Blog Archives

The beliefs are falling: The illusion of control in gambling

For the professional gambler, being in control of the situation is paramount. However, one of the psychological ploys that the gaming industry likes to exploit is the fact that gamblers often perceive they have more control than they have in actuality. Probably one of the most single influential contributions to the psychology of gambling was Ellen Langer’s series of experiments on the illusion of control in the 1970s. Her theories were based on the observations that some people treat chance events as controllable. For instance, it isn’t uncommon for dealers and croupiers who experience runs of bad luck to lose their job. Back in the 1960s, sociologists studying craps players noticed those throwing the dice behaved as if they were controlling the outcome of the toss. Typically, craps players threw the dice softly for low numbers and hard for high numbers. In a follow-up experimental investigation, psychologists showed that when playing with dice, people bet less money and were less confident if asked to bet after someone else had thrown the dice rather than throwing it themselves, even though the probability of success was the same in both situations. Ellen Langer argued that these behaviours are totally rational if gamblers believed their game was a game of skill.

The “illusion of control” was defined by Langer as being “an expectancy of a personal success inappropriately higher than the objective probability would warrant.” Put simply, gamblers think they have more chance of winning than they actually do. She tested for this in a series of experimental studies that supported her original idea (that under some circumstances, gamblers will produce skill orientations towards chance events). Langer’s experiments convincingly showed that players bet more when playing cards against a ‘nervous’ competitor than against a ‘confident’ one. She also demonstrated that players would sell previously bought lottery tickets for a higher price if they had picked the numbers themselves as opposed to having them picked by someone else. Her other groundbreaking experiments showed that certain factors such as the nature of the competition, the familiarity of the task, and the degree of personal involvement influence the belief that skill is a controlling force, stimulates the illusion of control, and produces skill orientations. In a later study involving the prediction of ‘heads’ or ‘tails’ after a coin was tossed, she also showed that early wins during chance games induced a skill orientation even though the activity was totally chance determined.

Many regular gamblers (such as roulette players) passionately believe their game is skill-based, and offer explanations of why they failed to win when their number doesn’t come up. Such beliefs have been tested experimentally by US psychologist Thomas Gilovich in a study of the biased evaluations in gambling behaviour. In three studies using people who bet on football games, Gilovich demonstrated that gamblers transformed their losses into ‘near wins‘. Gamblers pinpointed random or ‘fluke’ events that contributed to a loss but were unaffected by identical events that contributed to a win. I’m sure you can all think of instances like this when watching football. When your team loses, it’s not uncommon to berate the referee for a dodgy penalty decision or deride the linesman because he failed to spot an offside. You may end up blaming your team’s loss on one particular event. Had your team won with the dodgy decision going your team’s way, you would probably rationalise it and say your team would have won anyway because of their superior playing ability and skill. Gilovich also reported that gamblers spent more time discussing their losses and discounting them. For example, after a loss, a lot of time may be spent analysing a small incident of a few seconds duration even though the game lasted 90 minutes. What’s more, we make ourselves feel better by blaming the loss on something or someone external. Interestingly, exactly the same effects have been found in gambling activities in which losses could not easily be explained away (such as Gilovich’s experiments using computerised bingo gambling).

Many psychologists have consistently highlighted the irrational perceptions people produce while gambling. Many studies have evaluated the cognitive activities of gamblers while they play on slot machines or roulette using the ‘thinking aloud’ method. This basically involves getting gamblers to think aloud while they are gambling. Typical results have shown that erroneous and irrational perceptions of the gambling activity far outnumber the logical and rational perceptions. In these situations, gamblers attribute their success to personal factors such as skill whereas external factors (like bad luck) account for losses. For instance, in my own research on slot machine gamblers, I found that when slots players were winning they would attribute their success to their playing strategy and skill. When they lost it was because of something external in the gambling environment. For example, someone had put them off by talking to them or watching them while gambling. Similar findings have reproduced by psychological experiments in Canada, Australia and the USA (including some of my own). The illusion of control is just one of the many ways in which a gambler distorts the perceptions of their gambling. These are sufficient enough to show that psychological factors can influence the way in which people gamble and continue to gamble.

Dr Mark Griffiths, Professor of Gambling Studies, International Gaming Research Unit, Nottingham Trent University, Nottingham, UK

Further reading

Coventry, K. & Norman, A. (1998). Arousal, erroneous verbalizations and the illusion of control during a computer generated task. British Journal of Psychology, 89, 629-645.

Gilovich, T. (1983). Biased evaluation and persistence in gambling. Journal of Personality and Social Psychology, 44, 1110-1126.

Gilovich, T. & Douglas, C. (1986). Biased evaluations of randomly determined gambling outcomes. Journal of Experimental Social Psychology, 22, 228-241.

Griffiths, M.D. (1994). The role of cognitive bias and skill in fruit machine gambling. British Journal of Psychology, 85, 351-369.

Griffiths, M.D. (2011). Gambling, luck and superstition: A brief psychological overview. Casino and Gaming International, 7(2), 75-80.

Griffiths, M.D. (2013). Is ‘loss of control’ always a consequence of addiction? Frontiers in Psychiatry, 4, 36. doi: 10.3389/fpsyt.2013.00036

Griffiths, M.D. & Wood, R.T.A. (2001). The psychology of lottery gambling. International Gambling Studies, 1, 27-44.

Henslin, J. (1967) Craps and magic. American Journal of Sociology, 73, 316-330.

Langer, E. J. (1975). The illusion of control. Journal of Personality and Social Psychology, 32, 311-328.

Langer, E.J. & Roth, J. (1975). The effect of sequence outcome in a chance task on the illusion of control. Journal of Personality and Social Psychology, 32, 951-955.

Rogers, P. (1988). The cognitive psychology of lottery gambling: a theoretical review. Journal of Gambling Studies, 14, 111-134.

Rogers, P. & Webley, P. (2001). It could be us! Cognitive and social psychological factors in UK National Lottery play. Applied Psychology: An International Review, 50, 181-199.

Taylor, S. E. (1989). Positive illusions: Creative self-deception and the healthy mind. New York: Basic Books.

Wagenaar, W. A. (1988). Paradoxes in Gambling Behaviour. London: Erlbaum.

Problem thinking: Heuristics and cognitive bias in gambling

Once when I was playing roulette at my local casino, there was a run of seven reds in a row. I rarely bet by colour myself, but while I was laying my many 50 pence chips all over the roulette number grid, I told a friend standing next to me that many people would put a lot of their chips on black on the next spin of the wheel. And they did! I am no mind reader but what I do know about is the gambler’s fallacy. The gambler’s fallacy is a well-known psychological ‘rule of thumb’ where gamblers apply the law of averages to very small number sequences. Put very simply, I knew that most people would be thinking “by the law of averages the black is supposed to come up 50% of the time and hasn’t done so for the last seven spins”. While the 50% probability is true, the probability is based on very large sequences of numbers and not a few spins of the roulette wheel. What’s more, the roulette ball has no memory of where it landed before and every spin is independent of the last one. As it turned out, red came up again and there were some disgruntled and disbelieving gamblers. On the ninth spin, a black number finally came up.

The gambler’s fallacy is one of many psychological thinking patterns that are known as ‘heuristics’ (and sometimes called cognitive biases). The psychological effect of heuristics is to reduce uncertainty for the gambler. Open up any textbook on gambling and you will find that the gambler’s fallacy is referred to as the ‘representativeness bias’. This is because people expect to find a representative relationship between samples drawn from a small number of events (for example, eight spins on the roulette wheel), and the complete set of events (in this case, all the spins ever on all roulette wheels). When we gamble, we constantly process information (often unconsciously) in a consistently biased way. Humans tend to exhibit consistent biases when cognitively processing information in gambling situations. For instance, in psychological gambling experiments where people are asked to create a random sequence of imaginary coin tosses, they tend to produce sequences where the proportion of heads in a short segment is closer to 50% than chance would predict.

Over the last 35 years, psychologists have written about many different heuristics that gamblers use. One of the better known ones is the ‘availability bias’. This occurs when a person evaluating the probability of a chance event makes the judgement in terms of the ease with which relevant instances or associations come to mind. For instance, pools winners are highly publicised to invoke the idea that big wins are regular and commonplace when in fact they are rare. Availability biases can also be found when people actually gamble in lotteries. For instance, when selecting numbers, some people will pick (‘hot’) numbers that have come up more often and avoid the (‘cold’) numbers that by chance have not come up as often. For instance, during the week of the first ever triple rollover on the UK National Lottery it was noted by a number of newspapers that the number ‘13’ had come up much less than any other number in almost 10 years of lottery draws. Those gamblers prone to the ‘availability bias’ would be unlikely to pick this number. Of course, those prone to the ‘representativeness bias’ would be more likely to pick it! And that is one of the problems with ‘rules of thumb’ – it is almost impossible to know which heuristic will be applied in a given situation and it is quite possible for the same person to use a different heuristic in the same situation on different occasions.

Some of my favourite heuristics are those involving ‘illusory correlations’. These are superstitious behaviours where people believe two actions are related when in fact they are not. For instance, one seminal 1960s study of ‘craps’ players in US casinos (published by Dr. J. Henslin in the American Journal of Sociology) showed that players rolled the dice softly to get low numbers and rolled harder for higher ones. Other spurious examples are those people who have ritualised routines before they gamble, have ‘lucky chairs’ at the bingo hall, or those who carry lucky charms when they gamble. Most of these illusory correlations start by associative accident. For instance, a gambler might have three big wins at the roulette table and then notice that on all three of those occasions they wore the same pair of trousers. As a consequence, they might start to think that the trousers are somehow lucky and wear them on subsequent visits to the casino. When they win while wearing them, it bolsters the bias. The relationship between the winning and the trousers wearing is illusory but many gamblers display such irrational biases.

Psychological biases provide some insight into why some gamblers don’t learn from past losses and helps explain supposedly ‘irrational’ behaviour in the gambling process. Some psychologists claim that problem gambling is caused by defective reasoning, rather than personality traits, education or social environment.  They also claim that gamblers gamble, not because they have a bigger repertoire of heuristics, but because they select heuristics on the wrong occasions.

Dr Mark Griffiths, Professor of Gambling Studies, International Gaming Research Unit, Nottingham Trent University, Nottingham, UK

Further reading

Griffiths, M.D. (1994). The role of cognitive bias and skill in fruit machine gambling. British Journal of Psychology, 85, 351-369.

Griffiths, M.D. (1997). Selling hope: The psychology of the National Lottery. Psychology Review, 4, 26-30.

Griffiths, M.D. & Bingham, C. (2002). Bingo playing in the UK: The influence of demographic factors on play.  International Gambling Studies, 2, 51-60.

Griffiths, M.D. & Wood, R.T.A. (2001). The psychology of lottery gambling. International Gambling Studies, 1, 27-44.

Henslin, J. (1967) Craps and magic. American Journal of Sociology, 73, 316-330.

Kahneman, D. & Tversky, A. (1973). Availability: A heuristic for judging frequency and probability. Cognitive Psychology, 5, 207-233.

Langer, E.J. (1975). The illusion of control. Journal of Personality and Social Psychology, 32, 311-328.

Langer, E.J. & Roth, J. (1975). The effect of sequence outcome in a chance task on the illusion of control. Journal of Personality and Social Psychology, 32, 951-955.

Parke, J., Griffiths, M.D. & Parke, A. (2007). Positive thinking among slot machine gamblers: A case of maladaptive coping? International Journal of Mental Health and Addiction, 5, 39-52.

Tversky, A. & Kahneman, D. (1971). Belief in the law of small numbers. Psychological Bulletin, 76, 105-110.

Walker, M.B. (1992). The Psychology of Gambling. Pergamon, Oxford.

Wagenaar, W. (1988). Paradoxes of Gambling Behaviour. Erlbaum, London.

Play’s cool? Is the type of game played important in the development of gambling addictions?

Earlier today, I (and my research colleague Michael Auer) had a paper published in the journal Frontiers in Psychology arguing that the type of game that people gamble on is irrelevant in the acquisition, development, and maintenance of pathological gambling. We noted that anyone coming into the gambling studies field from a psychological perspective would probably conclude from reading the literature that problem and pathological gambling is associated with particular game types. More specifically, there appears to be a line of thinking in the gambling studies field that casino-type games (and particularly slot machines) are more likely to be associated with problem gambling than lottery-type games.

We argued that the most important factors along with individual susceptibility and risk factors of the individual gambler are the structural characteristics relating to the speed and frequency of the game (and more specifically event frequency, bet frequency, event duration and payout interval) rather than the type of game. Event frequency refers to the number of events that are available for betting and gambling within any given time period. For example, a lottery draw may occur once a week but a slot machine may allow 15 chances to gamble inside one minute. In this example, slot machine gambling has a higher event frequency than lottery gambling. Bet frequency refers to the number of bets or gambles placed in any given time period. Using lottery playing as example, Dr. Jonathan Parke and I noted in a 2007 book chapter on structural characteristics, that multiple tickets (e.g., 10 tickets) can usually be purchased as frequently as desired before any single lottery draw. In this instance, bet frequency would be equal to 10 but event frequency would be equal to 1. Therefore, event frequency can often be much lower than bet frequency and it is possible for players to spend more than they can afford even with a low event frequency.

Dr. Parke and I have stated that further empirical research is needed into the relationship between event frequency and bet frequency. This is because researchers often assume that event frequency and bet frequency have a strong relationship (i.e., the higher number of betting/gambling events – the higher the frequency of betting/gambling). However, this may not be the case.

Another important gaming parameter is event duration. This refers to how fast the event in question is (e.g., a reel spin on a slot machine might last three seconds). Here, it is important to note that duration of the betting/gambling event is different from event frequency (although they may be inextricably linked in so much as the length of a betting event will obviously limit the frequency with which they can take place). Again, Dr Parke and I noted that a betting event lasting two hours (e.g., a soccer game) could not have an event frequency greater than one in any 2-hour period but could have a betting frequency of over 100 with the advent of in-play betting.

In-play betting and gambling (which I examined in a previous blog) refers to the wagering on an event that has started but has not yet finished. This means gamblers can continue to bet on an event (e.g., a soccer or cricket match) and perhaps more importantly, adapt their bets according to how the event is progressing.  For instance, in the UK, during the playing of almost any soccer match, a gambler can bet on everything from who is going to score the first goal, what the score will be after 30 minutes of play, how many yellow cards will be given during the game and/or in what minute of the second half will the first free kick be awarded. What I argued in a previous blog is that ‘in-play’ gambling activities have taken what was traditionally a discontinuous form of gambling – where a gambler made one bet every weekend on the result of the game – to one where a player can gamble continuously again and again. In short, the same game has been turned from what was a low event frequency gambling activity into a potentially high frequency one (and gone from an activity that had little association with problem gambling to one where problem gambling is far more likely among excessive in-play gamblers).

Another important (and related) structural characteristic is payout interval. This is the time between the end of the betting event (i.e., the outcome of the gamble) and the winning payment (if there is one). The frequency of playing when linked with two other factors – the result of the gamble (win or loss) and the actual time until winnings are received – exploits the psychological principles of learning. This process of operant conditioning conditions habits by rewarding (i.e., reinforcing) behaviour (i.e., through presentation of a reward such as money). To produce high rates of response, those schedules which present rewards intermittently (random and variable ratio schedules) have shown to be most effective. Since a number of gambling activities (most notable slot machines) operate on random and variable ratio schedules it is unsurprising that excessive gambling can occur.

To highlight the irrelevance of game type, consider the following two examples that demonstrate that it is the structural characteristics rather than the game type that is critical in the acquisition, development and maintenance of problem and pathological gambling for those who are vulnerable and/or susceptible. A “safe” slot machine could be designed in which no-one would ever develop a gambling problem. The simplest way to do this would be to ensure that whoever was playing the machine could not press the ‘play button’ or pull the lever more than once a week. An enforced structural characteristic of an event frequency of once a week would almost guarantee that players could not develop a gambling problem. Alternatively, a problematic form of lottery could be designed where instead of the draw taking place weekly, bi-weekly or daily, it would be designed to take place once every few minutes. Such an example is not hypothetical and resembles lottery games that already exist in the form of rapid-draw lottery games like keno.

The general rule is that the higher the event frequency, the more likely it is that the gambling activity will cause problems for the individual (particularly if the individual is susceptible and vulnerable). Problem and pathological gambling are essentially about rewards, and the speed and frequency of those rewards. Almost any game could be designed to either have high event frequencies or low event frequencies. Therefore, the more potential rewards there are, the more problematic and addictive an activity is likely to be and this is irrespective of game type as games such as diverse as lotteries and slot machines could have identical event frequencies and event durations.

Dr Mark Griffiths, Professor of Gambling Studies, International Gaming Research Unit, Nottingham Trent University, Nottingham, UK

Further reading

Griffiths, M.D. (1993). Fruit machine gambling: The importance of structural characteristics. Journal of Gambling Studies, 9, 101-120.

Griffiths, M.D. (1994). The role of cognitive bias and skill in fruit machine gambling. British Journal of Psychology, 85, 351-369.

Griffiths, M.D. (1999). The psychology of the near miss (revisited): A comment on Delfabbro and Winefield. British Journal of Psychology, 90, 441-445.

Griffiths, M.D. (2008). Impact of high stake, high prize gaming machines on problem gaming. Birmingham: Gambling Commission.

Griffiths, M.D. (2012). Mind games (A brief psychosocial overview of in-play betting. i-Gaming Business Affiliate, June/July, 44.

Griffiths, M.D. & Auer, M. (2013). The irrelevancy of game-type in the acquisition, development and maintenance of problem gambling. Frontiers in Psychology, 3, 621. doi: 10.3389/fpsyg.2012.00621.

Griffiths, M.D. & Wood, R.T.A. (2001). The psychology of lottery gambling. International Gambling Studies, 1, 27-44.

Meyer, G., Hayer, T. & Griffiths, M.D. (2009). Problem Gaming in Europe: Challenges, Prevention, and Interventions. New York: Springer.

Parke, J. & Griffiths, M.D. (2006). The psychology of the fruit machine: The role of structural characteristics (revisited). International Journal of Mental Health and Addiction, 4, 151-179.

Parke, J. & Griffiths, M.D. (2007). The role of structural characteristics in gambling.  In G. Smith, D. Hodgins & R. Williams (Eds.), Research and Measurement Issues in Gambling Studies. pp.211-243. New York: Elsevier.