S.M.M.

Stop Manipulating Me · A Field Guide to Psychological Influence

ENTRY No. T21.17
CATEGORY Negotiation Pressure
CLEARANCE Public / Essential
EDITION 01
Dossier · Manipulation Tactic

Bracketing

Anchoring · Midpoint Targeting · Widespread
Caution
How It WorksSEC 01

Bracketing structures a series of offers so that the seemingly fair compromise — the midpoint between the two sides — lands on the mover’s actual target. It marries anchoring (Galinsky; Kahneman & Tversky) to the human pull toward compromise: because “splitting the difference” feels equitable, a negotiator who can control both the distance and direction of the endpoints can steer the “fair middle” wherever they like. If their true goal is a given number, they open far enough on the other side that the eventual split arrives precisely there, while the outcome wears the costume of mutual concession.

This is standard, largely benign bargaining behavior — nearly everyone opens away from their target and expects to move, and there’s nothing dishonest about aiming for a favorable middle. It becomes manipulative when the opening is extreme specifically to distort the reference point and the counterpart is pushed toward a reflexive split before independently evaluating either figure. The tell is a midpoint that is suspiciously convenient, an aggressive first number that makes a still-favorable middle look moderate by contrast, and a fairness framing (“let’s just meet in the middle”) offered in place of any evidence about actual worth.

The defense is to anchor to reality rather than to their geometry. Come in with a researched target and reservation value so that offers are judged against objective worth, not against the counterpart’s opening. Decline to treat a midpoint as neutral — it is fair only when both endpoints are honest — and if the opening is extreme, re-anchor the range with your own well-justified number so any future split reflects the item’s real value. Kept on objective criteria, the conversation is about worth, and the bracketing math loses its grip.

Warning SignsSEC 02
  • The midpoint is suspiciously round. Their opening plus yours splits neatly to a number that happens to be exactly what they'd want.
  • Push toward "let's just split it." The move steers you to a reflexive compromise before you've evaluated either figure independently.
  • Extreme opening, then reasonableness. An aggressive first number sets up a middle that only looks moderate by contrast.
  • Counter-offer mirrors your move. Each of your concessions is matched to keep the running midpoint parked on their target.
  • Fairness framing over data. "Meeting in the middle" is offered as inherently fair, independent of what the item is actually worth.
Frequently Paired WithSEC 03
  • Highball · T21.2
    combines
  • Lowball · T21.3
    combines
  • Anchoring · T8.2
    home
How the Hook LandsSEC 04
  • Stage 01 · Set the far anchor
    The mover opens with a number deliberately extreme relative to their true goal, chosen so the eventual midpoint lands on target.
  • Stage 02 · Invite the split
    As figures converge, they nudge toward "meeting in the middle," framing the midpoint as the obviously fair outcome.
  • Stage 03 · Land on target
    Compromise bias does the rest: the split feels equitable, but the middle was engineered to sit exactly where they wanted all along.
Counter-ProtocolSEC 05
Defense: Anchor to your own researched value — a midpoint is only fair when both endpoints are honest.
  • Anchor independently. Enter with a researched target and reservation value (Galinsky on anchoring). Judge offers against your data, not against their opening.
  • Don't reflex-split. "Meeting in the middle" is fair only if both endpoints are legitimate. Refuse to treat the midpoint of a rigged range as neutral.
  • Re-anchor the range. If their opening is extreme, reset with your own well-justified number so any future midpoint reflects reality.
  • Track the true value. Fisher & Ury: use objective criteria — market comps, cost data, standards — so the conversation is about worth, not geometry.