The spinner experiment
In one of the most disturbing demonstrations of cognitive bias, researchers spun a wheel of fortune in front of participants. The wheel was rigged to stop at either 10 or 65. Participants were then asked: what percentage of African nations are in the United Nations?
Those who saw 10 estimated around 25 percent. Those who saw 65 estimated around 45 percent. The wheel was random. The question was factual. The answer had nothing to do with the spinner. And yet the random number anchored the participants' estimates so powerfully that their answers differed by 20 percentage points. This is anchoring bias: the first number encountered becomes a reference point from which all subsequent judgments are made, even when that number is meaningless.
Anchoring does not require the first number to be relevant, accurate, or chosen deliberately. It requires only that it appears first. The brain establishes a range based on the first signal it receives, and then adjusts from there. The adjustment is typically insufficient — a phenomenon called insufficient adjustment — because the brain is anchored to the original number and does not move far enough from it, even when it should.
Random numbers anchor estimates as powerfully as relevant ones
Negotiation and pricing
Anchoring is most visible in negotiation. The party who makes the first offer has a significant advantage, not because the first offer is necessarily better, but because it establishes the anchor. A salary negotiation that opens with a low offer sets a low anchor, and subsequent counteroffers will cluster around that anchor, even if the employer would have been willing to go higher. The candidate who anchors first often ends the negotiation better — not because they asked for more, but because they shaped the range.
Retail pricing exploits anchoring deliberately. The "original price" that is crossed out — $199, now $99 — is an anchor. The original price is probably inflated to establish the anchor, not to reflect reality. The customer evaluates $99 against the anchor of $199, perceives it as a good deal, and buys. The retailer knows that the first price shapes the customer's perception of value, regardless of what the product is actually worth.
Real estate is one of the most anchor-susceptible markets. The listing price sets the anchor. Buyers make offers relative to the listing price. Even when the listing price is deliberately inflated, it shapes what buyers consider reasonable. A house listed at $800,000 will attract offers averaging perhaps $750,000. A house listed at $900,000 will attract offers averaging perhaps $830,000. The house did not change. The anchor did.
In expert judgment
Anchoring affects professionals as much as laypeople. Studies with judges, accountants, and financial analysts show that providing an initial number — even one described as arbitrary — shifts their final estimates substantially. A group of judges asked to roll dice before sentencing gave sentences that correlated with the die roll. Not because the dice were meaningful, but because they established an anchor from which the judge adjusted upward or downward.
Experienced negotiators know this. The opening position in any negotiation is deliberately extreme — not because the party expects to get it, but because it establishes an anchor that the final agreement will be measured against. The extreme opening compresses the range in the negotiator's favor. This is why the first number matters more than it should: it sets the terms of the debate before the debate begins.
In forecasting, anchor-and-adjust leads to predictable errors. When people estimate the range of a forecast — best case, worst case, most likely — they anchor to the most recent data point or the first number considered. Then they adjust insufficiently. The result is a forecast that reflects recent experience rather than true probability, and a range that is too narrow because the adjustment from the anchor does not reach far enough in either direction.
Overcoming anchoring
The first defense against anchoring is awareness. Before evaluating any number, ask: is this number meaningful, or is it an anchor? Who set this number, and why? If a salesperson tells you the competitor's price is "much higher," that is an anchor. If a recruiter mentions a salary range, that is an anchor. If a news story mentions a stock price at a particular moment, that is an anchor for how you perceive current value.
The second defense is deliberate consideration of alternatives. Instead of adjusting from the anchor, consider the problem from scratch. What would you think if you had not seen the anchor? What is the base rate? What are alternative explanations? This requires cognitive effort, but effort is what insufficient adjustment lacks. The brain wants to adjust from the anchor. You must override that impulse.
The third defense is considering the anchor from the outside. Ask: what would a reasonable person think if they had no prior information? This external perspective helps break the grip of the first number, because it removes the anchor from its privileged position as the reference point.
Anchoring bias is not a curiosity. It is a systematic distortion that affects every numeric judgment you make. The first number you see in any decision context will shape your perception of the entire range. Know this. Question the first number. Ask where it came from and whether it deserves its anchoring power. Most of the time, it does not.
Continue exploring