Why Telling Someone What To Do Is the Fastest Way to Make Them Do the Opposite

The Stanley Parable opens with a narrator. His voice is calm, precise, and faintly smug, and he immediately begins telling you what Stanley — you — is going to do. "Stanley got up from his desk and walked through the open door," he says. There is an open door. You are meant to walk through it. And the first thing I did, the very first thing, was stand completely still for thirty seconds staring at my desk chair, then walk in the exact opposite direction, then try to pick up an object that wasn't interactive, then spend four minutes investigating a broom cupboard that the narrator had specifically not mentioned. I did not go through the door. I did everything except go through the door. The narrator, to his considerable credit, had something to say about all of it.

I was not trying to find hidden content. I was not testing the game's systems in any calculated way. I was experiencing something more immediate and less rational: the moment someone told me what to do, I wanted to do anything else. Not because the alternative was better. Not because the door seemed suspicious. Just because the instruction was there, and something in me immediately resented it.

This is not a personality flaw, though it has occasionally presented as one in my working life. It is a documented psychological phenomenon with a name and a substantial body of research behind it, and The Stanley Parable is probably the most precise interactive demonstration of it ever designed.

The Game

The Stanley Parable was originally a Half-Life 2 mod created by Davey Wreden, released in 2011, and then rebuilt as a standalone game by Wreden and William Pugh in 2013. An Ultra Deluxe edition with expanded content arrived in 2022. The premise is deliberately minimal: you play as Stanley, a data entry worker who arrives one day to find his office mysteriously empty. A narrator — voiced with tremendous relish by Kevan Brighting — begins describing what Stanley does. You play. The narrator narrates. The game has over forty possible endings, ranging from poignant to absurd to actively hostile, and most of them are reached by ignoring or subverting the narrator's instructions in one way or another. The broom cupboard has its own ending. Of course it does.

The design achievement is that The Stanley Parable makes resistance the mechanic. It doesn't just tell a story about control and free will — it makes you enact the problem of control and free will through the act of playing. When you follow the narrator, you feel vaguely passive. When you deviate, you feel briefly triumphant, and then complicit, and then, in some of the later endings, genuinely unsettled. The game is structured to make you notice your own responses in real time, which is a remarkable thing for any piece of media to manage, and it does it primarily by finding the exact right target for your reactance and poking it with a long, elegant stick.

Don't Tell Me What To Do — The Research

In 1966, a social psychologist at Duke University named Jack Brehm published a monograph titled "A Theory of Psychological Reactance." His central observation was simple: people place value on their freedom to choose, and when they perceive that freedom to be threatened or eliminated, they experience a motivational state — reactance — that drives them to restore it. Crucially, this restoration isn't necessarily rational. The threatened behaviour doesn't become more objectively appealing; it becomes more subjectively appealing, because the act of choosing it becomes an assertion of freedom. The broom cupboard isn't more interesting than the open door. It's just that nobody told you to go to the broom cupboard.

Brehm's original experiments were elegantly simple. Participants were offered a choice between several consumer goods they'd rated similarly. When the experimenter indicated that one option was unavailable — or, in some versions, simply implied that the participant was expected to choose a particular one — the threatened option's appeal increased substantially. People wanted what they were told they couldn't or shouldn't have, not proportionally to its actual value, but proportionally to the perceived threat to their freedom.

The effect is remarkably robust across contexts. A 2002 meta-analysis by Rhonda Dickson found consistent reactance effects in health communication: messages telling people to exercise, quit smoking, or eat better frequently backfire, particularly when the framing is directive or high-pressure. The message intended to change behaviour actually stiffens resistance to it. Public health campaigns that tell people what they should do — strongly, repeatedly, with an implicit "or else" — can produce the opposite of the intended effect in a meaningful proportion of recipients. This is not because those people are contrary. It's because they're human, and their autonomy radar is extremely sensitive.

The modern application is somewhat uncomfortable if you spend any time in a management role, or as a parent, or as anyone who has ever tried to get another person to change their behaviour through instruction. Telling people what to do triggers reactance. Framing the same desired behaviour as a choice — especially their choice — dramatically reduces it. The research on motivational interviewing, a clinical technique developed by William Miller and Stephen Rollnick for use in addiction counselling, shows that people change behaviour most readily when they feel their autonomy over the choice is acknowledged and respected. The counsellor who says "here's why you should quit" reliably gets more resistance than the one who says "what do you think would happen if you did?"

The Narrator Knows You're Going to Disobey

What makes The Stanley Parable so precise as a psychological artefact is that the narrator anticipates the reactance. He has scripted responses to deviation that are themselves a kind of authority — "Stanley walked past the broom cupboard without a second thought," he says firmly, as you stand inside it — and those responses escalate the reactance further, because now you're not just resisting instruction but being actively narrated into compliance. The game becomes a kind of arms race between the narrator's framing and your desire to assert that you, not him, are the one in charge of what Stanley does.

The deeper design move is that the game eventually turns this on you. Several of the endings involve choices that feel like rebellion but are, in fact, exactly what the game predicted and accounted for. You think you're defying the narrator; you're playing an ending he had ready. The game is arguing, with considerable wit, that even our resistance to authority can be anticipated and contained — that the desire for autonomy is itself a behaviour that can be designed around. This is unsettling in a way that takes a moment to settle in, and it's one of the reasons the game has stayed relevant far longer than a four-hour walking sim has any right to.

What To Do With This

The reactance research has two useful directions, depending on whether you're the one giving instructions or receiving them. If you're receiving: recognising reactance in yourself is more useful than suppressing it. The next time you feel the pull to do the opposite of what you've been told, it's worth pausing to ask whether the resistance is based on the quality of the instruction or purely on the fact that it was given. Sometimes instructions are wrong or worth questioning. Sometimes you're just standing in the broom cupboard because the narrator mentioned the door. Knowing which is which is surprisingly useful information.

If you're giving instructions — to a colleague, a child, a patient, anyone — the reactance literature is fairly clear about what works better: frame the desired behaviour as a choice, acknowledge the person's freedom to decline, and resist the urge to escalate when the first instruction doesn't land. The harder you push, the more the resistance firms up. The narrator who says "you can go through the door, if you want" produces a very different outcome from the one who says "Stanley walked through the door." The Stanley Parable gets at this beautifully, and with considerably more entertainment value than most management textbooks.

Play All the Endings. Yes, All of Them.

The Stanley Parable: Ultra Deluxe takes three to four hours for a first playthrough and considerably longer if you're trying to see everything — which you should be, because the range from funny to genuinely moving is vast and several endings only make sense in the context of others. The game is available on PC and console, it's not expensive, and it is one of the most formally interesting pieces of interactive media made in the last twenty years. The narrator is one of the great video game performances. The broom cupboard is absolutely worth it. And if you find yourself resisting the urge to follow any of the above recommendations simply because I stated them as recommendations, congratulations: you have felt the thing this post is about, and I respect it completely.