The $8.67 vs. $1.96 Lottery Ticket

In 1975, Ellen Langer ran a lottery experiment. Some participants chose their own ticket. Others had a ticket assigned to them. Then she asked everyone what price they'd accept to sell it back.

People who chose their ticket demanded $8.67. People who were assigned a ticket accepted $1.96.

Same lottery. Same odds. Same ticket value. But the act of choosing made people value their ticket more than four times as much.

Langer called this the "illusion of control" — a skill-chance confusion where factors associated with skill (choice, involvement, familiarity, competition) make people believe they can influence random outcomes.

The effect extended beyond lotteries. Casino craps players bet more when personally throwing the dice. People threw harder when they needed high numbers, softer for low numbers. Active involvement made randomness feel controllable.

This became one of the most cited findings in judgment and decision-making. And then the replication attempts began.

The Meta-Analysis: d = .62

Stefan and David (2013) published an updated meta-analysis of illusion of control experiments. The overall weighted mean effect size was D = .62 (CI: .49 to .75), slightly lower than the .68 reported by Presson and Benassi in 1996.

Moderate to large effects, consistently present across different experimental paradigms. Motivation for outcomes produced large effects when estimating success (d = .86) and skill (d = .83), with moderate effects for control estimations (d = .51).

The illusion of control looked like a solid, well-replicated finding. Until someone tested the specific mechanism.

17 Studies, 10,825 Participants, No Effect

Klusowski, Small, and Simmons (2021) ran 17 preregistered experiments with 10,825 participants — online and in laboratories. Their goal: test whether choice specifically causes an illusion of control.

Their finding: "Choice rarely makes people feel more likely to achieve preferable outcomes."

When effects did appear, they weren't caused by choice creating a new illusion. They reflected participants' preexisting beliefs. People who already believed they could influence outcomes sought out choice. Choice didn't create the belief.

As Klusowski noted: "As we started running studies to verify that having a choice in a lottery makes us believe we're more likely to win, we weren't replicating the effects."

This is a meaningful distinction. The illusion of control exists as a general phenomenon — the meta-analytic evidence supports that. But the specific idea that the act of choosing creates the illusion? That's on shakier ground.

2024: The Plot Thickens

Two 2024 studies added nuance. First, researchers replicated the "Henslin effect" — showing that active physical involvement (not just mental choice) can create a sense of control. The effect was real but relatively small, requiring a large sample (N = 1,692) to detect reliably.

Second, a neural imaging study confirmed that people show different reward-processing patterns when they choose a gambling option versus having one assigned. The illusion of control has a neural signature. But the effect may be more about physical agency than abstract choice.

The emerging picture: people feel more control when they physically DO something, not just when they mentally pick something. Throwing the dice matters more than choosing the ticket.

The Deeper Finding: People WANT Control Regardless

Here's where it gets interesting. While the illusion of control has replication problems, a different finding is rock-solid: people prefer having choice even when it doesn't help them.

Botti and Iyengar (2004) demonstrated that people prefer choosing at the cost of subsequent satisfaction. When people choose among all bad options, they feel worse than when the same bad option is imposed on them. The act of choosing makes them feel more responsible for negative outcomes.

Yet they still prefer to choose.

Leotti, Iyengar, and Ochsner (2010) went further in "Born to Choose." They found that animals and humans prefer options that lead to further choice, even when the expected value is identical and choosing requires more energy. In economic terms, this preference for control is irrational. In biological terms, it's deeply hardwired.

Botti's 2023 paper in the Journal of Consumer Psychology formalized this as the "choice is better" heuristic: consumers prefer more choice freedom regardless of whether it actually makes them better off. This operates as an automatic heuristic, not a rational evaluation.

The Choice Paradox at Scale

A cross-national analysis of 63 countries found that choice freedom predicts subjective well-being more strongly than wealth. But here's the paradox: increased choice freedom is also associated with greater depression and stress.

People want choice. Choice makes them feel alive and autonomous. And choice also exhausts them, burdens them with responsibility, and creates opportunities for regret.

Botti and Iyengar's "tragic choices" research on parents deciding about infant life support illustrates the extreme: perceived personal causality for devastating decisions generates more negative emotions than having those decisions made by others. Yet people resist giving up their right to decide — even when deciding destroys them.

What This Actually Means

The original illusion of control story was simple: give people choice and they'll think they control the outcome. The real story is more complex and more useful.

People don't necessarily believe choice gives them more control (that effect is weaker than we thought). But people prefer having choice anyway — even at a cost. This isn't an illusion. It's a need.

The practical implications diverge from the original narrative:

Don't assume choice creates confidence. The Klusowski et al. replication failure suggests you can't make people believe your product will work just by letting them choose options.

Do assume people need to feel autonomous. The Botti and Iyengar research consistently shows that agency matters independently of outcome. Let people feel like they're driving.

Structure the choice. The paradox of choice research (too many options overwhelm) combined with the autonomy preference (people want options) means the sweet spot is 2-3 meaningful choices at each decision point. Not zero. Not twenty.

Frame actions as chosen. "You chose to explore this feature" feels different from "Here's what's next on your tour." The content can be identical. The framing of agency changes the experience.

Respect the cost of choosing. If your onboarding forces people to make too many decisions, you're taxing the very autonomy they value. Meaningful choices about direction, not trivial choices about configuration.

The Design Principle

Don't try to create an illusion of control. The science says that doesn't work reliably.

Instead, offer genuine autonomy within structured boundaries. Let people choose their path, not just their options. Make the choosing easy (fluency matters here too) and the consequences clear.

People don't need to believe they control everything. They just need to feel like the choices that matter are theirs to make.

---

This post is part of the MIFGE research series. Based on Langer (1975), Stefan & David (2013) meta-analysis (D = .62), Klusowski, Small & Simmons (2021, 17 studies, N = 10,825), Botti & Iyengar (2004, 2006), Botti (2023), and Leotti, Iyengar & Ochsner (2010).