Cognitive Biases · CB-02
The first number you see shapes every estimate that follows — even when that number is random and you know it's random.
The tendency to rely disproportionately on the first piece of information offered (the 'anchor') when making subsequent judgments, even when the anchor is arbitrary, irrelevant, or explicitly known to be uninformative. Later estimates are adjusted from the anchor, but the adjustment is typically insufficient, leaving the final judgment biased toward the initial number.
Formally demonstrated by Amos Tversky and Daniel Kahneman in their landmark 1974 paper 'Judgment under Uncertainty,' including the famous wheel-of-fortune experiment where a spun number (rigged to land on either 10 or 65) measurably shifted participants' subsequent estimate of the percentage of African countries in the UN.
The Mechanism
A random spin changes a serious estimate by dozens of percentage points
Participants knew the number came from a spun wheel — visibly random, admittedly irrelevant — and still anchored on it by roughly twenty percentage points. Knowing about the bias intellectually did not prevent it from operating; this is one of the most replicated and robust findings in all of cognitive psychology.
01 · THE ANCHOR DOESN'T NEED TO BE RELEVANT, PLAUSIBLE, OR EVEN EXPLICIT
Irrelevant anchors move judgments almost as much as relevant ones
Anchoring effects appear even when the anchor is a random number, an obviously wrong reference point, or even just the last few digits of your phone number used as a prompt — the mind adjusts from whatever number it was recently exposed to, regardless of that number's actual informational value.
02 · ADJUSTMENT FROM THE ANCHOR IS TYPICALLY INSUFFICIENT
People move away from the anchor, but not far enough
The standard explanation (anchor-and-adjust) holds that people do try to move away from an anchor toward a more accurate estimate, but stop adjusting too soon — leaving the final judgment systematically biased in the anchor's direction, rather than landing on an anchor-independent number.
03 · IT'S EXPLOITED DELIBERATELY IN NEGOTIATION AND PRICING
Whoever sets the first number shapes the entire subsequent range
Skilled negotiators and retailers deliberately set an aggressive first number (a high initial asking price, a 'suggested retail price' next to a discount) precisely because it anchors the counterpart's subsequent judgment of what's reasonable — a well-documented, deliberately weaponized application of the same effect Tversky and Kahneman identified.
Where It Fails / Inversion
Where it fails / inversion
Not every first number creates a meaningful anchor — anchoring effects are weaker when the anchor is obviously absurd, when the estimator has strong independent expertise in the domain, or when there's real time and incentive to think carefully rather than rely on a quick adjustment. Treating anchoring as an unavoidable, universal force can lead to overcorrecting against perfectly reasonable initial reference points.
How To Use It
Worked example · negotiating a salary or a deal price
Whoever states the first number in a negotiation sets the anchor the rest of the conversation adjusts from — which is why well-prepared negotiators arrive with their own researched number ready to state first, rather than waiting to hear the other side's opening offer and being anchored by it. If the other side anchors first, deliberately and explicitly reset the frame (citing independent data) rather than simply negotiating down from their number.
How to use it
Before finalizing any estimate — a price, a timeline, a probability — ask what number you saw or heard first, and consider deliberately generating your own independent estimate before being exposed to anyone else's, precisely to avoid anchoring on a number that may have no informational value at all.
See Also