Anchoring is the bias that survives knowing about it. Chapter 275 sets out the evidence: transparently random anchors work, warnings do not help much, incentives for accuracy do not help much, and Northcraft and Neale's estate agents produced valuations that tracked a manipulated list price while denying that it had influenced them.
Two defenses have experimental support, and both are procedural rather than attitudinal.
The first is to fill the slot before it can be filled for you.
Mussweiler and Strack's account of the mechanism is selective accessibility: the anchor makes anchor-consistent evidence easier to retrieve, so the adjustment is performed on a biased sample. An empty slot is easy to fill and an occupied one is much harder to displace, which means the timing of your own estimate matters more than its quality.
The protocol is therefore: before the meeting, in writing, decide three numbers. Your target. Your reservation value — the point at which you walk. And your best alternative if no deal happens, stated concretely rather than as a feeling.
Writing them down is not ceremonial. A number in your head is revisable under pressure without your noticing; a number on paper is an artifact you have to consciously overrule, and the act of overruling it is visible to you.
The second defense is consider the opposite, and Mussweiler, Strack and Pfeiffer tested it directly in 2000. Because anchoring runs on one-sided retrieval, forcing the deliberate generation of anchor-inconsistent evidence restores balance. Their participants who were prompted to consider reasons the anchor was inappropriate showed substantially reduced bias, and later work has applied the same debiasing to judicial and managerial decisions.
In practice: when a number lands, spend sixty seconds generating specific reasons it is wrong in the opposite direction. Not doubting it — enumerating. If the figure is high, list what would justify a much lower one; comparable transactions, defects, the seller's timeline, what it cost to produce.
The early signal that you have been anchored is that you find yourself arguing about how far to move from their number rather than about what the thing is worth. The conversation's unit has become adjustment, and adjustment is always insufficient.
The counter-response, once you have your own figure, is to re-anchor rather than to negotiate down: put your number on the table with a rationale, and let the midpoint move.
This counters Law 5.