Why craps players throw dice harder when they want high numbers
In 1975, Harvard psychologist Ellen Langer identified the illusion of control: people consistently act as though they can influence purely random events. Craps players throw dice harder when aiming for high rolls, gamblers wager substantially more money on lottery tickets they picked themselves than on random quick picks, and people mistake chance for skill whenever given personal choice or active physical involvement.
The Craps Table and the Instinct for Agency
Watch a craps table at a casino long enough and a clear physical pattern emerges. When players need a twelve or an eleven, they often rear back and launch the dice with noticeable force against the back wall. When they need a two or a three, their grip softens and they release the dice with a delicate, cautious flick. Mechanically, the velocity of the throw has no bearing on which combination of faces turns up; the tumbling cubes are governed entirely by physics that decouple intent from outcome. Yet the physical gesture reveals an instinctive human assumption that bodily effort can steer random trajectories.
In 1975, psychologist Ellen Langer formalized this phenomenon under the term 'illusion of control.' Langer defined it as an expectancy of personal success probability inappropriately higher than the objective probability would warrant. In purely chance-determined situations, people regularly behave as though their skill, attention, or physical engagement can sway the outcome. Rather than recognizing a sharp boundary between skill-based tasks like archery and chance-based tasks like rolling dice, human cognition often blurs the two, importing habits of effort and control into domains where they have zero mechanical effect.
Langer's Experiments and the Power of Choice
Langer's original experimental work demonstrated that the illusion of control does not require superstitious beliefs; it is reliably triggered by ordinary situational cues typically associated with skill. These cues include personal choice, active involvement, familiarity with the task, and competition. When any of these elements are introduced into a game of pure chance, participants systematically inflate their perceived control over the result.
In one of Langer's most famous field experiments, office workers were offered lottery tickets for a nominal fee. Half of the participants were allowed to select their own ticket from a deck, while the other half were handed a ticket chosen at random by the experimenter. Shortly before the drawing, the experimenter approached each worker and asked how much money it would take to buy the ticket back. The workers who had personally chosen their tickets demanded significantly higher sums to surrender them than those who had simply received an assigned ticket, despite knowing that every ticket possessed the exact same mathematical probability of winning.
Similar effects appeared when researchers introduced elements of familiarity and competition. Participants competing against an opponent who seemed hesitant and unconfident placed higher bets on chance outcomes than those competing against an opponent who appeared assertive and skilled. Introducing choices or physical actions—such as pulling a lever, cutting a deck of cards, or picking a specific symbol—consistently fostered a false sense of mastery over random events.
The Machinery of Contingency and Outcome Density
To understand why this illusion occurs, cognitive psychologists examine how people evaluate contingency—the statistical relationship between an action and an outcome. In laboratory judgment tasks, participants press or refrain from pressing a button and observe whether a light flashes. When the light flashes frequently on its own, regardless of the participant's actions, participants routinely report that their button presses are causing the light to turn on.
This distortion is driven in part by an 'outcome density bias.' Human judgment is disproportionately influenced by positive co-occurrences—instances where an action is taken and the desired event happens immediately afterward. When an outcome is frequent, these pairings occur repeatedly by chance alone. Because human working memory naturally tracks successful hits far more easily than non-events or instances where doing nothing produced the same result, the brain constructs a causal narrative connecting the action to the payoff.
Furthermore, the sequence of outcomes plays a decisive role. When individuals experience early success in a series of chance events, they develop a strong initial hypothesis that they possess an effective strategy. Once established, this belief is sustained through confirmation bias: subsequent wins are credited to personal technique, while subsequent losses are dismissed as bad luck, momentary lapses in concentration, or outside interference.
Depressive Realism and the Baseline Bias
In 1979, psychologists Lauren Alloy and Lyn Yvonne Abramson used contingency judgment tasks to test how mood affects the illusion of control, leading to the concept of 'depressive realism.' In their experiments, non-depressed participants regularly overestimated their degree of control over a light that was programmed to flash randomly without any true contingency. By contrast, participants with mild to moderate depression judged their lack of control with striking accuracy.
Alloy and Abramson suggested that a moderate illusion of control is the default cognitive state for healthy individuals, acting as a psychological buffer that maintains optimism, self-esteem, and motivation. Depressed individuals appeared to lack this self-serving filter, viewing their true agency in an indifferent environment with stark, unembellished realism.
Subsequent research has added nuance to the depressive realism hypothesis. Follow-up studies demonstrated that depressed individuals do not always make more accurate judgments across every context. Instead, non-depressed people tend to show a pronounced illusion of control primarily when outcomes are frequent or desirable, whereas depressed people may exhibit an inverse bias, underestimating their real control when they actually possess it. The broader finding remains clear: healthy human cognition carries a built-in lean toward overestimating personal agency.
High Stakes: Gambling and Financial Markets
While throwing dice harder in a casual game is harmless, the illusion of control has serious consequences in high-stakes environments. Commercial gambling operations frequently design games specifically to exploit skill cues. Modern electronic gaming machines offer players choices over which reels to stop, interactive bonus rounds, and customizable symbols. By replacing passive watching with active participation, these features induce a false sense of control that leads players to wager larger sums and persist longer in the face of losses.
In financial markets, the illusion of control regularly distorts investment decisions. Traders who conduct extensive research, actively monitor price charts, and execute frequent manual trades often develop unwarranted confidence in their ability to time market movements. Studies of individual investors have shown that those who trade most actively—exercising continuous personal choice—frequently underperform simple, passive buy-and-hold strategies due to transaction costs and misjudged market noise.
When an investor experiences a sequence of profitable trades early in their career, the illusion takes deep root. They attribute market-wide gains to analytical skill rather than broad macroeconomic conditions, encouraging excessive leverage and poor risk management that leave them vulnerable to sudden market reversals.
The Evolution of Misplaced Agency
The persistence of the illusion of control across cultures points to its potential adaptive value. In ancestral environments, failing to recognize an existing cause-and-effect relationship could be fatal, whereas falsely assuming a connection carried minimal cost. An individual who mistakenly believes that a specific ritual helped them find food wastes a small amount of energy, but an individual who fails to notice real patterns in animal tracks or seasonal changes misses vital survival opportunities.
Believing that one has agency also sustains goal-directed behavior during adversity. If people accurately perceived how heavily their successes depend on luck, timing, and uncontrollable social variables, they might succumb to helplessness and reduce their overall effort. The illusion of control fuels the persistence required to tackle difficult, uncertain tasks where skill and chance are genuinely intertwined.
The cognitive challenge is recognizing where real agency ends. While personal effort, deliberate practice, and strategic planning dictate performance in skill-based endeavors, they cannot alter the roll of the dice, the draw of a card, or the random fluctuations of complex systems. Distinguishing between environments governed by actionable feedback and those governed by sheer probability remains one of the most critical discernment tasks in human decision-making.
Key takeaways
•The illusion of control occurs when people treat purely random events as though they are responsive to personal skill, effort, or intention.
•Introducing skill-related cues—such as personal choice, active physical participation, task familiarity, or competition—consistently inflates perceived control over chance outcomes.
•Outcome density and positive co-occurrences bias the brain toward inferring causality whenever desired results happen frequently, regardless of actual contingency.
•While the bias can lead to financial losses and excessive risk-taking in gambling and trading, it also serves an adaptive psychological function by maintaining motivation and resilience.