Anchoring Bias

The first number or idea a person encounters tends to stick, coloring every judgment that comes after it — even when it has no real bearing on the answer.

Anchoring bias denotes the tendency to overweight an initial reference value when estimating an uncertain quantity, skewing every judgment that follows. Because the effect survives even when the anchor is transparently arbitrary, it reliably distorts decision distributions regardless of context.

Show a buyer a $40,000 sticker price first, and even a hard-won discount down to $34,000 will feel like a win — despite the car's fair market value sitting well below that. The opening number quietly redefined what counted as a "good deal."

One mock-jury study had participants hear either a 34-month or a 12-month sentencing demand from a prosecutor, with explicit instructions that the demand carried no binding weight. Jurors exposed to the higher figure still handed down meaningfully longer sentences. The anchor had imposed a high prior on the sentencing estimate, suppressing genuine evidence integration and producing a reference-point bias that survived the disregard instruction entirely — proof that the anchor functions as a near-fixed constraint on reasoning rather than something jurors could simply choose to ignore.

An early number shifts every later guess toward it, because people underadjust away from whatever they saw first. The first thing simply feels more important than it should.

The mechanism is a weighted-integration process: the initial anchor imposes a high prior weight on the estimate, and incoming evidence gets downweighted relative to it. This reference-frame constraint, operating at the level of contextual analysis, produces a systematic asymmetry in how beliefs get updated.

Form your own estimate before anyone shows you theirs, and treat the first number you hear as one data point, not the answer. Testing your response against very different starting numbers reveals how much the anchor actually moved you.

Debiasing starts with generating independent estimates before exposure to any anchor, then explicitly adjusting away from whatever value turns out to be salient. Counterfactual sampling and evidence reweighting further counteract the prior-induced asymmetry.

Anchor is arbitrary; Maliciously set anchor; Overreliance on anchor

An adversarial actor can deliberately seed high or low anchor values in negotiations, pricing, legal proceedings, or media framing to systematically skew target estimates before deliberation begins. Because anchors retain influence even when flagged as arbitrary, a bad actor can introduce a patently extreme first number—a wildly inflated asking price, an exaggerated damage claim, or a manipulative polling figure—knowing that final outcomes will still drift toward it. In information warfare, anchored narratives (e.g., inflated casualty figures or biased baseline statistics) planted early in news cycles constrain all subsequent reporting and public estimation within a distorted reference frame.

Generate independent personal estimates before exposure to external reference values and document reasoning to resist post-anchor revision. Apply debiasing protocols such as consider-the-opposite thinking or counterfactual sampling to surface evidence contradicting the anchor's implied range. Institutionally, delay anchor-bearing information until after independent estimates are recorded—analogous to blinded review in research—to substantially reduce prior-induced asymmetry in group decision outcomes.