Exploring gamblers’ experiences of problem gambling interventions: A qualitative study

Abstract A variety of services are available for those who experience problems with their gambling, though understanding about the effectiveness of and mechanisms behind these interventions remains limited. Insight into how users of these interventions experience them is limited further still. This is therefore an exploratory study of 10 problem gamblers’ experiences of the various interventions available to them, with a focus on their reasons for using them and their insights into the costs and benefits of each approach. The results described three main themes: (1) “degrees of investment” which illustrated the factors affecting investment in different interventions, highlighting the importance of the shared experience; (2) “social comparison” which highlighted how comparing one’s own gambling experience to that of others (both positively and negatively) bolstered recovery goals; and (3) “what works” whereby participants unpacked what they felt was effective for them and why, such as personalisation. Transcending these themes was the perception that “experience is expertise” and that due to their lived experience gamblers should be considered the experts on recovery over and above “trained professionals”. Overall, Gamblers Anonymous emerged as the most positive intervention as it facilitated participants’ sense of investment, encouraged positive social comparison, was deemed to be most effective and also epitomized the focus on “experience is expertise” as the groups are run by members rather than qualified practitioners. Therefore, not only does GA involve gamblers helping each other, but provides the context for physical connection, personalisation (through being listening and responded to individually) and support through social media.


Introduction
Globally, problem gambling affects between 0.7 and 6.5% of individuals, equating to between roughly fifty-four million people and six hundred million people worldwide (Calado & Griffiths, 2016). In the UK alone, problem gambling affects around half-a-million adults, with a further twoand-a-half million people at low or moderate risk (Sullivan, 2019).
A variety of services are available for those in need, with differing requirements and intended outcomes; some focus on complete abstinence whilst others focus on a reduction of gambling symptoms. It is not clear which route is optimal, and variation in approach reflects the absence of a single unifying theory to describe problem gambling. Nonetheless, various help services are available for those who

Cognitive-behavioural therapy
Cognitive Behavioural Therapy (CBT) focuses on identifying cognitive distortions about gambling (for example, illusions of control) and correcting them through cognitive restructuring techniques (such as reinforcing non-gambling behaviours; Chrétien et al., 2017;Tolchard, 2017). There is also a focus on weakening preservation patterns, irrational beliefs and "magical thinking" associated with problem gambling (Pallesen et al., 2005). CBT is the most empirically validated approach to treating problem gambling (e.g. Cowlishaw et al., 2014;Gooding & Tarrier, 2009;Ladouceur et al., 2001;Petry et al., 2006;Rash & Petry, 2014;Raylu et al., 2016;Ribeiro et al., 2021;Tolchard, 2017). A recent Cochrane review found 9 studies, some assessing CBT on an individual basis and some in a group scenario, all of which demonstrated that CBT significantly reduced a range of gambling recovery outcomes (ranging from reduction of DSM-IV criteria met, gambling behaviour, severity, craving, abstinence, and money spent). CBT has also been demonstrated to be an effective approach to problem gambling in a range of populations, with age, gender, ethnicity or psychiatric comorbidity not affecting its effectiveness (e.g., Champine & Petry, 2010;Cowlishaw et al., 2014;Ribeiro et al., 2021;Rossini-Dib et al., 2015;Tolchard, 2017), Often, however, individuals experiencing distress do not seek help, even if this distress is severe (e.g., Gainsbury et al., 2014;Kaufman et al., 2017, and commencing and adhering to treatment can present a challenge for individuals with gambling-related problems. Indeed, less than 10% of people who experience problems with their gambling seek formal help (American Psychiatric Association, 2018). This may be for a number of reasons including: a desire by people to handle their problems by themselves (or a belief in their ability to do so); feelings of shame, secrecy, embarrassment, pride and fear of stigma when admitting to, and having to seek help for, a gambling problem; denial about or minimisation of the problems associated with gambling; lack of knowledge regarding the options available; practical issues involved in attending and concerns about the content, quality and effectiveness of treatments available Kaufman et al., 2017;Suurvali et al., 2009) In order to combat these and other issues inherent to formal psychological intervention (for example, cost or lack of resources meaning long waiting lists), research into online and mobile interventions is increasing rapidly . These types of interventions also have the potential to include features associated with more effective healthcare communications, such as tailoring, interactivity, personalisation, and/or high message repetition (Parvanta, 2011). They have also been found to improve engagement and retention of treatment services (Milward, Day, Wadsworth, Strang, & Lynskey, 2015). Whilst research exists about what types of online and mobile interventions are available, generally studies have focussed on either quantifying effectiveness (for example, measuring the effect of an intervention on gambling urges or behaviours), comparatively little qualitative research has been done. The existing qualitative research has, rather than focussing on the actual experience of those using these types of interventions, focussed on the experiences of problem gambling itself (Järvinen-Tassopoulos, 2016) or exclusively on Internet-based counselling sessions (e.g., Rodda et al., 2013, Rossini-Dib et al., 2015.

The current study
Face-to-face interventions (such as CBT) and mutual aid groups (such as GA) frequently appear in the literature, yet few qualitative studies have been completed to understand exactly how users of these interventions experience them. Similarly, whilst online and mobile interventions have been effective and well received for individuals with substance and alcohol use disorders, one cannot assume the experience for those with gambling problems is comparable. For example, whilst gambling addiction and alcohol or substance addiction share many characteristics, some fundamental differences exist. An obvious example of this is method of consumption; you cannot consume alcohol or illicit drugs via the Internet or mobile phone, but online and mobile gambling is popular, and participation continues to grow. This might affect the way problem gamblers interact with, and experience, Internet, and mobile interventions, or may affect their choices about what interventions are appealing to them. The lack of insight into how problem gamblers experience gambling interventions may be hindering research efforts to find the most effective support for those in need.
With this in mind, the present study therefore aimed to explore problem gamblers' experiences of the various interventions available to them with a focus on their reasons for using them and their insights into the costs and benefits of each approach. By answering these questions, the results of this research will help researchers and other professionals make better decisions about support provisions for problem gamblers.

Design
A qualitative design was used to obtain detailed data sufficient for a rich and in-depth understanding of problem gamblers' experiences of gambling interventions. Data were collected using semi-structured interviews which were analysed using Thematic Analysis (Braun & Clarke, 2006). Data collection was completed between August and November 2019, just prior to the COVID pandemic.

Participants
The inclusion criteria for this study were: be an adult over the age of 18, speak and understand English, have sought some form of help for gambling problems, and be actively seeking help at the time of the study. Participants (N = 10) were 1 woman and 9 men who were between 28 and 68 years of age. One participant chose not to disclose his/her age. All were White British. They had sought a range of help including GA, online resources, books, mobile apps, CBT, and counselling services. All had disclosed a wish to stop gambling, and had sought some form of help at the time of study. Four participants were actively involved in GA, two had attended previously, all but two had accessed online help, half had undertaken CBT and counselling, two had tried mobile apps and two had used books. The demographics form was anonymous due to ethical concerns and the need to keep participant information confidential. As such, and therefore their demographics could not be linked with their interview data. Demographic data is presented in Table 1.

Procedure
A mixture of purposive and snowball sampling was used to recruit participants via links to organisations which are in direct contact with people who experience gambling problems. Emails were sent to GamAnon (a support group for family members of problem gamblers), Gamblers Anonymous, to other researchers with whom the researcher had established links with, and to individuals working in the gambling industry who were known to the researcher. Social media was also utilized. Participants were included if they were an adult over the age of eighteen years, had experienced problems because of their gambling (by their own admission, rather than specifically having been given a formal diagnosis), had experience of offline, online, or mobile interventions for problem gambling and were able to speak and understand English.
Participants were given an information sheet detailing the nature of the study and asked to provide informed consent (either by a hard copy in person or an emailed copy). They were also given a demographics sheet including questions on age, ethnicity, education, employment, types of gambling, time per week spent gambling, and what kind of help they had sought.
Interviews were predominantly conducted face-to-face (n = 6) though some were conducted by telephone (n = 4). An interview schedule comprising the following four open-ended questions was developed and used as a guideline: "Can you tell me about any experience you have with face-toface interventions?"; "Can you tell me about any experience you have with online interventions?"; "Can you tell me about any experience you have with mobile interventions?"; "Can you tell me about any experience you have with text-message interventions?". The interview schedule was developed to include these areas based on a review of the previous literature, and to capture experiences of any types of intervention participants may have experienced. Interviews lasted between 15-80 minutes and were recorded on a voice recorder. When the interviews were complete, participants were given the opportunity to ask any questions they might have and were asked for consent to be contacted about future studies. Data were subsequently transcribed, and participants were assigned pseudonyms to maintain their anonymity. The names of the audio files were then changed to include these pseudonyms.

Data analysis
Data were analysed using Thematic Analysis (Braun & Clarke, 2006). This method is data-driven and allows a flexible approach to discover how individuals find meaning in their experience. Thematic Analysis also encourages a structured approach to analysis because it is free from ties to any specific theoretical framework (Nowell et al., 2017). Thematic analysis is appropriate for interview studies and allows for a good overview and summary of a large amount of data. It also has the potential to generate "unexpected insights" (Braun & Clarke, 2006).
The analysis was conducted with a critical realist epistemological perspective (Bhaskar, 2008), allowing the researcher to accept the objective realities of the participants, whilst acknowledging that their realities sit within a cultural, societal and historical context.
All interview data were transcribed by the researcher. During this process initial thoughts and ideas were noted down. The transcripts were then analysed in six stages, in accordance with the guidelines set by Braun and Clarke (2006). The analytic procedure was as follows: (1) interview transcripts were read and re-read several times to ensure the researcher gained a good overall sense of the participants' accounts. Anything interesting or significant was noted, and potential themes were recorded in a research diary; (2) codes were then developed for both the semantic and latent content. These codes identified features of the data which the researcher felt to be important to the research question; (3) emerging themes were noted separately, and attention was focused on the interrelationships between these themes; (4) themes were then reviewed by returning to the original data set and comparing the themes against it. Themes that had insufficient data to support them or were too diverse were then deleted. This refinement was undertaken on two levels: Firstly, using the coded data thus ensuring a coherent pattern, and secondly, this coherent pattern was then considered in relation to the data set as a whole, to accurately reflect the data set. Any codes that had been missed in earlier stages were applied at this time; (5) themes were then defined and named; (6) finally the report was then produced, which included using verbatim examples taken from transcripts to illustrate elements within the themes. This was an interactive process involving ongoing discussions with the research team.

Results
Participants described having engaged in a wide range of face-to-face and online interventions such Gamblers Anonymous, Cognitive Behavioural Therapy, counselling, online self-help (such as forums) and gambling blocking computer software; most had used several different approaches. Through analysis and repeated coding three main themes and a transcending theme were developed, each with related subthemes. An overview is presented below followed by a detailed analysis of the themes and subthemes, illustrated with exemplar quotes from the transcripts.

Overview of themes
Participants' responses centered around three themes: i) "degrees of investment" which illustrated the factors affecting investment in different interventions, illustrating the importance of shared experience, ii) "social comparison" which highlighted how comparing one's own gambling experience to that of others (both positively and negatively) bolstered recovery goals, and iii) "what works" whereby participants unpacked what they felt was effective for them and why. Transcending these three main themes was the perception that "experience is expertise" and that due to their lived experience gamblers should be considered the experts on recovery over and above "trained professionals". These themes and their subthemes will be discussed below and illustrated using exemplar quotations.

Theme 1: degrees of investment
This theme describes the factors which affected the degree to which participants felt invested in different interventions and has three subthemes: "something to believe in", "emotional connection" and "physical connection".

Something to believe in
The most frequently discussed interventions identified by participants included Gamblers Anonymous, CBT, and counselling. Of these interventions, GA was described in the most detail, with the most enthusiasm, and experienced in the most positive way. GA was also the most frequently discussed of all the intervention types, with all but one participant having at least some experience of it. One of the most pertinent observations from the results was how highly participants regarded GA. In fact, GA was the single intervention which was spoken about with any real enthusiasm or optimism, with some going as far as describing the experience as transcendent. James said: " . . . now do I have a sort-of higher power now? Do I believe in a higher power? Yeah, I do, 'cause I'm kind of just a bit switched-on to it now. But, I didn't. And loads of people in the room don't. You know they call the room their sort-of higher power. This is the thing that gives them hope. This is the thing that they . . . that they truly believe in." (James) For James, his experience of GA has been so positive that he considers it as providing him newfound spirituality and something to "truly believe in"; a feeling which is echoed by the other members of his group. Participants conveyed that CBT did not provide for them something to "truly believe in", that is, when embarking on a programme of CBT, whereas attending their first GA meetings did provide this for them. This experience was shared by Paul: " . . . of various types, and, and, of various lengths of sobriety, you know, it's um, for me that's certainly the most valuable tool." - (Paul) Paul indicates that he has tried various other interventions with varying degree of success and recognises GA as providing the more valuable support of these. Daniel goes further than this, to say: For Daniel, he feels as though he has found a concrete solution in GA. The use of the word "answer" implies that he perceives his gambling addiction as something he has been seeking an answer to for some time, and in finding GA has finally been able to solve this problem.

Emotional connection
Participants largely attributed their positive experiences of GA to the feelings of emotional connection it provided them. Through their shared experiences of gambling addiction, participants felt they were able to connect with other members of GA in a way that they felt unable to with individuals outside of GA; their experiences allowed them to give and receive "authentic" empathy to and from others. Through attending GA, individuals felt understood, listened to, and ultimately not alone. James said: "What is common across all of the people in the room is that they crave connection . . . I think that's the thing -it's that . . . it's just having an acknowledgement that somebody understands you in the way that no-one has before." -(James) Human connection and genuine understanding are clearly incredibly important and powerful to problem gamblers. Daniel Said:

"The addiction is isolation and if there is a remedy it's connections."-(Daniel)
Daniel describes the connections made through GA as so powerful they are the "remedy" or "cure" for his gambling problems. Indeed, many of the other participants shared these feelings, describing GA as the physical embodiment of connection, the opposite of isolation, and the ultimate solution. In contrast, when talking about CBT, participants described feeling like the treatment providers were unqualified to help gamblers because they had no personal experience of gambling problems and thus were not able to understand them (or provide "authentic" empathy). Feelings and perceptions of being judged in formal treatment settings also contributed to participants' negative experience of these interventions.

Physical connection
Online and mobile interventions were also experienced negatively because-by design-they are devoid of physical interaction. Human contact and connection were clearly very important to the participants of this study, so much so they believed they were fundamental to their recovery. By design, online and mobile interventions typically offer much less, if any, interaction with others. Daniel said: "I'm a bit hesitant about the online thing as it currently is, because even if you're joining in, you're interacting, you're writing back. You know, you can go on an online dating website and exchange messages with someone, but you might never meet them." - (Daniel) When talking about online support, Daniel describes a reluctance to engage with online interventions, implying the interaction with others is lacking a fundamental element; he feels it is not worthwhile engaging with other individuals online because "you might never meet them". This clearly emphasizes the need for Daniel to form connections with people beyond the online domain. This lack of physical interaction with others also contributed to feelings of isolation and demotivation to continue recovery. Paul said: "When you're heavily in addiction it's very, er, solitary, and you're isolating so, for me the option of, you know, having something online that I can sit in my room and listen and learn and do online, for me would only enhance my addiction." -(Paul) Paul describes engaging with online interventions as an isolating experience and feels as though, as with the online forums, this experience would actually contribute to an increase in gambling, rather than offer any kind of solution.
The above demonstrates how GA provided members with positive, almost transcendent feelings through providing both emotional and physical connection with other likeminded people, which alleviated feelings of isolation brought about by gambling addiction. Through these features, GA represented hope through being something to believe in.

Theme 2: social comparison
This theme demonstrates how people compare themselves and their experiences to those of others with gambling problems. For those in face-to-face meetings with others this happened whilst in the room, but for those undergoing individual CBT or counselling therapy on engaging in online interventions this process of comparison could happen through hearing about others or reading the literature. This process of social comparison could happen in both a "productive" and "destructive" way.

Productive
Participants who had been to face-to-face GA meetings described using these meetings for both upward and downward social comparison; some described hearing the "horror stories" which helped them stay focused on their abstinence goals, such as Rick: For Luke, hearing that other people have been in his positive and managed to get out of it almost acts as a demonstration or proof that he too can do it, changing his mindset from thinking there is no possibility of his situation improving.

Destructive
Social comparison also resulted in counter-productive outcomes which were destructive and resulted in continuing or worsening gambling behaviours. For some, this type of outcome was more related to CBT/counselling: For James, attendance at formal treatments had the opposite of the intended effect; the short periods of abstinence achieved after sessions were actually used as justification of continuing gambling, making engaging in CBT/counselling an incentive to gamble (if he engages with CBT/ counselling he is allowed to gamble). Jane had a similar experience to James. She said: "I've done some CBT and I do find it's useful at the time but at the minute its stops, it's gone."-(Jane) All the other participants in this study had similar experiences to those described by Jane and James; CBT would help them achieve abstinence for short periods of time (usually the duration of the course), but they would relapse once the course of treatment is complete. For Ben, reading about other gamblers whom they perceived had more severe gambling problems than themselves served as justification that their gambling wasn't severe enough for them to have to abstain. For others, reading about how negatively gambling had impacted lives resulted in a negative affective state, for which they turned to gambling to alleviate.
The above describes how individuals compare themselves and their experiences to others, both in productive ways (which helped strengthen abstinence goals) and destructive ways (which led to continuing or worsening gambling behaviour).

Theme 3: what works
Throughout the interviews participants discussed the positive and negative features of each intervention in terms of three subthemes: "personalisation" "responsible gambling", and "the role of social media".

Personalisation
Personalisation was identified as one of the most important aspects of a "successful" intervention, or at least a positive experience of that intervention. Conversely, lack of personalisation lead to feelings of non-committal, demoralisation, boredom, and unwillingness to engage. Jane said:

"It has to be personal. It has to be at least a bit personal because if you don't make it relatively personal to the person then it's not going to work . . . like what's the point? i.e., it's no good just having generic texts that mean nothing. It has to be personal. If that can be achieved that's the way to go." -(Jane)
For Jane, personalisation is fundamental to the success of interventions, implying that she would not engage with an intervention which did not include this feature. It should be noted that Jane had no experience of any intervention which included personalisation, so rather than talking about the success of such an experience, she commented on the negative impact of an absence of personalisation-an experience shared by other participants. Martin said: "It would be really useful to kind-of personalise it because if you feel like you're just kind-of part of the system and you're just a number or just a spec on a chart or something then that really demoralises me -I can't speak for anyone else but I feel it really demoralises me, erm, whereas I feel that the more personal and personable you make it and kind-of, targeted for a specific person, you know, that actually makes you feel like somebody has actually listened to your specific things, and again, you're not just a . . . a number on a chart, that would be really helpful." -(Martin) Martin describes the lack of personalisation as "demoralising" because his own needs have not been considered, resulting in feelings of being overlooked and aloneness and being "just another number". As the results highlight the importance of mitigating feelings of isolation, and increasing feelings of connection and understanding, it is no surprise that personalisation of interventions is described as important to participants. Participants highlighted several ways in which this personalisation could occur in online or mobile interventions, most frequently suggesting that prompts or reminders could be sent to users at pre-set personalised times.

The role of social media
WhatsApp group chats were used to send supportive and encouraging messages, and reminders about the goals of abstinence. Daniel said: "On the WhatsApp group someone will say 'just for today '

. it's a sort of unity." -(Daniel)
Other than people sharing encouraging messages, they were also used by participants to ask for support ad-hoc when they felt tempted to gamble. James said:

"We have a WhatsApp group, erm . . . and for me that WhatsApp group is a safety net. So, if ever I get an itch, or if ever I am tempted . . . or . . . it's, it's there for me to go in case of emergencies." -(James)
James describes his GA WhatsApp group as a "safety net" implying that the group does serve an important function in helping him maintain his sobriety.
WhatsApp is referred to frequently throughout the data which might be an indicator that it plays an important role in recovery, especially within GA, however, it may equally be merely a reflection on how ingrained social media has become in modern life. Participants themselves seemed unsure about the exact role of social media for them. For most, it served as an extension of support groups (GA) between sessions, however the group might actually stop people from engaging in the face-to-face meetings. Daniel said:

"So, we've got a WhatsApp group which is the closest we've got to the online thing of it, but every week I find myself saying 'this isn't online GA' because people don't come to the meeting. They'll send like a 'thumbs up' icon. 'All good here. Love this group. You guys are helping me so much' and it's like, 'Well you've only been once ever'." -(Daniel)
Daniel perceives this extension of GA to a WhatsApp group as negative because of the potential to use it as an excuse not to physically attend meetings. As previously discussed, the power of actual human connection, and therefore the potential for physical GA meetings to provide this, is incredibly important, at least to the participants of this study. Thus, WhatsApp may be counterproductive in terms of meeting this need. However, WhatsApp has the potential to provide an extension of the GA support network.

Responsible gambling
Creating barriers to gambling was identified as an important and useful strategy, especially in the beginning stages of recovery. Successful barriers identified included software which blocks access to gambling related websites, and relinquishing control of finances (which were found to be highly effective). However, many interventions designed to act as barriers to gambling were found to be highly unsuccessful, namely "responsible gambling tools" such as self-exclusion and time limits. In fact, most participants reported them to be "pointless". Paul Said: This study included participants who gambled in a range of different ways-some online, some offline and some a mixture of both, yet there was also no positive experience of selfexclusion (in either online or offline settings) reported, and generally self-exclusion was seen as futile.

"Having in place those measures is like asking turkeys to vote for Christmas. I'm not going to put in place a block . . . for me it's very binary: either I don't want to gamble, or I do want to gamble. I don't want this grey area in between." -(James)
For James, self-exclusion is not effective because he lacks motivation to adhere to it. In this quotation James also highlights a recognition that his feelings towards gambling are much more important and outweigh the concept of blocking himself for a time, which infers the triviality of the concept of a block function. One can infer from this that, at least for James, the change needs to come from within.
Another problem with self-exclusion is that it can also be easily circumvented. This finding was consistent throughout the dataset, which shows a fundamental flaw in how responsible gambling tools are being implemented. The experience of this made participants feel as though the lengths gambling operators are going to ensure the safety of their users are woefully inadequate. James said:

I wrote to the Gambling Commission about six months into my recovery 'cause I just didn't believe that they had put in place enough barriers . . . no sorry, their members, not the
Gambling Commission, they're not a gambling firm, but you know, their members and the community that they . . . that they regulate, I didn't believe they were putting enough barriers in place . . . it was only after I'd gambled that they wrote to me my account had been blocked, so after they had taken my money. It was the principal that they could obviously recognise that they . . . there is a flag somewhere, but that flag is after the event. That is . . . to me, that was just morally wrong." -(James) For James, the burden of responsibility falls with the gambling operators which they are inadequately addressing, and thus failing to protect their users from harm. More concerningly, James believes that the gambling operator knowingly allowed continuation of betting even after he had attempted to self-exclude, which has very serious implications about the role gambling operators have in the development and continuation of gambling problems, especially for those who are proactively trying to help themselves. This experience was consistent across all participants, and all perceived this failure to protect consumers is deliberate.
The above describes the features of different interventions participants found to be positive which included personalisation (such as through tailoring prompt messages to suit individual), the utilization of social media as a form of "safety net" and to provide connection to others, and how responsible gambling tools can be utilised to help, especially in the beginning stages of recovery. These tools were however also seen as falling short of their intended purpose.

Transcending theme: experience is expertise
Participants therefore described "degrees of investment" in the interventions facilitated through having something to believe in, emotional and physical connection, as well as making "social comparisons" which could be either productive or destructive. They also more broadly discussed "what works" about the different interventions (and what doesn't).
Transcending the main themes was that experience is expertise and that gamblers are the authority on gambling addiction recovery over and above "trained professionals" with no gambling experience.
Regardless of whether the interventions were offline or online, the most important thing to participants of this study was that the help given to them came from other gamblers. GA was identified as the most positive intervention because it not only provided participants with something to believe in, emotional connection, physical connection, social comparison, a sense of personalisation (through being listened and responded to individually), but it also provided barriers to gambling through the support of the group and additional support in the form of social media, and support when participants felt that responsible gambling tools had failed them. The underlying reason GA was able to provide all of these attributes was that the support was being delivered by gamblers themselves. Paul said: "I guess if I was to compare the GamCare to Gamblers Anonymous, GamCare I was with, you know, a trained addiction counsellor, however someone who didn't have any personal experience of addiction and the way you feel, and, you know . . . it was very textbook in terms of responses, which . . . yes there is a science behind it, however, you know, in Gamblers Anonymous what it is, is there's that personal element of it, erm . . . you can identify with everybody in the room in some capacity, you know? Erm, the feelings, the behaviours, everything that goes around it. You know, the secrecy and the lies and everything that goes hand-in -hand with addiction. You can really identify and everybody in there has experience of it. You're not sitting there, almost you know, with a teacher who doesn't know what they're doing and then other people who do. Everybody in there is recognised as a gambler." -(Paul) Paul is contemptuous of treatment providers and, despite acknowledging their training, dismisses them as unqualified to treat gambling problems, implying that only through shared experience does one become qualified. This highlights the absolute importance of emotional connection (through "authentic empathy") and ranks those who have shared experience as higher than any scientific method. This may be because shared experience facilitates feelings of being understood and mitigates feelings of being judged.

'I was up until 3am looking at Venezuelan basketball on a Bet365 site, not even the game, just numbers on a screen, and people in the meetings are sort-of like nodding at you like 'Oh yeah, I've done that before' and there's no judgment as well. That's another massive thing about it." -(Daniel)
Daniel describes the difference in how he feels when sharing his experiences of gambling addiction with other gamblers. Although not explicitly stated, it is implied that he has previously received -or perceived -judgment from non-gambling others. For him, the space to be able to talk freely and without concerns about being judged come only from talking with other gamblers, making GA not only more impactful for him, but the overall superior intervention.
A key concept through the data, and resulting themes was that addiction and feelings of isolation are intrinsically linked, and so it is only through shared experience and authentic empathy from other gamblers providing feelings of being understood and accepted rather than judged can those affected by gambling addiction begin to recover.

Discussion
This study explored participants' experiences of problem gambling interventions. The analysis described three main themes relating to "degrees of investment", "social comparison" and "what works". Transcending these themes was the belief that "experience is expertise".
In terms of "degree of investment" the extent to which individuals felt invested in an intervention largely relied on that intervention providing both emotional and physical connection, which, coupled together, created a sense of having something to believe in.
Previous research has demonstrated a clear link between social isolation and loneliness and a host of health concerns, including damage to physical and mental health, as well as overall mortality rates. Low social support has also been found to be associated with a range of addictions including Internet, Smartphone and substance addictions (Dobkin et al., 2002;Herrero et al., 2019;Tudorel & Vintilă, 2018). Conversely, when social isolation and loneliness are reduced, health outcomes are improved (see, Leigh-Hunt et al., 2017, for review). It is perhaps not surprising then that emotional and physical connection were identified from the data as important and positive features of interventions. Dobkin et al. (2002) also demonstrated that increased social support was a predictor of addiction treatment retention, similar to the findings of the current study that emotional and physical connection contributing to investment in interventions.
Participants also highlighted the role of social comparison within interventions which had the potential to be both productive and destructive. For example, in Gamblers Anonymous (GA) meetings, listening to the "horror story" testimonials of others served to bolster goals of abstinence for some but for others served as motivation to continue gambling, justified through positive comparison of their own gambling ("it could be much worse"). Reading the negative impact of gambling on the lives of others also produced negative affect which some gambled to alleviate. A review of the literature did not retrieve any research relating to the role of social comparison in gambling addiction, though there is some evidence to suggest negative social comparison is associated with addiction in other areas such as social media (Onat Kocabıyık, 2021;Robinson et al., 2019) and brand addiction (Le, 2020). These studies suggest that, as with the results of the current study, social comparison to individuals perceived as "better than" oneself can be destructive, leading to worse addiction outcomes but that social comparisons to those perceived as "worse than" result in more productive outcomes.
The third theme described effectiveness and several features of interventions were identified to be the features which "work". Consistent with prior research, personalisation was identified as an important feature of an intervention which could include, for example, tailoring the timing and content of interventions to each individual (see, Parvanta, 2011, for review). Social media also played a key role, both as a "safety net" when individuals felt they needed ad-hoc support, but also as an extension of GA meetings, providing additional feelings of connection between meetings. A review of the literature did not return any research regarding the role of social media in mutual aid groups, or for gambling addiction recovery, suggesting that this research has produced a novel insight. Given the preponderance of social media in everyday lives, and the mixed nature of these results, this area should be explored further by other researchers in order to harness the possible potential of social media in gambling recovery. However, this area may need to be approached with caution, given the addictive nature of social media, and that the potential for further negative comparison to occur.
Responsible gambling tools also played a role however it was less clear what this entailed.
Gambling blocking software appears to be a very useful tool, especially in the beginning stages of recovery however other tools such as self-exclusion were seen as futile because they only operated on a surface-level, promoting abstinence rather than addressing a much bigger underlying problem. Research in this area is similarly unclear with mixed support for the effectiveness of selfexclusion. Some studies suggest it can be effective in reducing harm (e.g., Hayer & Meyer, 2011), and that it is an important part of addressing gambling-related harm from a public health point of view (Gainsbury, 2014), however, as with the results of this study, it is also viewed as an inflexible approach which facilitates abstinence, rather than control (Productivity Commission, 2010).
Participants therefore described "degrees of investment" in the interventions facilitated through having something to believe in, emotional and physical connection, as well as making "social comparisons" which could be either productive or destructive. They also more broadly discussed "what works".
Transcending these themes was the perception that "experience is expertise" and that those best placed to provide support to gamblers are other gamblers. Through their first-hand experience, other gamblers can provide: something to believe in, through being a living demonstration than things can get better; emotional connection, through shared experience; and the ability for social comparison (to their own story). This appears to be a relatively novel finding given that most of the previous research has focused almost exclusively on therapist-driven interventions such as Cognitive-Behavioural Therapy (e.g., Gooding & Tarrier, 2009;Ladouceur et al., 2001;Petry et al., 2006;Rash & Petry, 2014;Raylu et al., 2016;Tolchard, 2017). Comparatively little attention has been given to investigating or developing gambler-led interventions. Future researchers could focus on gambler-led or mutual aid interventions, as the results of this study suggest they have potential which may have been overlooked. Indeed, a number of interventions were discussed in the interviews, however Gamblers Anonymous was consistently reported as not only the most positive, but the only one which really helped. This may be because Gamblers Anonymous has the potential to incorporate all of the individuals features identified above; not only does it involve gamblers providing the support to other gamblers and helping each other, but individuals may feel more invested in it because it can provide both emotional connection (through sharing experiences with like-minded others) and physical connection (by being present in the room with other members), providing something for individuals to believe in. GA also facilitates social comparison (through listening to other members testimonials), and many participants reported being part of a GA WhatsApp group, which provided them additional support.
One explanation for the high regard of GA as opposed to other interventions could be that it is delivered in a group scenario whereas all other interventions reported on in this study were individual. Certainly it seemed it was group scenario which facilitated the feeling of connection, solidarity and social comparison highlighted by participants. The participants in this study had not participated in any other group interventions (such as group CBT). An avenue for future research would be to include those who had experienced other group interventions as well as GA to explore whether it is the group scenario in general, or something inherent to GA, which is the key to its success.
The evidence of the effectiveness of GA is patchy and inconsistent; some studies suggest that there is no significant difference between GA and other types of therapy; some suggest that it is significantly less effective than other types of therapy, and others suggest that it can be effective in conjunction with CBT (see, Schuler et al., 2016 for review). On the contrary, the participants of this study describe their experiences of GA as positive, life-changing, and even transcendent. Almost all of those who engaged with GA were still active (and abstinent) members after many years. With these results in mind, it is surprising that there is such a dearth of research into GA as a recovery pathway, in-and-of itself. More exploration into GA as a recovery pathway is needed.

Limitations
This study does however present several limitations. Primarily, as is inherent to qualitative research, the sample size was very small and thus findings are not representative of the wider gambling population. Furthermore, whilst every effort was made to recruit a diverse sample, most participants were white British males which may have limited the results. Future research should focus on aiming to capture the experiences of a more ethnically and gender diverse range of participants.
Qualitative analysis involves an unavoidable degree of subjectivity on the part of the researcher. To try and mitigate this subjectivity affecting the results as much as possible, the researcher was mindful of their own personal biases and idiosyncrasies throughout the analytic process. In this case, the researcher was mindful that they had personal experience of a problem gambler in their personal life which may have affected results. It could however be argued that this experience allowed the researcher to connect deeply with participants and provide an amount of authentic empathy and understanding, which may have resulted in them feeling more comfortable divulging personal information.
That said, the researcher was also mindful of their status as a non-gambler may have affected participant responses. The results of this study demonstrate that understanding and authentic empathy through shared experience are of great importance, and participants may have felt unable to be fully open with the researcher. To try to mitigate this, the researcher was mindful to maintain unconditional acceptance, staying open and non-judgmental throughout the interviews. These data could help to design quantitative studies which test the validity of the present findings.
To avoid researcher bias, during the analytic process the research team held repeated discussions to ensure the results were as close to the data as possible. It is hoped that this is also managed by presenting explicitly the researchers' own theoretical commitments, and by providing a transparent account of the analytic process.
Due to data being collected pre-COVID-19, the results of this study may not be reflective of the current situation with regards to GA, CBT and indeed all post-COVID-19 addiction treatment (as most of this was transferred online due to the lockdown restrictions). The results of this study do however provide a rare basis for comparison between pre-and post-COVID-19 treatment, allowing future researchers to explore the differences and possible benefits of online treatment in a unique way.

Conclusions
Participants therefore described their experiences in terms of three key themes and a transcending theme. Across all interviews GA came out as most positive as it facilitated their sense of investment, encouraged positive social comparison, was deemed to be most effective and also epitomized the focus on "experience is expertise" as the groups are run by members rather than qualified practitioners. Therefore, not only does GA involve gamblers helping each other, but provides the context for physical connection, personalisation (through being listening and responded to individually) and support through social media. The results of this study will help future researchers and other professionals make better decisions about how to provide support for those with gambling problems.