reported that in a number of replication studies that problem gambling rates had stabilized or decreased. Looking at these jurisdictions in more detail, she reported that all of them had introduced comprehensive services for problem gamblers including public awareness campaigns, helplines, and professional counselling programmes. She concluded that the relationship between increased opportunities to gamble and problem gambling may be moderated by the availability of helping agencies/services for problem gamblers. In areas of the US (like Montana and North Dakota) that saw an increase in problem gambling following the introduction of casinos, no public awareness campaigns or services for problem gamblers were introduced. Consequently, it appears that the increased availability of gambling opportunities do not necessarily equate to increased levels of problem gambling.

Professor Peter Collins of Salford University (UK) also reviewed this evidence and concluded that if a jurisdiction introduces new forms of gambling and does nothing else, it will most likely see an increase in problem gambling. However, if the jurisdiction combines the introduction of new forms of gambling with appropriate prevention and treatment services, it is likely to decrease numbers of problem gamblers. Collins noted that in the national South African gambling prevalence study that the country witnessed a decline in problem gambling over a two-year period following the introduction of the National Responsible Gambling Program.

Overall, evidence suggests that gambling availability has a positive, but complex, relationship to the prevalence of problem gambling. The relationship is not linear and there are many other factors that determine problem gambling.

Dr Mark Griffiths, Professor of Gambling Studies, International Gaming Research Unit, Nottingham Trent University, Nottingham, UK

Further reading

Abbott, M.W.  (2007). Situational factors that affect gambling behavior. In G. Smith, D. Hodgins & R. Williams (Eds.), Research and Measurement Issues in Gambling Studies. pp.251-278. New York: Elsevier.

Collins, P.  (2007). Gambling and governance. In G. Smith, D. Hodgins & R. Williams (Eds.), Research and Measurement Issues in Gambling Studies. pp.617-639. New York: Elsevier.

Delfabbro, P.H. (2002). The Distribution of Electronic Gaming Machines (EGMs) and Gambling-related Harm in Metropolitan Adelaide. Report Commissioned by the Independent Gambling Authority of South Australia.

Griffiths, M.D. & Parke, J. (2003). The environmental psychology of gambling. In G. Reith (Ed.), Gambling: Who wins? Who Loses? pp. 277-292. New York: Prometheus Books.

Shaffer, HJ., LaPlante, DA., LaBrie, RA., Kidman, RC., Donato, AN., Stanton, MV. (2004). Toward a syndrome model of addiction: Multiple expressions, common etiology. Harvard Review of Psychiatry, 12, 367-374.

Productivity Commission. (1999). Australia’s gambling industries, Report No. 10. Canberra:  AusInfo.  Available at

Volberg, R.A. (2004). Fifteen years of problem gambling prevalence research. What do we know? Journal of Gambling Issues, 10.