Anchoring Effect Simulator: How Initial Values Bias Judgment

simulator beginner ~6 min
Loading simulation...
Mean Estimate: ≈ 260 (true value: 500, anchor: 100)

With an anchor of 100, a true value of 500, and an adjustment factor of 0.4, the average estimate lands around 260 — only 40% of the way from the anchor to the truth. This 48% bias demonstrates how even arbitrary anchors create systematic distortion in judgment. The histogram shows the full population clustered far from the true value.

Formula

estimate = anchor + α · (true_value - anchor) + ε
ε ~ N(0, σ²) (individual noise)
anchor_bias = |mean_estimate - true_value| / true_value × 100%

What Is the Anchoring Effect?

The anchoring effect is one of the most powerful and pervasive cognitive biases. First demonstrated by Amos Tversky and Daniel Kahneman in 1974, it describes how an initial piece of information — the anchor — disproportionately influences subsequent judgments, even when the anchor is completely arbitrary.

The Classic Experiment

In Tversky and Kahneman's original study, participants spun a rigged wheel of fortune that landed on either 10 or 65. They were then asked whether the percentage of African countries in the United Nations was higher or lower than that number, and then to estimate the actual percentage. Those who saw 65 estimated an average of 45%, while those who saw 10 estimated only 25% — a massive difference driven by a clearly random number.

Anchoring-and-Adjustment Model

This simulator implements the anchoring-and-adjustment model: each person starts from the anchor and adjusts toward what they believe is the correct answer, but the adjustment is insufficient. The adjustment factor α controls how far people move — with α = 0.4, people only travel 40% of the distance from anchor to truth. The noise parameter adds realistic individual variation around this biased mean.

Reading the Visualization

The top histogram shows the distribution of estimates in the population. The red dashed line marks the true value, the cyan line marks the anchor, and the white line marks the mean estimate. The gap between the true value and the mean estimate is the anchoring bias. The bottom scatter plot shows individual estimates, revealing both the central tendency toward the biased mean and the spread of individual variation. Try moving the anchor dramatically — notice how the entire distribution shifts, even though the true value hasn't changed.

FAQ

What is the anchoring effect?

The anchoring effect is a cognitive bias where people's numerical estimates are systematically pulled toward an initial value (the anchor), even when that anchor is arbitrary or irrelevant. First described by Tversky and Kahneman in 1974, it is one of the most robust findings in judgment and decision-making research, replicated across cultures, expertise levels, and domains.

How does anchoring-and-adjustment work?

In the anchoring-and-adjustment model, people start from an anchor value and adjust until they reach a plausible estimate — but they typically stop adjusting too soon (insufficient adjustment). The adjustment factor α measures how far people move: α = 0 means no adjustment (stuck at the anchor), α = 1 means full correction to the true value.

Can experts resist anchoring?

Research consistently shows that expertise provides limited protection against anchoring. Real estate agents are anchored by listing prices, judges by sentencing demands, and doctors by initial diagnoses. While experts may show slightly less anchoring than novices, the effect remains substantial and practically significant across all studied professional domains.

What are real-world examples of anchoring?

Anchoring affects salary negotiations (first offer sets the range), pricing (original price anchors perceived value of discounts), legal judgments (prosecution demands anchor sentences), medical diagnosis (initial hypothesis anchors subsequent investigation), and financial forecasting (past values anchor future estimates).

Sources

Embed

<iframe src="https://homo-deus.com/lab/cognitive-biases/anchoring-effect/embed" width="100%" height="400" frameborder="0"></iframe>
View source on GitHub