Consultation response

Online games design and reverse withdrawals

Consultation response for online games design and reverse withdrawals and proposed changes to the design of online slots.

Published
2 February 2021

Last updated
29 August 2023 - Changes

Document actions
Print or save

Online games design and reverse withdrawals

Prohibiting auto-play functionality for online slots

Due to our evolving view of risk and research which cites auto-play as a potential contributing factor to gambling related harms, we proposed removing auto-play for online slots, which is the product most heavily associated with auto-play functionality. This research into the risks posed by auto-play was supported by feedback from our interim Experts by Experience (EbE) group and by data provided by GamCare (opens in new tab) from its online support forum for service users.

Consultation question

Q17. Do you agree with the proposal to prohibit auto-play on slots?

Respondents’ views

Although the majority of respondents did not agree with the proposal to prohibit auto-play on online slots, 29% either agreed or strongly agreed. While operators were the least positive about the proposal with around 18% in favour, members of other organisations and individual respondents were also not supportive with negative response rates of 73% and 65% respectively.

Quite a few operators raised concerns about the evidential basis for prohibiting auto-play and mentioned that it was not clear how we had formed the proposal to ban it. Questions were asked about the correlation between auto-play usage and problematic behaviour. Similar views were expressed by trade bodies with the BGC mentioning that auto-play was one of the areas it has been looking into along with some of its members and that the industry are willing to share data with academics which could be used to assess the impact of auto-play on gambling related harms.

One operator indicated that they were concerned we were over-representing the views of those who had been harmed by their gambling, namely that we were giving too much prominence to the GamCare (opens in new tab) survey results and the feedback from our interim Experts by Experience group. In addition a number of operators and a trade body pointed to previous data shown to us by the (then) Remote Gambling Association, which they said showed that auto-play use did not correlate with gambling harms, and therefore in their view did not pose a disproportionate risk to players.

A number of operators and one individual suggested that reducing the maximum number of auto-play spins from 100 may be more appropriate than prohibition, with suggested limits ranging from 10 to 50. Other suggestions included limiting the maximum stake permitted using auto-play or limiting the total number of times auto-play could be used per session.

Operators, trade bodies and some members of the public suggested that auto-play could be used as a way to control gambling expenditure. This was because consumers may use auto-play to play a set number of spins which could be determined beforehand and that it was easier to stick to compared with individually committing to each spin and keeping track of how many they had played.

Operators expressed views that the current requirements in place for auto-play, such as requiring a customer to set a loss limit before commencing the gamble, provided consumers with a suitable amount of control and that they required a customer to think about the total financial commitment before using the feature. A number of operators felt that prohibiting auto-play and removing these additional controls could be counterproductive to the aims of the proposal. One respondent suggested updating the controls around auto-play to ensure they were the best versions, another suggested auto-play could be required to stop for wins over a certain amount.

Several operators stated that in light of the proposal to set the minimum game cycle duration to 2.5 seconds, the removal of the auto-play feature seems unnecessary. Further, several operators pointed out that auto-play does not necessarily speed up play. For instance, one operator referenced figures which indicated that the average game cycle was one minute slower per 50 spins when auto-play was in use.

A small number of respondents informed us that removing auto-play would be detrimental to the player experience, that being required to press a button each time would be exhausting or tedious, and that removing auto-play would be unfair on some of the customer base.

Several consumers raised concerns relating to disabilities or other physical conditions for whom access to play might be affected by the proposal to remove auto-play. Consumers also expressed concerns noted in other sections of the consultation about the impact this proposal would have on bonus play through.

A group of operators raised concerns that the proposal would make unlicensed gambling sites more appealing and the practice of using auto-play would become unregulated due to circumvention. A few respondents mentioned that auto-play was already prohibited in other jurisdictions, whilst one individual stated the current requirements on auto-play meant it was more heavily regulated than in some other countries. One respondent asked whether it was our intention to remove the auto-scratch functionality from other products such as scratch cards.

Additional research on auto-play

While analysing the consultation responses it became clear that our proposal was not supported by some respondents, partly because of the apparent popularity of auto-play for some consumers and partly because of the potential for unintended consequences caused by the:

Removal of current pre-commitment features of autoplay. At present consumers using auto-play are required to set a loss / spin limit up to a maximum of 100 spins, with some responses highlighting the risks which may be associated with the removal of these pre-commitment tools.
Potential knock-on effects on how consumers play slots should auto-play be removed, with some respondents flagging the risk of faster play, longer sessions or some consumers increasing stakes/loss.

Given the breadth of views expressed we considered three additional strands of data collection and research to inform our response. This additional evidence covered different perspectives, including general slots players, as well as those who have been harmed by their gambling.

Target cohort	Aim	Vehicle	Type	Number of responses
Online slots players (consumer research)	Seek views on auto-play from online slots players, explore how it’s used and whether it correlates to intensity of play and to potential harms.	Online, with nationally representative sample	Primarily quantitative, with some qualitative content	358 online slots players / 190 auto-play users
Those who have played online slots and experienced harm from their gambling	Seek targeted view on what (if any) role auto-play featured in the harm they experienced	Survey hosted on GamCare (opens in new tab) service user forum	Primarily qualitative, with some quantitative content	33 service user forum responses
Operator transaction data for auto-play users	Explore the quantum of use of auto-play by online slots players (numbers of spins, sessions, accounts using auto-play etc.)	Data request to operators	Quantitative	2 x operator [business to consumer] datasets 1 x supplier [business to business] research

The response rates (190 from the consumer research and 33 from the GamCare (opens in new tab) survey) have been robust enough to help inform our position, especially as the consumer survey analysis has been derived from a nationally representative sample of 2,000 respondents.

The consumer research³ highlighted that auto-play has been used by 41% of online slots player respondents during the last 12 months (in addition 10% said they’d used it but >12 months ago) and that the feature was used most by the under 35 age cohort, where 51% had used auto-play during the last 12 months. This compares to 35% for those aged 35-54 and 24% past year use for those aged 55 and over.

In terms of frequency of use (past 12 months auto-play users only), the split was roughly equal (one-third each) between those who use auto-play more than half of the time, less than half of the time and very rarely. The under 35 age cohort were more likely than any other age cohort to use auto-play more than half of the time (37%).

Analysis of the responses reinforce the potential risks posed by auto-play for some users. 42% of respondents who use auto-play agreed that they lose track of their play whilst using auto-play, and almost one-third agreed that auto-play makes it difficult for them to stop gambling. This supports the academic research which points to the potential for dissociation for some players when using the feature.

Simultaneous play on multiple machines⁴ has been identified in the offline environment as a risk to players, as has online functionality “which allows customers to play two online gambling products simultaneously via a split screen function”⁵. In the context of these risks, the finding that 34% of respondents who had used auto-play agreed that they gamble on other activities at the same time as using auto-play (i.e. when the reels are spinning) does not seem compatible with an aim to keep players safe. While this may not happen during every session it still indicates a sizeable minority who are exposed to this risk.

The research findings support our concerns about the potential intensity impact of auto-play, where we sought to understand the relationship between auto-play and the three main dimensions of intensive slots play (time, spend and speed). The findings indicate that in the view of the respondents, auto-play correlates to a larger degree with increased speed of play, but still for a significant minority it also correlates with spending more money and spending more time on a game as a result of auto-play than originally intended.

The majority (58%) of auto-play users agreed that auto-play had resulted in them playing a game faster than they had intended.
Almost half (45%) agreed that auto-play had resulted in them spending more money than they had originally intended, and
38% agreed that auto-play had resulted in them spending more time on a game than they had intended.

While it is not possible to attribute causality in terms of auto-play and intensity of play, these findings imply that there is a strong relationship between them.

The consumer research indicated that for some players auto-play was used as a tool to mitigate some of the potential risks they faced. 45% of auto-play users agreed that it helped them budget by setting a maximum they are prepared to lose and 35% agreed that auto-play helps them make sure they take a break in play.

As per some of the consultation responses, a sizeable proportion of respondents (34%) agreed that the feature increased their enjoyment of gambling.

A small number (around 7%) of respondents indicated that they felt using auto-play improved their chances of winning, despite it having no such impact.

The purpose of the research into the views of those GamCare (opens in new tab) service users who have used auto-play and been harmed by their gambling was to understand better the role that auto-play might have played in the issues they faced. The survey was shared through GamCare’s publicly accessible online forum for one month and received 33 responses⁶. It found that the majority of this group (60%) felt that auto-play either contributed (45%) or was one of the main causes (15%) of the harms they experienced.

The majority of respondents agreed that “auto-play contributed to me gambling my allocated funds more quickly than I had planned” (67% agree), “I lost track of my play whilst using auto-play” (61% agree) and “Auto-play contributed to me spending more money than I intended” (55% agree).

Overall, two-thirds of the respondents felt that online slots would be made safer by a prohibition of auto-play.

The data provided by two operators appears to show that auto-play stakes tend to be on average at a slightly lower level than stakes for all slots spins - namely there is a higher prevalence of smaller stakes on auto-play. In addition, the average stake decreases as the proportion of auto-play use increases.

It also indicates that sessions involving auto-play tend to last longer and have a longer average spin cycle, in comparison to session data for all slots spins.

One operator’s data indicates that auto-play use increases with age.

The supplier research presented to us based on data from a licensed operator, indicated that as the inferred risk score of players increased, the proportion of spins using auto-play decreased. It also showed non-linear relationship between auto-play usage and risk factors such as financial loss; as the proportion of spins using auto-play increased up to around 30% for an individual consumer, financial loss decreased, before levelling off and increasing once again past 40%.

Our position

After consideration of our original position, the responses to the consultation and further research, we will introduce the proposed requirement to prohibit auto-play for online slots.

We proposed removing auto-play due to the potential link to gambling related harms identified through academic research as well as feedback from our interim Experts by Experience (EbE) group and by data provided by GamCare (opens in new tab) from its online support forum for service users.

The consultation responses highlighted a broad spectrum of issues, including opposing views on whether auto-play is harmful or is used by some as a gambling management tool. It is possible that the different views are valid for different players or at different times for the same player particularly as it is clear that use of autoplay is not a niche activity.

Operator data shared with the Commission in 2019 was presented as evidence that auto-play is not a contributor to harms. This was not a view we shared as the data examined the relationship between auto-play use and a consumer’s propensity to self-exclude⁷ from gambling - rather than correlating auto-play use against a significantly more robust measure such as PGSI.

We were keen to augment the existing academic evidence base by seeking the views from those who have been impacted by gambling related harm and we do not agree with the suggestion that in doing so we have disproportionately listened to one group at the expense of others. Our approach to understand whether these responses are indicative of a wider sentiment from consumers led us to seek the three stands of further, targeted evidence.

Our analysis of the additional GamCare (opens in new tab) service user survey responses showed that auto-play appeared to play a role in the harm experienced by the majority of this cohort.

The additional consumer research supports the findings of the academic work looking at auto-play and indicates that for a sizeable minority of auto-play users there appears to be a correlation with issues around intensity of play, as well as dissociation expressed through agreement that players lose track of play whilst using auto-play. It also highlights that for around one-third of respondents, the use of auto-play can make it difficult to stop gambling. The finding that 34% agree that they gamble on other activities at the same time as auto-play is spinning reels is troubling. It adds weight to the argument that auto-play can be a contributor to gambling harms, at least for some players, some of the time.

This is a key finding in the context of player safety, especially as removing auto-play actively inserts friction and provides less opportunity for simultaneous play across multiple products. This is of particular importance given the introduction of a 2.5 second spin speed for slots which could be circumvented by using auto-play across multiple products.

The consultation responses, consumer research and GamCare (opens in new tab) research all provided evidence that consumers may use auto-play to facilitate the undertaking of other activities whilst gambling. Amongst them were extreme examples, such as hiding a mobile phone from family members which continued to auto-play. While such responses are not indicative of the wider usage of the feature, they highlight concerns about potentially harmful behaviour enabled by auto-play.

While we recognise the potential benefits afforded by the additional controls on auto-play required by RTS (around pre-commitment), and the fact that a number of respondents indicated that it encouraged them to take breaks in play; we note that auto-play does not provide the only opportunity for slots players to use these facilities or for operators to encourage or require players to use them.

On balance we do not assess that these mitigating factors sufficiently counter the potentially sizeable harms associated with auto-play.

The consultation has provided helpful evidence that for some players with particular health conditions auto-play can be an enabler of play. While we recognise our proposals can provide a challenge for those players, they are also likely to be exposed to the risks associated with auto-play identified previously.

The structural use of auto-play and its potential to facilitate play on multiple products at once does not seem compatible with the requirements for operators to keep players safe. It is therefore our decision to proceed with the prohibition of auto-play for online slots.

We will be vigilant where licensees seek to circumvent the intention of this prohibition and where necessary that will prompt further action.

This requirement will come into force on 31 October 2021.

RTS requirement 8C

The gambling system must require a customer to commit to each game cycle individually. Providing auto-play for slots is not permitted.

References

³ 190 respondents who have used auto-play.

⁴ Machines Research Programme: Report 1 – Theoretical markers of harm for machine play in a bookmaker’s. Wardle, Parke & Excell (2014).

⁵ Advisory Board for Safer Gambling: advice to the Gambling Commission on actions to reduce online harms

⁶ While the forum is a safe and secure space for GamCare service users, the forum’s accessibility means that anyone could complete the survey

⁷ Research has shown that self-exclusion is not a satisfactory proxy for problem gambling or gambling related harm. See - Griffiths, M.D. & Auer, M. (2016). Should voluntary self-exclusion by gamblers be used as a proxy measure for problem gambling? Journal of Addiction Medicine and Therapy, 2(2), 00019

Previous section
Prohibiting player-led “spin stop” features Next section
Prohibiting effects that give the illusion of “false wins”

Last updated: 29 August 2023

Show updates to this content

Following an audit corrected link formatting issues only.

Cookies on the Gambling Commission website

Online games design and reverse withdrawals

Contents