Elite professional online poker players: factors underlying success in a gambling game usually associated with financial loss and harm

Abstract

Most gamblers lose money, and this means that a behavioral dependence to gambling can cause harm. However, some professional gamblers win consistently, and there is little academic literature on their psychology and how they differ from disordered gamblers. To contribute to this understudied area, we qualitatively analyzed interviews with 19 elite online professional poker players, by examining factors from the disordered gambling and decision-making literatures. Like disordered gamblers, participants displayed aspects of a behavioral dependence to gambling, but contrastingly did not generally experience harm. Other contrasts included their rational approach to statistical thinking, a general self-reported tendency to not be impulsive, and their social connections with other experts. One factor that did not yield clear contrasting results was whether or not they experienced early big wins. Parallels with the decision-making literature included their assessment of decision quality based on expected value rather than realized outcomes, their reluctance to take risks outside of their ‘circle of competence,’ and their ‘active open-minded’ thinking style. This study contributes to gambling psychology via an in-depth exploration of an understudied group.

Keywords:

• Gambling
• gamblers
• online betting
• psychology
• poker professionals
• professional gambling

Introduction

Gambling is a losing proposition for most people. In 2020 the Chief Executive of a major UK gambling operator confirmed this by saying to a government select committee that: ‘I am not going to sugar-coat it: 99% of the customers who play on our sites will lose, so you’re probably losing more if you play more’ (Alexander 2020). Most of the people with the heaviest levels of involvement with gambling are disordered gamblers (Rockloff 2012). Their behavioral dependence on gambling, for example as expressed as a preoccupation with gambling, results in monetary losses (American Psychiatric Association 2013). The resulting fallout can spread through families (e.g. via intimate partner violence, as described by Hing et al. 2022) and communities (Langham et al. 2015). These two constituent factors, a behavioral dependence to gambling, and harms from gambling, co-occur so consistently that valid scales of disordered gambling such as the Problem Gambling Severity Index (PGSI) contain items pertaining to both factors (Browne and Rockloff 2019). However, the above quoted statistic implies that about 1% of gamblers win, and an even smaller percentage will win consistently in their career as a professional gambler. Successful professional gamblers can presumably gamble frequently without experiencing harm, and further study of this group may yield insights relevant to the theoretical understanding of disordered gambling among the majority of heavily-involved gamblers. The present study contributes to understanding in this area by qualitatively analyzing data from 19 interviews with elite professional online poker players. This research therefore follows in the footsteps of previous qualitative gambling research (Cassidy et al. 2013; Reith and Dobbie 2013).

Professional gamblers play skill-based gambling games to make money. The study of professional gamblers might yield insights relevant to measurement issues researchers and clinicians face when applying standard gambling disorder criteria to skill-based games. For example, the PGSI (Ferris and Wynne 2001), and measures of faulty gambling beliefs such as the Gambling Related Cognitions Scale (GRCS; Raylu and Oei 2004) have been criticized for containing content inappropriate for skill-based games. For example, a winning poker player might rationally respond ‘always’ to the PGSI item, ‘Have you gone back on another day to try to win back the money you lost?’ (Laakasuo et al. 2016). However, doing so would be enough to place the player in the moderate-risk category on the scale (Ferris and Wynne 2001). Similar issues have been highlighted for at least one other PGSI item (Laakasuo et al. 2016), suggesting that the most widely-used gambling self-report scale is poor at differentiating between professional and disordered gamblers in skill-based games.

Similarly, measures of faulty gambling beliefs, which aim to uncover incorrect patterns of thinking amongst gamblers, might also work less well for skill-based gambling games (Russell et al. 2019). For example, the GRCS has two illustrative examples. ‘I have specific rituals and behaviours that increase my chances of winning’ and, ‘A series of losses will provide me with a learning experience that will help me win later.’ A winning poker player might well rationally agree with these statements (Palomäki et al. 2020). Specific rituals and behaviors in live poker include the need to sit at the table in a similar way after looking at one’s cards in order to avoid giving away physical ‘tells’ about the strength of one’s cards (Caro 2003). Additionally, if a poker player can study past losses, then this may reveal strategic insights that can then due to the skill element of poker increase their chances of being successful in the future. Therefore, studying the behaviors and thinking-styles of successful poker professionals could provide insights crucial to the development of items which could better detect harmful engagement and irrational thinking styles among skill-based gamblers.

The study of professional poker players could also contribute to the decision-making literature, in particular the expertise sub-field (Ericsson and Charness 1994). There is a tradition of exploring expertise in chess (Chase and Simon 1973; De Groot 1978; Gobet and Simon 1996), another game known for its professional players and circuits. However, poker and chess are very different games. While the best chess player will win in the majority of matches, hundreds of hours of poker might need to be played for skill to reliably dominate over short-term luck (Potter van Loon et al. 2015). (The number of hours needed for skill to dominate luck in poker depend on the player’s winrate and the volatility of their game; see Chen and Ankenman 2006). Professional and expert gamblers have of course been the subject of previous studies. Some classic studies include Ceci and Liker’s (1986) and Rosecrance’s (1988) studies on expert horseracing bettors, Wagenaar and Keren’s (1985) study of blackjack experts, and Hayano’s (1982) study of professional poker players in a Californian cardroom. However, since those studies, especially in the past two decades, online gambling has profoundly changed the gambling landscape. Gambling online is one of the strongest risk factors for disordered gambling (Allami et al. 2021), a fact that can be observed amongst amateur poker players (Shead et al. 2008; Hopley and Nicki 2010; Mihaylova et al. 2013).

But online gambling can also provide unique opportunities for professionals. Online gambling has transformed the life of professional gamblers by creating a wide array of either lower-cost or higher-frequency gambling opportunities (Cassidy 2013; Schüll 2016). Poker might be the most popular gambling game for online professional gamblers, given that poker is played against other gamblers, meaning that successful poker players do not get their accounts closed down or restricted by online operators, unlike successful sports bettors (Kaunitz et al. 2017).

Poker was also the first gambling game to explode in popularity online. One of the main poker sites, PokerStars, was founded in 2001 (Weston 2022). In 2003 the amateur player Chris Moneymaker parlayed a $86 online tournament buy-in on PokerStars into a $10,000 entry into the televised 2003 World Series of Poker Main Event (Raskin 2014). Moneymaker outlasted many famous poker players in the tournament including Phil Hellmuth (27th), Phil Ivey (10th), and Dan Harrington (7th), and won the $2.5 million first place prize. Between the television exposure, and general media hype around Moneymaker’s amateur status, $86 initial entry, and his last name, online poker became incredibly popular for the remainder of the decade. This interest was reflected economically, and for example in 2005 an online poker operator called PartyGaming floated on the London stock exchange and instantly became one of the UK’s 100 largest listed companies (Reuters 2005). This increase in amateur interest about poker created the conditions for professional players to thrive.

It should be noted that ‘poker’ is not a single game, and a wide range of variants exist with differences over betting types (e.g. limit versus no-limit), formats (e.g. Texas hold ‘em, Omaha, and 7-card stud), events (cash games versus tournaments), stakes (low versus high), and environments (in-person versus online). The mathematics and theory can be very different across variants, as driven by their different betting types, formats, and events. For example, tournaments generally present a fixed maximum loss (the entry fee), but then provide the potential of a win that can be many magnitudes of that entry fee (e.g. turning $10,000 into $2.5 million as Moneymaker did). By comparison, cash games might offer higher chances of smaller wins, and yet also provide the possibility of losses continuing for as long as the player adds money to their stack of chips in play. Players can be experts at one specific combination of factors and complete novices at another. These differences mean that players will be more likely to make mistakes under some combinations of factors than others. This means that specific players will have better and worse chances of profiting in some variants than others. As will be detailed below, we aimed to recruit a sample of experts with a diverse range of specialist expert areas as defined over the factors described in this paragraph.

For example, Phil Hellmuth has a record 16 World Series of Poker in-person tournament wins, and might succeed at this format for example due to his skill at reading other players’ body language (Slepian et al. 2013), a skill which poker theorists emphasize for in-person environments (Caro 2003). Hellmuth has been quoted as summarizing this skill as follows, ‘I seem to look right into people’s souls sometimes. I don’t know what it is’ (Schmidt and Hoppe 2012). However, Hellmuth might expect to lose in the long-run while playing high-stakes poker online against a single opponent with deep expertise of online poker software (Schüll 2016) and knowledge of ‘game theoretic optimal strategy’ (GTO) in that specific format (Bowling et al. 2015). Additionally, a winning player needs to have sufficient money on hand (‘bankroll’), which is determined by the volatility and winrate of their play (Chen and Ankenman 2006; Chin and Ingenoso 2007). Playing a significant number of hands while staying within one’s bankroll could be considered a key poker skill (Schmidt 2009). Winning players also need to control ‘tilt,’ an emotional state where a player is more likely to make mistakes and rash decisions (Palomäki et al. 2013). Having the mental discipline to combat and mitigate tilt (i.e. ‘steam control’) is a skill in which many poker experts have shared their expertise in (Angelo 2007; Tendler and Carter 2011). The poker economy is therefore a complex ecology, where expertise is a matter of the relative skill between each player under the current set of playing conditions. There are therefore many ecological niches for different players to potentially profit from and exist in. The online poker ecology changed significantly in April 2011 when the US Department of Justice seized assets from the three largest online poker operators serving the US market, an event known in poker circles as ‘Black Friday’ (Rose 2011). These actions effectively made online poker illegal in the US, and led to US-based professionals facing the choice of having to move abroad to continue playing online, focus on in-person games, or retire from poker.

Given this cultural significance, prior research has investigated a number of facets of online poker players (Palomäki et al. 2020). Some research has explored online poker communities (Parke and Griffiths 2011; O’Leary and Carroll 2013), or leveraged the large datasets made possible from online poker (Smith et al. 2009; Potter van Loon et al. 2015). Other studies have explored poker expertise via novel tasks (Liley and Rakow 2010; Linnet et al. 2010; Seale and Phelan 2010; St. Germain and Tenenbaum 2011; Linnet et al. 2012; Leonard et al. 2015; Leonard and Williams 2015), self-report scales (Laakasuo et al. 2014; 2015), or via matchups against AI poker agents (MacKay et al. 2014; Moravčík et al. 2017; Brown and Sandholm 2018; Newall 2018; Brown and Sandholm 2019). Other research has explored poker expertise using both qualitative (Bjerg 2011; Jouhki 2011; Radburn and Horsley 2011; McCormack and Griffiths 2012; Recher and Griffiths 2012; Zaman et al. 2014; Talberg 2018; 2019), and quantitative methods (Weinstock et al. 2013; Hopley et al. 2014; Biolcati et al. 2015).

However, as noted before, poker expertise is relative and there are a range of experiences even within the ‘professional gambler’ category. In particular, gamblers at risk of harm might self-report as ‘professionals’ to researchers due to the perceived status of this label (Hing et al. 2015). Many of these self-reported professionals engage in bets with negative expected value such as bets on electronic gambling machines and experience harms from their gambling (Hing et al. 2016), and self-reported poker professionals can experience non-financial harms from the amount of time they spend gambling (Bjerg 2010). We note that previous investigations of poker professionals appear to tap out at maximum self-reported yearly earnings of approximately $150,000 (McCormack and Griffiths 2012; Weinstock et al. 2013; Hopley et al. 2014). Given that these studies were published in the early 2010s, the players in these samples may have struggled to maintain those earnings in the years since, given the degradation in the online poker ecology after the US ban (Rose 2011). With the underlying uncertainty of earnings in professional poker, and the need to pay taxes and living expenses, many of these professionals may now be earning significantly less at poker or have returned to the conventional labor force due to these reduced net earnings from professional poker.

The present study contributes to this literature by performing in-depth qualitative interviews with 19 elite online professional poker experts. While it is important to maintain the anonymity of these participants, we believe that the information presented in the Participants section supports our claim that they represent an elite level of online poker expertise. This is a relatively small community of players, which previous research has highlighted may be reluctant to participate in academic research (Talberg 2019). Our collection of this sample was likely assisted by the first author’s background as a former professional player and longstanding member of the online poker community (Newall 2013), as was mentioned by several of the participants: ‘If it was not you there’s no way I would do this’ (Participant 6); ‘I agreed to speak with you because I have a deep appreciation for your approach and the way that you speak about the game in general’ (Participant 14). Potential weaknesses of this recruitment approach are considered in the Discussion.

Aims

The study used semi-structured interviews to address two aims. The first aim was to explore risk factors from the disordered gambling literature. Specific risk factors explored were their experiences of behavioral dependence and harm (Browne and Rockloff 2019), the presence of irrational gambling beliefs (Raylu and Oei 2004), impulsiveness (Ioannidis et al. 2019), their social networks amongst other gamblers (Russell et al. 2018), the extent to which they engaged in multiple forms of gambling (LaPlante et al. 2014), and the presence of early big wins (Custer and Milt 1985). Exploration of these factors within this group can be of theoretical interest to the gambling psychology literature. For example, if elite professionals tend to show an opposite pattern to disordered gamblers on a factor such as impulsiveness, then this suggests that a reversal of this known risk factor for gambling harm could act as protective factor. The second aim was to find other consistent predictors of success related to the decision-making literature. Any findings here could potentially contribute to the expertise literature (Ericsson and Charness 1994), or also help stimulate new research avenues or treatment approaches for disordered gambling. The final analysis included factors of evaluating decisions based on their expected value (Rubinstein 2011), the extent to which they focused on risks within their ‘circle of competence’ that they knew well (Graham 2003), and their endorsement of an ‘active open-minded’ thinking style (Baron 2008).

Method

Ethical approval for this study was obtained from CQUniversity in Australia and KORUS Øst and Sykehuset Innlandet HF in Norway. The participants received an information sheet prior to the interview and were told that they were free to withdraw at any time, and provided verbal consent at the start of each interview for the interview to be audio recorded. Interviews were conducted from May to November 2021, by both authors remotely via Zoom, and then transcribed by the second author. The interviews lasted between 62 and 124 min (Mean = 87 min).

Participants

Given the present research aims, we recruited a purposive sample of elite online poker professionals (i.e. a non-probabilistic sample). Online poker contains a number of unique ecological niches (which change over time) that players can specialize in, and we recruited experts who are considered by their professional peers to be among the very best at various specialties. While all of these experts began their poker careers online, they also had diverse experiences. Many experts focused their careers around their income from playing, while some experts focused on being paid poker coaches. One expert had coached over 1,000 players, and another offered a coaching service for high-stakes players. The experts were diverse in terms of when they began playing poker, with some experts starting before the Moneymaker boom from 2003. Another expert started playing only after the US ban, and specialized in a form of poker only popularized in the US from 2012 onwards. Experts who were playing throughout the 2000s also showed diverse responses to the US online ban, with one expert who was amongst the most feared high stakes players at his specialty retiring then, while other experts moved overseas to continue playing high stakes online games. In response to the ban, some US-based experts transitioned to in-person games, with some specializing in high stakes ‘mixed’ games which involve several different formats, and others tournaments. Some experts from non-US countries played online continuously up until the present day. Some experts were still playing poker at the time of interview, while others had retired. The retired players generally had sufficient assets from their poker winnings to live as they wished.

There are a number of markers of success in professional poker, such as overall net earnings, tournament earnings, and cash game winnings. However, many of these markers may be difficult to verify. Each year the World Series of Poker runs a number of in-person tournaments generally costing in the range of $1,000 to $50,000 to enter, with the winner of each tournament usually receiving a payoff of usually at least $100,000 and a prestigious ‘bracelet.’ Possession of a bracelet is a prestige marker amongst professional poker players, with the number of bracelets being one way to differentiate in-person tournament players. Up until the 2021 World Series, 268 players had won more than one bracelet since its inception, with 182 of these players winning all of their bracelets from 2000 onwards (Wikipedia 2022). Seven of the 19 experts had won at least one bracelet, and five of these had won more than one bracelet. We are not aware of any previous research on poker experts which has obtained a sample of participants scoring in a similar range on an objective marker of success such as this. This is a remarkably high proportion, as the participants were recruited based on their expertise in online poker. This verifiable and public statistic helps support our claim that these participants represent an elite level of poker skill that has not been represented in the previous academic literature, but without deanonymizing them. This does not mean that the bracelet winners were necessarily the most skilled experts in the sample; many experts may have simply focused on cash games or other tournaments, where their winnings may well have been higher than the bracelet winners’. However, bracelets are a prestige marker in the poker expert community, and this is one shorthand for describing the overall level of expertise in the sample.

Net overall career earnings is a concept that non-players can struggle to understand. As such, professionals, especially those at the elite level, can be reserved about discussing their level of overall earnings. Because of this, in our interviews we generally avoided direct questions about participants’ overall career earnings. However, some participants revealed related information in response to broader questioning. One expert self-reported that his proudest accomplishment was winning $1 million net in a year of online play. We believe, based on the interviews and our own knowledge of these players as members of a small community over a number of years, that participants had average peak annual earnings somewhere in the six-to seven-figure range. There is considerable public data on the earnings of high-stakes online players, so we are withholding further details to protect participants’ anonymity and privacy. Two experts self-reported having losing days of $1 million or more; however, these losses were only temporary and were eventually recouped.

Interview guide

The interviews involved participants broadly narrating their personal histories in poker and gambling, but also included probes consistent with the following risk factors from the disordered gambling literature. As a marker of behavioral dependence (Browne and Rockloff 2019), we asked participants about the proportion of their waking hours spent on poker at their peak. We asked participants whether their gambling led to any harms for themselves or other people in their life (Browne and Rockloff 2019). A question about decision-making planning horizons probed for impulsiveness versus deliberativeness (Ioannidis et al. 2019). We asked participants about memorable early experiences based on the amount of money won or lost, given that early big wins are a risk factor for disordered gambling (Custer and Milt 1985). We asked participants about whether any faulty beliefs about randomness may have affected them (Raylu and Oei 2004). Disordered gamblers tend to engage in multiple gambling formats (LaPlante et al. 2014), so we asked participants about their gambling activities outside of poker, as well as their levels of engagement with different poker formats. Disordered gamblers also tend to have many other disordered gamblers in their social network (Russell et al. 2018), so we asked participants about their networks amongst other poker players. These questions were broadly framed, so as to reduce possible demand effects whereby participants might attempt to guess and provide a ‘correct answer.’

We asked participants about the importance of ‘expected value’ to them. Expected value represents the statistical long-run average return from any gambling situation, and has been emphasized by poker theorists as a rational theory of gambling (Sklansky 1999). For example, a poker player may judge average earnings of $100 an-hour from one game, but only an average of $75 from another game. Joining the first game would be a rational decision, regardless of the profits or losses of the actual session. Expected value has a long history in the field of decision-making, and was emphasized for example by the earliest decision theorists (Rubinstein 2011).

Finally, we asked participants about their attitudes toward investments, and also provided participants to reflect on any topics or issues of personal significance to them which was not covered by the rest of the interview.

Analysis

The present research explores how risk factors from the disordered gambling literature and concepts from the decision-making literature map on to a group of elite online professional poker players. This is a research question which could plausibly be analyzed by one of several analytical approaches, such as for example grounded theory, template analysis, or flexible pattern matching. With grounded theory approaches, theory development begins at the stage of data analysis. But this was not deemed as appropriate here, since the interview guide was based on established literatures among other groups (e.g. disordered gamblers). Template analysis is one way of fitting findings into a priori themes, which would be more appropriate given these research questions which were set before the interview stage began (Brooks et al. 2015).

Ultimately, we decided on an approach called flexible pattern matching approach for analysis (Bouncken et al. 2021). This approach is highly similar to template analysis, to such an extent that template analysis has been termed a type of flexible pattern matching (Bouncken et al. 2021). Overall, flexible pattern matching investigates the congruence of present findings with prior theory. These findings can then be presented flexibly depending on the given research question, and can be described either numerically (e.g. number of cases meeting a predicted pattern), or described narratively. Furthermore, it allows for findings to be developed iteratively, with exploration of the data potentially leading to the discovery of further relevant theories. Given the present work’s aims, it was deemed that the substantial prior literature on a related but distinct group (disordered gamblers) favored the flexible pattern matching approach over for example a grounded theory approach. In terms of the relevant findings outlined in the Aims section, all of the factors from the disordered gambling literature were outlined prior to data collection. The decision-making literature is perhaps broader, and has fewer logical parallels with the presently analyzed group, making it harder to outline relevant theories for matching prior to data collection. The finding around the importance of expected value was outlined prior to data collection, but the findings on the circle of competence (Graham 2003) and active open-minded thinking (Baron 2008) were arrived at inductively after the initial data analysis stage.

After transcription, the two authors began separately analyzing the dataset. The first author’s analysis was based on the above-described aims, while the second author conducted a broader-based thematic analysis, aimed at exploring a broader range of findings from the data (results not reported here). During this time the two authors met to discuss with the other initial themes deriving from their analysis. After an initial round of coding by the first author, the two authors met to discuss potential interpretations of the current findings. This discussion involved both the factors that were determined before the interviews, as well as the two factors which were based from the data. The first author then performed a second round of data analysis, after which the findings were finalized. We note that qualitative methodologists have varying views on the extent to whether coding outcomes should be subject to formal quantitative tests of interrater reliability (Braun and Clarke, 2021). Although no formal tests were performed, the two authors are in agreement on the present study’s findings. The first author’s analysis was conducted in Nvivo software (QSR International Pty Ltd. 2020), and the second author’s coding was conducted in MAXQDA 2022 (Verbi Software 2021).

Findings

Many signs of behavioral dependence, but not harm

Disordered gamblers tend to experience a behavioral dependence to gambling, meaning that they firstly spend a lot of time gambling, might think a lot about gambling during other activities such as spending time with family, and can become upset or irritable if unable to gamble. Participants all reported a peak level of obsession with poker that displayed many similarities to this pattern. They all spent long periods of time playing poker and engaging in related activities, such as studying, browsing online forums, or talking to other professionals. Some participants reported optimizing other areas of their life, such as diet or exercise programs, in order to maximize their performance at poker. In disordered gamblers, losing track of time while gambling is one common occurrence from a behavioral dependence, with for example some electronic gambling machine gamblers betting for many hours on end, until forced to leave by the venue closing or by running out of money. One participant told a story that shows a similar level of dissociation from the schedules of normal life, during a period in which he had moved abroad to continue playing online poker after the US ban:

Me and X were walking back to the coffee shop when we lived in the same apartment building. And we would go and we were walking back and we were like, arguing about which day of the week it was. And he thought it was like a Thursday I thought it was like a Saturday. And it was like Tuesday or Monday: neither of us were even close. (Participant 4)

However, unlike in disordered gamblers, this high level of engagement with poker did not generally lead to participants experiencing harm. Instead, this dedicated focus at developing their expertise, which parallels findings from the broader expertise literature, maximized their opportunities from online poker’s boom years. Many participants were able to smoothly grow small initial deposits into life-changing amounts of money, either through profitable cash game sessions, prizes in tournaments, or a combination of the two. Beyond these initial small amounts, further deposits were often in effect transfers of previous winnings rather than additional funds or players’ ‘own’ money from outside poker. Although four participants did report losing all their poker money on at least one occasion after having spent a significant amount of time playing¹ these experiences were transitory for all except one participant, and were outweighed by for example the level of personal development achieved from poker, e.g. ‘Poker allowed me to grow as a person’ (Participant 1).

This pattern of low levels of harm, in contrast to the disordered gambling literature, also extended to other people in participants’ lives. Disordered gamblers can for example engage in intimate partner violence, most often after losing money. By contrast, participants generally did not harm others with their gambling:

When I came back from playing poker, I didn’t want my partner to be able to tell, you know, either that I'm like, I'm walking on air because I'm up or that I'm like slamming doors because I lost like, I don’t want to create emotional turbulence. Ideally, I'm not experiencing it at all, but I certainly don’t want to bring it home with me. (Participant 8).

A rational approach toward statistical thinking

Participants’ overall positive experiences from gambling may have partly followed from their rational approach toward statistical thinking. Disordered gamblers have many faulty beliefs about randomness, such as the gambler’s fallacy where for example several blacks on the roulette wheel lead to the fallacious thought that red is ‘due’. By comparison, all participants reported being less influenced by these mistaken ideas. Some participants flatly rejected the idea that any faulty beliefs about randomness may have affected their play. Other participants were more open-minded, with some believing that biases such as the gambler’s fallacy represented an inherent bias in the human brain. However, these participants generally also stated that these beliefs did not affect their play, or affected their play much less than other poker players:

I can feel the attractiveness of that thought. But then I know what I'm supposed to do, which is play it the way I normally would, and I play it the way I normally would. (Participant 5).

In part, this avoidance of faulty beliefs may have been assisted by the fact that participants all endorsed an alternative model of statistical thinking with a long history in the decision-making literature: expected value. This is a rational way of thinking about gambling, which has not been subject of as much investigation among disordered gamblers. Focusing on the long-run expected value of a decision allowed participants to judge the quality of their decisions independently of random chance. Some participants reported that this concept helped them even beyond the poker table:

So I think [not] calculating basic expected values like this is one of the mistakes that people in real life, just make a mistake on it everything they do, like from buying insurance to gambling, any pretty much monetary decision they make, they just don’t calculate. (Participant 17).

Expected value can be a difficult concept for people to understand, since it divorces the evaluation of a decision from its outcome. A casual poker player may win a large pot and be happy with their play; a player who understands expected value may contrastingly be dissatisfied by their play if they deduce that the win was not sufficiently statistically likely. A casual player may lose a large pot and be dissatisfied with their play; a player who understands expected value may contrastingly be satisfied with their play if they deduce that statistical wins would be over time likely if that win were repeated enough times.

Members of small expert communities

Disordered gamblers tend to have many other disordered gamblers in their social network. All participants had a similar and yet beneficial tendency to exist in often small, but sometimes extremely close communities of expert players. These people in participants’ professional poker networks were often first met online, either as forum posters, or as players who existed in similar online games. Some participants reported growing to spend time in-person with other professionals that they had met online, e.g. some participants reported moving abroad with other players after the US ban of online poker. These relationships served multiple purposes. One purpose involved the provision of mutual support during the long periods of online poker that could be repetitive, solitary, and frequently frustrating (whichever form of online poker an expert specializes in):

So having people that can help you get out of your own head, I think is tremendously important. And also, that includes being able to sometimes tell someone else, that you’re not feeling great. That actually understands what you’re talking about. Because, as we said earlier, variance² is a very difficult thing to understand. And if you tell someone that doesn’t play poker, and how things are going or how variance is affecting you, they have no clue. But someone that plays poker on a similar or higher level, as you are, or on a slightly lower level even, will have a good understanding. (Participant 1).

For some participants, there were financial aspects to these relationships. For example, a player could sell half of their action in a specific game to someone not playing at that table. The person playing would agree to pay half of the profits if they won. The other person would agree to reimburse half of the net losses if the person playing lost. Arrangements like this might last a single session or continue for months at a time. Oftentimes, these financial relationships were made based purely on trust:

And just trust when people believe you’re trustworthy with money, like you don’t even have to be the best player in the world to say, hey, like, there’s this, you know, pretty big fish playing 1500–3000. Like, I am about so start a session, do you want a piece, you know, I just, I just had enough relationships in poker where I had, you know, 5–10 guys I could, you know, send that text out to pretty quick and get a yes. (Participant 12).

Many participants reported that in the early days of online poker, much information was shared by top players for free on public forums. However, they all reported that as time went on and as poker became more competitive, that this information sharing gravitated toward smaller private networks:

I know that all of the elite players now have these exclusive groups. So like, was a bit like, almost in the same way that like race horses have – what is the term. There’s like a lineage by you know, you want to find race horses that have a connection to a previous race horse from generations ago. I’ve heard this is the case with study groups now. So people were like, a degree removed from X or, or two degrees from X, or that his study group produced a lot of great players. (Participant 10).

This type of information sharing could cross an ethical line if it resulted in collusion: two or more players in the same game sharing holecard information and adapting their play as consequence. This is illegal and unethical behavior in poker circles. We had no indication that any of the participants crossed over this ethical line.

A tendency to not be impulsive

Disordered gamblers tend to be impulsive. By contrast, 15 of the 19 participants reported having a very deliberative approach toward the achievement of their long-term goals. Although these are only self-reports, rather than responses on a validated scale, it does show an overall pattern amongst these people who succeeded in an area with many short-term temptations. While there are many external reasons determining whether a participant might choose to either play online poker or study in order to improve their game, some participants reported substantial investments in study, which could add up to 50% of their overall time spent on poker:

For me, it was always a lot of study, like if you say, 50–50, that’s probably close. (Participant 2).

Online poker games are always available, and it can be tempting to always play poker rather than do things such as studying which have longer-term payoffs. This was summarized well by one participant when reflecting on their approach to their overall career over the years (which involved success in both cash games and tournaments):

I would say I’d say I think I have a fairly long run focus, I think that I'm pretty good at thinking about the big picture for situations and not getting bogged-down in maybe, like random events that might happen, you know, my particular frustration or joy or that one particular moment, but, yeah, like trying to find kind of a nuanced approach to, what will this look like over the long term. I’d say that’s it. That’s generally how my mindset has been, in poker and life. (Participant 16).

However, there was also variation within the sample, with these data meriting further discussion for the remaining four participants. One participant gave an equivocal answer in response to this line of questioning. The remaining three participants reported taking quite a short-term approach to their decisions, consistent with a tendency toward impulsiveness. Each of these three participants reported at least one period where they struggled to control their gambling. For one of these participants, issues only occurred at one point in their career, and coincided with a period of problematic drinking which has now been treated. These issues were strictly resolved, and the participant was able to play professionally (with a career of over 15 years and counting) since then in a controlled manner like the majority of the sample. Another participant reported that he had a history of at times making negative expected value bets on blackjack and sports betting, and that this was an instinctive response that could be triggered by periods of losing. However, this tendency was now controlled by this participant, for example by reducing the easy availability of money for these types of gambling. The last participant with a self-reported tendency toward impulsiveness experienced a great deal of success early on in his poker career, but problems later on, before eventually seeking out treatment for gambling disorder. Since this participant’s experiences differed markedly from the majority of the sample on several (but not all) dimensions, this topic will be explored more fully in the Discussion section. However, this overall pattern, of a skew toward deliberativeness, with the minority of impulsive professionals having a greater risk of experiencing some negatives consequences, does suggest that impulsiveness and its resulting behaviors may be one useful input for predicting a gambler’s outcomes in skill-based gambling.

No clear pattern on memorable early experiences

Disordered gamblers tend to report having a big win early on in their gambling history. In response to a question about memorable experiences early in their poker careers, participants told a range of stories related to both wins and losses. Unlike the previous finding, this finding is not appropriate for a numerical count of participants in different groups, given that participants could report multiple early experiences. Overall, participants reported a mixture of early wins and losses.

Some participants reported intense early losses or mistakenly played hands (such as a folded royal flush – the best possible hand in poker), which might reflect these participants’ competitive motivation to win and rational approach toward achieving this aim:

And, you know, say I was 18–19 and had $50, and I sat down at a 4–8 limit hold ‘em game and promptly lost my money really quickly. And then I put in another $50. And I lost that, and then another $50. And I lost that. And this was an absolute pile of money for me at this stage of my life. Like I've probably borrowed the second and third. You know, I work in the video store Sundays, and this was really bad, but I couldn’t afford to lose any more. I didn’t really know what I was doing… So, you know, I've gone home, and the only thing I immediately resolved to do the next day when I woke up was, let’s learn everything we can about this game, and get our money back next week. (Participant 15).

Other participants reported having memorable wins early on in their careers. Some of these wins came from tournaments, where the top prizes can be large multiples of the initial entry fee. For several of these participants, these early wins were extremely helpful to their careers, as it could either alleviate the need to withdraw living expenses while attempting to grow a bankroll to play poker with, or allow them to immediately play at higher stakes:

And then I remember like, I won, I won like a $10 tournament for like, $3k. And that was like a big win at that point. And I kind of started taking some shots at bigger and starting moving up. (Participant 11).

Early big wins are thought to create risk for disordered gamblers, as they keep returning to try and recreate that win. One participant told an interesting story that was the opposite of this pattern. The participant won a ‘freeroll’ tournament before ever depositing, which provided cash winnings but without having to pay an entry fee. In percentage return terms this is as extreme an early big win as is possible. However, this participant never proceeded to gamble in a problematic fashion, either at the time or later on in his career:

And I ended up winning a free roll tournament for what seemed like an absolutely astounding amount of money, something like over $1,000, maybe $2,000 or so. And then I remember afterwards I was so high on life that I that I went and played $100 buy-in or maybe 200 NL. And I lost. I lost like $400 out of my winnings, maybe I had like $1,400 and I was down to like $1,000, I think it was $1,600–$1,200 or something right? But I never lost such an astounding sum gambling before. I immediately cashed out my entire bankroll and I didn’t play online poker again for many years. (Participant 19).

The circle of competence in risk taking

This finding reflects features relevant to both the disordered gambling and decision-making literatures. Disordered gamblers tend to engage in multiple gambling formats, rather than sticking to just one specific type of gambling. Participants had a clear opposing pattern to this regarding the risks that they took across poker, gambling more broadly, and also in investing.

In poker, participants had a tendency to stick within their chosen specialty format. For example, one player strictly played his chosen specialty for 16 years, and never played other poker formats. While this example was perhaps extreme, another participant reported playing at least 95% of his overall lifetime hands in his chosen speciality. Another participant reported essentially not even knowing the rules of some other poker formats. Some participants changed specialty over time, but this was generally done as a strategic move to stay ahead of changing trends in the popularity of different poker formats. For example, some participants decided to specialize in mixed games, where multiple poker formats are played in turn at a single table. This was also seen as a valid specialty, as nobody could be the very best in the world at so many different poker games.

Participants generally reported a similar reluctance to take too many different risks in gambling more generally. Some participants reported essentially never gambling outside of poker. Others reported occasional low-stakes social gambling, where the small negative expected value of playing blackjack at a casino, for example, was consciously accepted as the price of entertainment, similar to the cost of buying tickets to a show. However, the remainder of their gambling activities outside of these examples were selectively chosen based on considerations of creating expected value. One participant reported regularly betting on sports, but only after deliberatively making this another specialist area of expertise, in response to the changing ecology of poker games. Some others reported gambling on intermittent positive expected value opportunities, for example by betting on Joe Biden to become US president after the votes had been counted in 2020, or positive expected value gambling on a slot machine when a casino offered a reward which exceeded the expected cost of playing (hence creating a profitable gambling opportunity):

I think gambling has never really held much interest for me. I would say that it was 0% of the appeal. I mean, I do have some like early recollections of the excitement of winning like $20 in the poker game or also losing $20 on a poker game. I think that’s why I don’t really enjoy any other form of gambling, the only time I've ever played like blackjack or slots or something is when I've gotten some sort of comp³ that was like, I can only clear that by just mashing the slots button until I did. I didn’t even know what was happening and what I was rooting for, how many cherries I was trying to get in a row. (Participant 8).

This reluctance to take unnecessary risks extended to their investment choices. Expert investors use the term ‘circle of competence’ to describe the universe of risky investments that a given person knows a lot about (Graham 2003). Participants’ descriptions of their own risk-taking in poker, gambling, and investing suggest patterns of staying within their own circles. However, the investments and circles varied across participants. Two participants kept their money within broader gambling-related enterprises; the participant who also specialized in sports betting stated that this was quite a capital-intensive type of gambling, while the other preferred to invest within his specialty of emerging gambling-related opportunities. Several participants invested early on in cryptocurrencies, perhaps because cryptocurrencies helped them to transfer large sums of money after the 2011 US ban. The remainder of participants tended to invest in the stock market in a way that respected the expertise of professional investors. An investor can buy and sell individual stocks as a way to try and beat the market. But several participants stated that their own personal circle of competence lay within poker, and not within beating the stock market. Therefore, those participants tended to invest in the stock market in a passive way, by buying index funds which buy and hold say every stock in the US market (Bogle 2000), and which do not require the exploitation of superior knowledge to get good returns:

That’s not a game that I can that I can reasonably beat and nor is it really worth me playing because I don’t enjoy it. I don’t read company reports 10Ks and things like that. So, why should I deserve any extra returns from picking? So my general philosophy is just to save as much money as possible and invest in the cheapest and the most widely available index funds. (Participant 7).

Active open-minded thinking

Decision-making researchers have highlighted the importance of ‘active open-minded thinking,’ as a decision-making style which avoids making quick and certain conclusions, and which involves updating evidence in support of multiple contrasting possibilities (Baron 2008). Many participants responded to general questions about poker and their career with answers consistent with their thinking style. It is difficult to quantify this proportion, as there were no explicit prompts regarding this topic in the interview guide. However, the following quote captures the essence of this thinking style, and its potential usefulness for a poker player:

Open mindedness is maybe the single most important attribute, just the ability and the flexibility to change my pre-existing beliefs, and to open myself up to learning new ways of thinking. And I think the ability to admit that I'm wrong, and the ability to just be very open about changing my viewpoints is what’s most important, because poker is still evolving, people are still learning about it, you know, you can still be always improving. (Participant 13).

Discussion

Gambling online is a strong risk factor for disordered gambling (Allami et al. 2021). However, online gambling has also provided a number of new opportunities and tools for professional poker players (Cassidy 2013; Schüll 2016). Still, little is known about the psychological profiles of these elite professional online gamblers. The present research contributes to understanding in this area, by analyzing interviews with 19 elite online poker professionals. Some of the risk factors from the disordered gambling literature yielded clear contrasting patterns among this group. While disordered gamblers tend to experience behavioral dependence and harm, most of this group also experienced behavioral dependence, but without harm. While disordered gamblers tend to suffer from faulty gambling beliefs, this group did not. While disordered gamblers tend to have social networks with other disordered gamblers, this group had strong social networks with other experts. While disordered gamblers tend to be impulsive, a majority of this group did not, although with some level of variation that should be subject to further research. While disordered gamblers tend to engage in multiple gambling formats, this group tended to only gamble in gambling formats that they specialized in. Disordered gamblers also often report experiencing early big wins. Interestingly, this group did not display a clear pattern in this area, with some participants experiencing early big wins, and others ruminating on significant early losses. This group also displayed certain themes without clear precedence in the disordered gambling literature. Participants all indicated the importance of evaluating decisions based on the expected value of different options, tended to make investments within their own circle of competence, and displayed a thinking style called active open-minded thinking. These findings can potentially inform the psychology of gambling in several ways.

Some gambling researchers have criticized measures of gambling disorder for mixing items relating to both behavioral dependence and harm, and suggested that harm reduction is more important for policy (Browne and Rockloff 2019). The present results support this perspective with respect to skill-based gambling games, as many participants experienced signatures of behavioral dependence, but not harm. This lack of harm contrasts this group from previous studies of non-elite poker players, amongst whom it has been argued that harm can occur from the loss of time spent gambling (Bjerg 2010; Griffiths et al. 2010). Overall, this suggests that a harm-based approach (Browne et al. 2018) could avoid the measurement issues the PGSI has been argued to suffer from in skill-based games (Laakasuo et al. 2016). Measurements of faulty gambling beliefs have also been argued to lack validity for skill-based games (Russell et al. 2019), and the present results could help researchers to innovate alternative items. For example, participants did not judge their decisions based on their results, and were less concerned about losing than disordered gamblers typically are. Items therefore pertaining to ‘results-based thinking’ could therefore work better, and this is a thinking style that other poker professionals suggest avoiding (Sklansky 1999; Duke 2018). Participants could also be extremely humble about their own level of knowledge in gambling, and had a tendency to always want to improve their understanding, areas which could also yield items able to reliably differentiate them from disordered gamblers.

The present results could also inform other aspects of harm reduction in gambling. One core aspect of current CBT treatments for gambling disorder, which are considered the current gold standard treatment approach (Petry et al. 2017), focus on the correction of fault gambling beliefs. However, some participants in this group reported that they could also find such faulty beliefs intuitively attractive. It may be that their knowledge of rational beliefs around expected value were key to avoiding behavioral manifestations from faulty beliefs, and current gambling treatments could be augmented by placing greater emphasis on teaching disordered gamblers about expected value. In fact, several participants reported that this concept was broadly helpful beyond gambling, and some even said that this concept should be taught to children in school.

The patterns found around impulsivity accord with some previous research amongst poker players (Weinstock et al. 2013), and should be subject to further research, perhaps utilizing a quantitative correlational design on a larger sample of poker professionals. This leads to further discussion of Participant 18’s experiences, who was initially highly successful at poker (hence our contacting him for an interview), but who then experienced significant harm both personally and to his family, before successfully receiving treatment for gambling disorder. This participant had much in common with the rest of the sample, including his rational approach toward statistical thinking and his membership of expert communities. Yet he also showed some differences. Notably, he mentioned difficulties around impulsivity, and issues with substance abuse while playing poker. This participant had a fearlessness about risk, which helped his career initially, but led to later issues which contrasted him from a majority of the sample including: playing poker formats that he was not an expert in, engagement with non-skilled gambling games, and also some bad investments. Further study of larger samples of skill-based gamblers is therefore an important topic for future research.

Limitations

This research was subject to various limitations. Although the first author’s background as an insider to this group helped with recruitment of 19 participants from a small and unique pool, this approach may have led to certain biases. Participants may have tried to provide ‘correct’ answers to certain questions, instead of responding honestly. At the time of interview some professionals had been retired for up to 10 years, and this may have affected the accuracy of some of their answers. Although many attempts were made to recruit females for the study, all participants were male, which does reflect the gender imbalance at high-stakes poker. Many factors may have influenced the types of players willing to participate in the study. For example, theft and scams were at times reported amongst the online poker community, and a player who had suffered from uncontrollable bad luck such as this may have been unwilling to talk about their experiences to researchers. Obtaining as broad a range of experiences as possible should maximize the insights available from this group. Some previous research has collected usable responses to a quantitative survey from a sample of 306 experts from a poker forum (Zhu et al. 2022), and this could be a worthwhile methodology to complement the present findings.

The present study only explored factors with precedent from either the disordered gambling or decision-making literatures, and further study of this group could produce novel insights, or insights relevant to other fields. Finally, we did not collect demographic information from participants, such as their highest level of education obtained or field of study. Some participants discovered poker either at high school or undergraduate level, with some dropping out of education to play poker full-time, with some continued to obtain undergraduate or postgraduate degrees. It is possible that some of the decision-making factors that they endorsed, such as expected value, might only benefit people with above-average levels of quantitative skills.

Conclusions

Increasingly, people who have been in some way harmed by gambling have been contributing to gambling research as members of the ‘lived experience’ community (Ortiz et al. 2021). The present research demonstrates how the lived experiences of professional gamblers could also help lead to a fuller understanding of the psychology of gambling, and opens a number of possibilities for future research that could be facilitated by further engagement from this group, via for example their recruitment for larger-scale quantitative surveys.

Ethical approval

Ethical approval for this study was obtained from CQUniversity in Australia and KORUS Øst and Sykehuset Innlandet HF in Norway.

Disclosure statement

Philip Newall wrote two poker strategy books for Two Plus Two Publishing in 2011 and 2013. Philip Newall is a member of the Advisory Board for Safer Gambling – an advisory group of the Gambling Commission in Great Britain, and in 2020 was a special advisor to the House of Lords Select Committee Enquiry on the Social and Economic Impact of the Gambling Industry. In the last three years Philip Newall has contributed to research projects funded by the Academic Forum for the Study of Gambling, Clean Up Gambling, Gambling Research Australia, NSW Responsible Gambling Fund, and the Victorian Responsible Gambling Foundation. Philip Newall has received open access grant funding from Gambling Research Exchange Ontario. Niri Talberg declares no conflicts.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.

Notes

1 The poker community refers to losing all of one's bankroll as "going busto." This term applies equally to cash game and tournament players.

2 “Variance” is the term used by poker players to describe how short-term results can deviate from a player’s long-term expected value, and can be used in a more colloquial sense than this term’s formal statistical definition (Chen and Ankenman 2006).

3 “Comps” are complimentary items and services given out by casinos to encourage players to gamble, such as free meals, tickets to shows, or even free accommodation to stay at the casino’s hotel.