The Impact of Speed of Play in Gambling on Psychological and Behavioural Factors: A Critical Review

Games with fast speeds of play are frequently associated with problem gambling. For example, it has frequently been observed that problem gamblers seeking intervention or treatment for their disordered gambling often report rapid forms of gambling (such as electronic game machines [EGMs]) as a primary cause of their disordered gambling (e.g., Griffiths 2008; Meyer et al. 2009; Turner and Horbay 2004). In the psychological gambling literature, speed of play is inextricably associated with event frequency, a structural characteristic referring to the number of gambling events within a given time period (and operationalized as the time interval between successive wagers on any given gambling game [Griffiths and Auer 2013]). For example, the event frequency of a bi-weekly lottery is twice a week, whereas the event frequency on an EGM that spins 12 times a minute is five seconds. A fast speed of play has been identified as one of the key features that appeal to gamblers and is therefore more likely to be associated with both higher levels of gambling participation generally, as well as gambling-related harm (Parke and Griffiths 2007). Of concern is evidence suggesting games with fast speeds of play, such as EGMs, are particularly appealing to problem gamblers (Griffiths 2008).

Several theoretical propositions exist that attempt to account for the relationship between high event frequency gambling participation and disordered gambling. For instance, the rapid sequencing of gambling stimuli accompanied with reward (i.e., ‘the constant cycling of player action’; Dow-Schull, 2012) means that that fast, rhythmic, and continuous nature of EGM gambling facilitates an immersive state of lowered conscious awareness for peripheral information. This may give rise to the gambler experiencing a dissociative state, and it has been argued that such psychological states, facilitated by games with fast speeds, are pleasurable to the gambler (Griffiths et al. 2006). During such dissociative experiences, the need for more conscious and deliberate decision-making is limited, providing negative reinforcement to gamble by reducing tension and escaping wider psychological distress that may be experienced in everyday life (Fang and Mowen 2009). However, Norman and Shallice (1986) argue that there are specific situations where the routine activation of behaviour, at the expense of top-down executive control, is maladaptive. Unsurprisingly, among the situations Norman and Shallice (1986) identify include those where potential danger can be experienced, or situations that require planning and decision-making. Given that gambling is a situation requiring the constant updating of goals and adjustment of behaviour, as well as a situation where harm may be experienced, it may be maladaptive for gambling features such as speed of play to facilitate dissociative experiences.

The appeal of games with fast speeds of play, particularly amongst problem gamblers, may also be explained by Gray’s (1970) Reinforcement Sensitivity Theory. The theory postulates that the Behavioural Approach System (BAS) motivates behaviour to seek out reward (Gray 1981, 1991). The subsequent reward, which is exciting and pleasurable to the individual, reinforces the behaviour and consequently leaves individuals highly sensitive to potential rewards and makes extinction of the behaviour difficult. Pickering and Gray (1999) argue that dopaminergic fibres ascending from both the substantia nigra and ventral tegmental areas of the brain, that innervate the basal ganglia, together with motor, sensorimotor, and prefrontal regions, are assumed to drive this system. It has been demonstrated that those with abnormalities in dopaminergic functioning, as well as ventro-medial prefrontal cortex structures, are at risk of developing problem gambling due to abnormalities in the way reward and punishment is processed (Goudriaan et al. 2004). Therefore, it is perhaps unsurprising that gamblers with increased sensitivity to reward will be attracted to games with high event frequencies, as such games are more likely to provide increased levels of reward in a relatively shorter period of time.

Alternatively, sensitivity to punishment or loss is seen as a protective factor in the persistence of risk-taking behaviour (e.g., Gray 1991). Games with high event frequencies also deliver relatively higher rates of loss, and therefore conceptually, one could predict that such factors result in fast games being avoided for gamblers with higher levels of punishment sensitivity. Paradoxically, research demonstrates that this is not the case for gamblers with high levels of sensitivity to reward and punishment. For example, Gaher et al. (2015) argue that the increased sensitivity to punishment results in further gambling to alleviate the negative mood state caused by the loss, which results in loss-chasing behaviours. As a result, reinforcement sensitivity theory is able to predict that those high in either reward sensitivity, and/or punishment sensitivity, would be attracted to and persist on games with fast speeds of play for different reasons.

The rapid and continuous pace of play afforded by gambling games with high event frequencies may potentially interfere with a gamblers’ ability to process new information, update goals, and/or make adjustments in their behaviour to avoid undesirable consequences. Response modulation is a cognitive process whereby the individual disengages attention on the ongoing activity to re-evaluate and adjust behaviour according to the current reinforcement rate of the behaviour in question (e.g., Derevensky et al. 2011). Behavioural perseverance despite negative consequence is a hallmark sign of a wide range of clinical disorders including psychopathy (Newman et al. 1987), borderline personality disorder (Davey 2008), and disordered gambling (Thompson and Corr 2013). Consequently, if a gambler is not afforded the opportunity to pause and reflect between gambling events, it is less likely that they will respond adaptively to punishment (e.g., financial loss). High event frequency games allow less opportunity for such reflection and adaptation of behaviour and are therefore more likely to lead to behaviour symptomatic of problem gambling. In support of this notion, experimental evidence suggests that when problem gamblers are forced to pause for five seconds between events, they do not persist in gambling longer than non-problem gamblers (Corr and Thompson 2014; Thompson and Corr 2013). However, it is unclear whether this effect is due to increased reflection time, or more simply, that the pause made the game less enjoyable. Both factors are not necessarily mutually exclusive.

Whilst these theoretical models have high face validity in explaining why fast speeds of play are associated with disordered gambling, a significant problem remains in that the empirical relationship is largely correlational. The argument can be made that a weak empirical association between fast speeds of play and disordered gambling is potentially harmful to scientific research into this relationship, as it assumes an extensive knowledge-base has already been established. Therefore, one of the goals of the present review is to identify the gaps in the current understanding relating to the impact of high event frequency on gamblers across the entire spectrum of problem gambling behaviour. An additional reason for carrying out the present review paper is to collectively establish what is already known in terms of the psychological and behavioural factors that high event frequency games impact. This is to facilitate the development of gambling harm-minimisation approaches which focus on specific factors that enhance a gamblers’ self-control. As far as the authors are aware, no previous literature review has ever examined speed of play in gambling as the single focus although more general reviews of structural characteristics in gambling have devoted small sections of such overviews to theoretical descriptions of event frequency (e.g., Griffiths 1993; McCormack and Griffiths 2013; Parke and Griffiths 2006, 2007).

Based on the studies reviewed, there appears to be an overall trend from the experimental findings that games with high event frequencies are perceived as more exciting and more enjoyable by gamblers, and which is likely to be one of the core factors accounting for the popularity of EGMs. This is a finding that applies to gamblers across the entire problem gambling continuum. This evidence is supported and complemented by the qualitative data surrounding speed of play, where the reasons gamblers show a preference for such fast games include the instant gratification they provide, and the lack of waiting around between gambling events which appeals to the gamblers’ ‘lack of patience’. However, while both problem and non-problem gamblers rate faster games as more enjoyable when compared to slower game speeds, some studies (e.g., Linnet et al. 2010) demonstrate that enjoyment ratings for fast games are significantly higher amongst problem gamblers. Furthermore, some studies found that problem gamblers also report a significantly higher desire to continue gambling on faster games when compared to the same ratings made by non-problem gamblers, as well as problem gamblers also reporting a greater reduction in tension when playing faster games. Taken together, these findings appear to support previous notions that games with fast speeds are particularly appealing to those displaying signs of disordered gambling (e.g., Griffiths 2008).

In terms of the behavioural impact of speed of play, results demonstrated a varied set of findings. Several studies reported that games with faster speeds of play resulted in more games being played compared to slower games (e.g., Loba et al. 2001; Delfabbro et al. 2005; Ladouceur and Sevigny, 2006), which is perhaps unsurprising given the fact that a higher event frequency affords the gambler the opportunity to make more bets in a given period of time compared to games with slower event frequencies. Several studies also found that problem gamblers reported more difficulty in stopping gambling compared to non-problem gamblers at fast speeds of play (an effect that disappeared when game speed was slowed) or that problem gamblers were significantly more likely to continue gambling until asked to stop at fast speeds compared to non-problem gamblers (an effect that was also found at slower speeds, though to a lesser extent). One study (i.e., Mentzoni et al. 2012) showed that speed interacted with problem gambling status, demonstrating that problem gamblers significantly increased their average bet sizes in games with fast speeds. Taken together, these results suggest problem gamblers have more difficulty in exercising self-control compared to non-problem gamblers regardless of speed of play, but that this effect is exacerbated with fast game speeds. However, several studies showed that speed of play had no impact on variables including both the amount of time and money spent gambling, number of bets placed, desire to continue gambling, and illusion of control. The trend appears to point towards an overall deleterious impact of speed of play on gambling behaviour and self-control, but results are inconsistent. This inconsistency is most likely due to the relatively small amount of studies conducted investigating speed of play, the varied methodologies used among this small sample, and the methodological limitations they possess (particularly the relatively small sample sizes). Coupled with this, the present review clearly demonstrates that there is a lack of studies with longitudinal designs and that those studies with small sample sizes include relatively few individuals with gambling problems making it difficult to provide any definitive conclusions regarding the impact of speed of play on both problem and non-problem gamblers and/or the differences between them.

The examination of the impact of speed of play on gambling behaviour in a real gambling venue, using commercially available gambling products, has the advantage of assessing gambling behaviour in its’ natural environment. While this adds a great deal of ecological validity to the findings, a drawback is that tight experimental control measures are sacrificed. For example, it has been found that the structural changes made to machines in such studies are often not made in isolation. That is to say, speed of play was not the only variable manipulated, making it more difficult to ascertain casual influence on gambling behaviour. Of note, several of the laboratory-based studies also fall victim to this limitation, but as a whole, experimental research in the laboratory environment has the added advantage of implementing higher levels of experimental control and more easily isolating the impact of speed of play on gambling behaviour. Whilst it is acknowledged that gambling comprises a wide range of structural and situational characteristics (McCormack and Griffiths 2013; Parke and Griffiths 2007), a more parsimonious approach is essential to learn more about specific structural characteristics (in this case speed of play) and its’ influence on the gambler.

Another factor potentially driving the inconsistent findings is the nature of the speed of play manipulation in this body of research. As a case example, the ‘fast’ speed of play condition in Ladouceur and Sevigny’s (2006) experimental study consisted of a five-second event frequency, whereas this would not even qualify as the slow condition in both Mentzoni et al.’s. (2012) and Linnet et al.’s. (2010) experiments, and only matched the speed of the slow condition in Delfabbro, Falzon, and Ingram’s (2005) experiment. This has important implications in the way results are interpreted and reported. It may be more beneficial for research findings to be interpreted in terms of the event frequency itself (measured in seconds, for example), rather than any subjective interpretation of what makes for a ‘fast’ or ‘slow’ condition. This way, results can be standardised and made to be more comparable across studies, and also has the added advantage of helping to ascertain speed thresholds where self-control becomes facilitated or degraded. It would also be advisable for speed manipulations to be anchored and manipulated proportionally around industry standard event frequencies, which occur approximately every three seconds on EGMs, allowing event frequency speed results to be assessed against existing industry benchmarks. Furthermore, there are relatively few studies that manipulate speed comparable to the faster pace of games found on online gambling platforms, emphasising the need for investigations evaluating the impact of both decreases and increases in speed of play.

There is also the argument that the impact of speed of play may not be immediately visible by assessing direct and overt gambling behaviour in some cases. The impact may be more subtle, and not captured within a relatively short experimental session, where the effects of speed on behaviour may take impact over a more sustained period of time by influencing executive functions vital for self-control not assessed in these studies. For example, these studies did not assess core executive functions such as response inhibition, reflection impulsivity, or response modulation, functions which act as the antithesis to a more impulsive style of response, and functions that act as predictors of risk-taking behaviours (Mahmood et al. 2013). Emerging evidence has demonstrated that executive control capabilities can be influenced by structural characteristics in a gambling context, characteristics that include stake size (Parke et al. 2015), as well as speed of play (Harris and Griffiths 2016). Furthermore, evidence suggests that facilitating response inhibition in a gambling context leads to a preference for less risky gambling-related decisions (Verbruggen et al. 2012).

The present review identified just one study utilising a qualitative self-report approach, the findings from which supported the empirical studies in which gamblers frequently report a preference and increased levels of enjoyment in games with high event frequencies. Obtaining first-person perspectives may offer fruitful information not otherwise available to gambling harm-minimisation research studies via the experimental method. Interviews and focus groups may provide insights into alternative ways of facilitating self-control during gambling, without excessively slowing down the speed of the game, which has been shown to have a consistent detrimental impact on the enjoyment of gambling. For example, gamblers report the need for ‘a constant fix’, so one avenue of exploration may be to find ways of providing breaks in play to facilitate self-control and allow for response modulation, but whist simultaneously making the time between gambling entertaining, such as the use of non-gambling mini-games. It is also advisable that gamblers are involved in the design process of such measures, much the same way that gambling focus groups were used to help create persuasive system designs to facilitate monetary limit adherence tools in a study conducted by Wohl et al. (2014).

Despite much reference to problem gambling being associated with games with high event frequencies (e.g., EGMs), research actually investigating the impact of speed of play on gamblers is in its relative infancy. The majority of the limited empirical evidence points towards the notion that games that have a faster speed of play are more enjoyable and desirable by an array of gamblers, but that this comes at the cost of impaired self-control. The increased number of bets placed, increased time spent gambling, and the reduced ability or willingness to stop gambling during fast games, appears particularly applicable to, but not limited to, problem gamblers, suggesting close attention should be paid towards implementing measures to facilitate self-control during rapid forms of gambling. Slowing down game speed has shown some (but inconsistent) support for reducing risk-taking and facilitating self-control, although evidence suggests this would likely reduce gambling enjoyment and detract from the experience of gambling. As a result, potential perverse and unintended consequences may result from slowing game speeds, in the form of compensatory behaviours or a migration to other products. For example, if game speed is slowed, this may result in gamblers making higher volume bets to compensate for the reduced event frequency, or playing multiple products simultaneously to essentially allow for the same volume of gambling in the same period of time. Alternatively, slowing game speed on EGMs may result in gamblers migrating to online forms of gambling, where speeds of play can be much faster, and where the online environment itself can give rise to increased risk-taking and reduced inhibitions (Suler 2004).

As an alternative, researchers should investigate ways of implementing harm-minimisation tools that have the effect of making gambling safer by facilitating self-control, but that are less conducive to detracting from the overall enjoyment and experience of gambling such as slowing game speeds. One possibility mentioned is the use of non-gambling mini-games during breaks in play to both provide a chance to take a break and modulate behaviour, but maintain entertainment levels. Existing research also suggests that while breaks in play in isolation may be detrimental to gambling behaviour by increasing cravings and negative subjecting emotion (Blaszczynski et al. 2016), when breaks are accompanied by responsible gambling messages that are not overly paternalistic and allow gamblers to engage in self-appraisal (Monaghan and Blaszczynski 2010), or allow a gambler to focus on external emotional factors (Harris et al. 2016), gambling behaviour is shaped more positively.

Therefore, rather than slow down the speed of the game (which would likely decrease the pleasure of gambling for those without any problems), gambling operators should utilize gambling tools that promote responsible gambling (Harris and Griffiths 2016). There is now growing empirical evidence that some responsible gambling tools can help decrease the time and money spent playing among individuals who gamble intensely on games with fast speeds of play including pre-commitment tools such as limit-setting features (Auer and Griffiths 2013) and personalised feedback based on actual gambling behaviour (Auer and Griffiths 2015, 2016). Unfortunately, such tools can only be used on those gambling games where playing behaviour can be electronically tracked such as those online and/or those that require a loyalty card or player card to gamble. However, some operators in some countries (such as Norsk Tipping in Norway and Svenske Spel in Sweden) use mandatory player cards that tracks all gambling behaviour both online and offline and such a system could be implemented by other operators in other countries.

Finally, further research is required to ascertain the psychological mechanisms that mediate the relationship between speed of play and overt gambling behaviour. It is possible that the total impact of high event frequencies on the gambler is not immediately captured within short, single-session experimental procedures (which is why, as mentioned above, longitudinal research is needed), or that it is not immediately observable using overt gambling behaviour. What may be required is the use of proxy measures deemed essential for the application of self-control, particularly relevant in a gambling context. Such measures may include core executive functions that act as the antithesis to impulsivity, for example response inhibition and response modulation.