Skip Navigation Links

Helping Professionals

An Introduction to Conceptual Models of Problem Gambling

by Darryl Upfold

As a number of authors have pointed out (Lesieur & Rosenthal, 1991), it is probable that sociological, psychological and biological processes are involved in the etiology of problem gambling. Models of problem gambling typically emphasize one of these processes over the others. Three models will be reviewed: the disease model, the social learning model (including the cognitive-behavioural model) and the psychodynamic model.

The Disease/Medical Model

This model proposes that problems with gambling are the result of a disease — a condition that an individual “has” or “does not have.” The model states that compulsive gambling is progressive (it will continue to worsen if the individual continues to gamble), and irreversible (a compulsive gambler will not be able to return to non-problem gambling).

Whether gambling is “compulsive” is assessed by using the GA 20, a questionnaire developed by Gamblers Anonymous (GA).

Rosecrance (1985-86) describes the five major components of the disease model:

  1. There is a single phenomenon that can be labelled “compulsive gambling.”
  2. Compulsive gamblers are qualitatively different from other gamblers.
  3. Compulsive gamblers gradually lose control, and are eventually unable to stop gambling.
  4. Compulsive gambling is a progressive condition, and progresses inexorably through a number of stages, beginning with initial success at gambling; then experiencing less success; irrational optimism about winning; psychological distress; chasing losses and possibly engaging in illegal activities to get money to gamble; unsuccessful attempts to cut down or quit gambling; and eventually hitting bottom.
  5. Compulsive gambling is a permanent and irreversible condition. The only cure is total abstinence. If the gambler were to resume gambling, all of the symptoms described above would manifest again.

The major implication of this model for counsellors is that there is a predetermined need for the compulsive gambler to embrace abstinence as the treatment goal. For a compulsive gambler, any gambling is framed as a relapse that will inevitably lead to compulsive gambling. The disease model also advocates the use of self-help groups (such as Gamblers Anonymous) and maintaining abstinence by dealing with past or hidden underlying psychological problems through therapy. Medications may also be used to treat gambling compulsions and impulsive behaviour and antidepressants are prescribed for gamblers dealing with depression and suicidal tendencies.

Although there is scarce research support for the disease/medical model, there is some support for the role that biological processes may play in compulsive gambling. Although Rosenthal’s typology (1987) referred to physiological or psychobiological models rather than specifying the disease/medical model, the research that is often cited to support physiological or biological models should be considered in discussing the disease/medical model. This body of research has studied the role of EEG brain waves, plasma endorphin levels (linked to arousal), and brain chemical imbalances (Ferris, Wynne & Single, 1998). The direction of this research is to attempt to identify abnormalities that would differentiate compulsive gamblers from other gamblers.

There are two main criticisms of the disease model. The first is that the model does not seem to be able to account for those who may experience problems with gambling from time to time, but who do not progress to the stage of persistent, recurrent compulsive gambling. The second criticism is that the disease model, with its emphasis on internal disease or medical processes, may seem to absolve the compulsive gambler from taking responsibility for the problem. The counter-argument is that even if the compulsive gambler cannot be responsible for having the disease, he or she is responsible for taking action to deal with the disease.

The disease model has clinical utility for those who think of their gambling experience in a way that is highly consistent with the five main components of the model. When this is the case the individual may embrace the theory and practice of the disease model. Alternatively, when the individual seeking help for a gambling problem does not think of his or her problem in the same terms as the model, he or she is likely to resist engaging in treatment, as he or she would (rightfully) conclude that this approach is not “going to work for him or her.”

Therefore, in assessing and developing a treatment plan, attention to the principles of matching the individual to the appropriate intervention is crucial. The counsellor needs to have a thorough understanding of whether the gambler thinks about his or her problem in a way that is consistent with the components of the disease model.

The Social Learning Model

The social learning model views gambling as operant behaviour, subject to reinforcement schedules, which strengthen gambling behaviour. Initially, money was theorized to be the positive reinforcement, but more recent thinking, supported by research, has elevated the importance of the reinforcing properties of physiological arousal in this model.

Interestingly, Skinner (1953) argued that the individual’s gambling behaviour is a function of his or her previous reinforcement history. He theorized that initial success with gambling led to an increased likelihood that the gambling behaviour would continue, even if the reinforcement ratio declined, i.e., the individual won less often. Skinner was able to demonstrate this learned pattern in rats and pigeons. However, his hypothesis was not tested on humans. Later, Custer (1982) emphasized the importance of the early big win in the development of pathological gambling.

Social learning theorists still wonder why people affected by problem gambling tend to persist even when there is very little reinforcement, i.e., they lose far more than they win. Two explanations have been suggested. One, reinforcement schedules that pay off only intermittently (like gambling) are known to produce a greater persistence in the behaviour after the reward is stopped, than would a schedule of continuous reward. This is known as the “partial reinforcement extinction effect” (pree). Another aspect of the pree that accounts for persistence is that persistence does actually pay off in a sense, because the gambler will eventually experience a win if he or she continues to gamble. Thus persistence to gamble, even after many losses, is strengthened by the win (Griffiths, 1995).

The second explanation considers the role of arousal. Brown (1987) proposed arousal as a key determinant of problem gambling, thus including physiological concepts in the realm of the social learning model. According to Brown, individuals have different psychophysiological arousal needs, and some learn to use gambling to regulate their arousal needs. Thus, arousal, not winning, may be the primary reinforcer of the gambling behaviour.

In addition to arousal, Brown (1987) proposed that there are six mechanisms involved in the development of gambling problems:

  1. affective states, like anxiety or depression
  2. cognitive distortions about gambling
  3. behavioural reinforcement schedules
  4. social and institutional determinants, like the opportunity to gamble
  5. subcultural conditions, such as the prevailing attitude toward gambling, and the prevailing values of the individual’s social context and reference groups.
  6. internal fantasy relationships with personifications such as “lady luck,” and the gambler’s parents.

Brown explains that an individual is introduced to gambling as a function of social and institutional determinants and sub-cultural conditions. Continued (and perhaps escalating) gambling depends on behavioural reinforcement schedules and individual psychophysiological arousal needs. Gambling is elevated to a central role in the individual’s life when affective and cognitive factors and internal fantasy relationships are present, as well as the continued influence of reinforcement schedules, arousal needs, social and institutional determinants, and subcultural conditions. Brown’s analysis is clearly broader in explanation than other social learning models.

Because the social learning model takes the position that behaviour is acquired through a complex interaction of a variety of internal and external factors, it suggests that gambling can be conceptualized as a continuum — from problem-free gambling to varying degrees of problem gambling. Further, the development of gambling problems does not follow set stages (although there may be patterns of gambling), nor are serious gambling problems considered permanent and irreversible. The course that problem gambling can take varies from one individual to another.

Counsellors working within the social learning model approach treatment differently from the disease/medical model, preferring to focus on empowerment and solutions rather than assessing for hidden, past or underlying problems. They are more concerned with dealing with the present problems rather than knowing what led up to the problem. From this perspective, goal options vary and may include harm reduction, controlled gambling or abstinence.

Stein (1993) added an interesting perspective to the social learning model from developmental theory, by viewing compulsive gambling as a symptom of cognitive developmental delay in the transition from adolescence to adulthood. The compulsive gambler is seen as having a limited way of understanding his or her environment and developing solutions to his or her problems. As a result, his or her problem-solving strategies — namely, gambling to gain pleasure or diminish pain — are developmentally delayed.

The cognitive-behavioural model is a more specific application of the social learning model.

The Cognitive-Behavioural Model

The cognitive-behavioural model is based on the theory that behaviour is initiated, maintained (or discontinued), based on principles of learning (such as imitation, observational learning, schedules of reinforcement), and cognition (how the individual attends to, interprets, and draws conclusions about events that go on around him or her).

According to Ferris, Wynne and Single (1998), the cognitive-behavioural model described by Sharpe and Tarrier (1993) is the best developed of the social learning models. Sharpe and Tarrier proposed that gambling behaviour is acquired through operant and classical conditioning. Gambling behaviour is reinforced on a partial and variable reinforcement schedule, through a combination of financial rewards and increased physiological arousal levels. The gambler learns that wins will be intermittent, and with persistence, wins will occur.

If the pattern of reinforcement continues, associations may be formed through classical conditioning. The gambling environment becomes associated with increased arousal, and may generalize to other gambling-related stimuli. These stimuli act as triggers for gambling because they have the ability to increase arousal.

Cognitions play an important role in the maintenance of gambling. Gamblers who misinterpret the causes of wins and losses, and develop a belief system that they will be successful at gambling, become more susceptible to problem gambling (Refer to Section 3.8, “Cognitive Therapy for Problem Gambling” for a discussion of cognitive factors).

Whether or not the gambler will gamble when faced with an urge (triggered by exposure to a conditioned stimulus), seems to be related to the gambler’s coping skills — the individual’s ability to control increased arousal, analyze distorted cognitions, delay reinforcement, and apply problem solving skills (Sharpe & Tarrier, 1993). (Refer to Section 3.9, “Teaching Clients Skills for Change and Relapse Prevention.”)

The Psychodynamic Model

The psychodynamic model proposes that personal problems, like excessive gambling, lie within the psyche, and are an attempt at self-healing or a strategy of resolving unconscious psychic conflicts. The source of the conflict is beyond voluntary control. Driven by these conflicts, the individual continues to engage in gambling behaviour in an unsuccessful attempt to ease the psychological pain.

Within the psychodynamic model there are several psychodynamically based theories that have been put forward to explain psychopathological behaviour, including excessive gambling. Three main components are:

  1. Gambling is an unconscious substitute for pre-genital libidinal/aggressive outlets.
  2. Gambling involves an unconscious desire to lose — a wish to be punished in reaction to guilt.
  3. Gambling is a medium for continued enactment (but not resolution) of psychological conflict (Griffiths, 1995, p. 11).

Rosenthal (1986) also claimed that the majority of compulsive gamblers were narcissistic. They have feelings of inadequacy that lead to the creation of a fantasy world in which gambling is seen as the solution to their problems. This fantasy allows them to feel important, respected, powerful and independent, and, in fact, many gamblers report such feelings of power and respect while “in action.” Rosenthal notes that the narcissistic personality is susceptible to swings of arousal and depression, and that gambling behaviour is an attempt to regulate these swings.

Rosenthal also claimed that gamblers are preoccupied with freedom from responsibility, intimacy, and the constraints of consciousness. Defence mechanisms, the psychodynamic equivalent of coping skills, are believed to be primitive in the compulsive gambler. Defence mechanisms prevent exploration and resolution of the unconscious conflicts, allowing the problem to continue.

Another predominant theme in psychodynamic theory is loss, including physical separation such as death of a loved one, loss of job or status, or divorce; or emotional losses, such as loss of self-esteem, or a sense of security. The theory is that pervasive loss is a central issue for the compulsive gambler, resulting in intolerance of future losses. Gambling is both precipitated by an emotional loss, and also a way to avoid taking any further emotional risks. Compulsive gambling is seen as a fixation or regression to the psychic development of a child who is beginning to perceive the loss of a mother’s love. According to this theory, gambling is used as a substitute for the mother’s love.

Critics of the psychodynamic model have pointed out that the model is more theoretical than scientifically based, since concepts like “psyche” and the “unconscious” are largely untestable. Much of the “evidence” for this model is based on single case histories and elaborate conjecture (Ferris, Wynne & Single, 1998).

However, the psychodynamic model has made a contribution by advocating that counsellors understand the internal, unseen processes, in addition to focusing on the gambling behaviour itself. Many gamblers may indeed benefit from counselling that explores the issues of loss, emotional intimacy, and the drive to create the fantasy of importance, respect, power and control that is often central to the gambling experience. For some gamblers, stopping gambling may bring these unconscious conflicts into greater awareness, creating an uncomfortable tension that could be a factor in relapse if left untreated.

Related to the psychodynamic model is the personality, or trait, theory of gambling. In contrast to the more theoretical nature of the psychodynamic model, a great deal of research has been conducted to identify personality factors that may be related to problem gambling (Ferris, Wynne & Single, 1998). Compulsive gambling has been linked to high scores on measures of depression (Blaszczynski & McConaghy, 1988), although it is unclear whether depression is the cause or the outcome of problem gambling.

Research findings have linked compulsive gambling to antisocial personality disorder (Wildman, 1997). For example, there is a relatively consistent finding that studies using the Minnesota Multiphasic Personality Inventory (MMPI) have found pathological gamblers to have elevated scores on the psychopathic deviation (Pd) scale, as well as depression (Taber et al., 1987). Other research has focused on three personality dimensions: sensation seeking, extroversion, and locus of control (Griffiths, 1995). However, Griffiths noted that the research has been inconclusive.

An interesting area of research has suggested that there are different types of traits associated with different types of gambling (Ferris,Wynne & Single, 1998). For example, Selzer (1992) noted that skill gamblers are more likely to have personality disorders, while luck gamblers are more likely to have affective disorders.

Although the value of personality studies remains doubtful, and the notion that pathological gamblers possess a unique set of variables or traits overly simplistic (Knapp & Lech, 1987), enough research has been done to suggest that counsellors need to be familiar with personality disorders, and to be able to screen for affective disorders such as depression and anxiety.


Bandura, Albert. (1977). Social Learning Theory. Prentice-Hall, NJ: Englewood Cliffs.

Blaszczynski, A. & McConaghy, N. (1988). scl-90 assessed psychopathology in pathological gamblers. Psychological Reports, 62 (2), 547-552.

Blume, S. (1988). Compulsive gambling and the medical model. Journal of Gambling Behaviour, 3 (4), 237-247.

Brown, R. (1987). Gambling addictions, arousal, and an affective/decision-making explanation of behavioural reversions or relapses. International Journal of the Addictions. 22 (11), 1053-1067.

Custer, R. (1984). Profile of the pathological gambler. Journal of Clinical Psychiatry, 45 (12, Sect 2), 35-38.

Ferris, J.,Wynne, H. & Single, E. (1999). Measuring Problem Gambling in Canada: Final Report – Phase 1. For the Inter-Provincial Task Force on Problem Gambling, Submitted on behalf of the ccsa, April 22, 1999. http://www.ccsa.ca/

Griffiths, M. (1995). Towards a risk factor model of fruit machine addiction: A brief note. Journal of Gambling Studies, 11 (3), 343-346.

Knapp, T. & Lech, B. (1987). Pathological gambling: A review with recommendations. Advances in Behaviour Research and Therapy, 9 (1), 21-49.

Lesieur, H. & Rosenthal, R. (1991). Pathological gambling: A review of the literature (prepared for the American Psychiatric Association Task Force on DSM-IV Committee on Disorders of Impulse Control not elsewhere classified). Journal of Gambling Studies, 7 (1), 5-39.

Rosecrance, J. (1985-1986). “The next best thing”: A study of problem gambling. International Journal of the Addictions, 20 (11-12), 1727-1739.

Rosenthal, R. (1987). The psychodynamics of pathological gambling: A review of the literature. In Galski, Thomas, et al. (Eds.). The Handbook of Pathological Gambling (pp. 41-70). Springfield, IL: Charles C Thomas.

Rosenthal, R. (1986). The pathological gambler’s system for self-deception. Journal of Gambling Behaviour, 2 (2), 108-120.

Selzer, J. (1992). Borderline omnipotence in pathological gambling. Archives of Psychiatric Nursing, 6 (4), 215-218.

Sharpe, L. & Tarrier, N. (1993). Towards a cognitive-behavioural theory of problem gambling. British Journal of Psychiatry, 162, 407-412.

Skinner, B.F. (1953). Science and Human Behaviour. New York: Macmillan.

Steel, Z. & Blaszczynski, A. (1998). Impulsivity, personality disorders and pathological gambling severity. Addiction, 93 (6), 895-905.

Stein, S. (1989). A developmental approach to understanding compulsive gambling behaviour. In H. Shaffer (Ed.), Compulsive Gambling: Theory, Research, and Practice. Lexington, MA: Lexington Books/D. C. Heath and Company.

Taber, J., McCormick, R., Russo, A., Adkins, B., et. al. (1987). Follow-up of pathological gamblers after treatment. American Journal of Psychiatry, 144 (6), 757-761.


Back to Helping the Person Affected by Problem Gambling


DISCLAIMER: Information on this site is not to be used for diagnosis, treatment or referral services. CAMH does not provide diagnostic, treatment or referral services through the Internet.
CAMH accepts no responsibility for such use. Individuals should contact their personal physician, and/or their local addiction or mental health agency regarding any such services.
Technical enquiries: webmaster@problemgambling.ca