Sealed-Bid Premortem¶
Foresight ritual — instantiates Winner-Conditioned Valuation Correction
Just before an irreversible sealed bid goes in, the team imagines it won and the deal went sour, then works backward to surface why — dragging the hidden reasons winning is bad news into view while the number can still change.
A sealed bid is a one-shot, irreversible commitment: you submit a number, and if you win, there is no ascending contest to reveal what rivals thought and no chance to revise. All of the correction has to happen before the envelope is sealed — which means it cannot come from feedback, only from imagination. Sealed-Bid Premortem supplies that imagination as a disciplined ritual. The team stipulates the outcome — "it is two years later, we won this bid, and it went badly" — and works backward to enumerate the reasons. Its one defining move is to run prospective hindsight on the winning branch specifically: not "will we win?" but "given we win, what will have made winning a mistake?" That framing pulls out precisely the winner's-curse hazards a forward-looking plan suppresses — the ways the shared value could be overstated, the reasons the seller is transacting now, and the places where the team's several "independent" signals are secretly the same signal. Each surfaced hazard is fed back into the number before submission.
Example¶
A contractor is hours from submitting a sealed lump-sum tender of ≈$47M for a bridge retrofit. The team runs the premortem: "We won, and two years on we're bleeding money — why?" The reasons come fast. The geotechnical survey — the same one handed to every bidder — probably understated the rock and groundwater conditions, and we won because we were the most optimistic reading of it. The authority put the job out to tender this quarter partly because its own engineers had flagged escalating risk, so the timing of the sale is itself information. And three of the "independent" subtrades in the estimate all sourced their steel quote from the same supplier, so the low numbers that made the bid competitive were really one number wearing three hats.
Those map cleanly onto three hazards: shared-value overstatement, seller motive, and correlated signals. The team raises the tender toward ≈$52M with an explicit contingency, and comes within a hair of no-bidding. Nothing here required new data — only the discipline of imagining the failure while the number was still editable.
How it works¶
- Stipulate the win and the failure. Assume the bid won and the outcome was bad; that fixed frame is what licenses candor the ordinary go/no-go conversation suppresses.
- Work backward through the winner's-curse hazards. Ask specifically where the common, shared value could be overstated; why the seller is transacting now and what that reveals; and where nominally independent estimates share a hidden source.
- Resolve each hazard. Convert every surfaced reason into either a concrete adjustment to the bid or an explicit, recorded "accepted."
- Do it before the seal. The exercise only has value ahead of an irreversible submission — there is no second pass after the envelope goes in.
Tuning parameters¶
- Imagined time horizon — how far after the win the failure is set. A near-term failure surfaces execution and information hazards; a long-horizon one surfaces slow-burn value erosion.
- Room composition — who participates. A diverse, partly-adversarial group surfaces correlated-signal and seller-motive hazards a homogeneous deal team would talk past.
- Feedback mode — whether hazards return as qualitative flags or as quantified add-ons to the bid. Quantified feedback bites harder but risks false precision.
- Kill threshold — how many or how severe the surfaced hazards must be to convert the exercise into a no-bid rather than a markup.
When it helps, and when it misleads¶
Its strength is reaching the hazards that quantitative correction assumes away. A valuation model can shade for a known number of rivals and a known noise level; the premortem[1] is what surfaces that the noise is correlated, that the seller's timing is a signal, and that the shared survey everyone trusts is the shared survey everyone is wrong about. For a one-shot sealed bid with no feedback loop, imagined foresight is the only correction available after the model has done its part.
Its failure modes follow from its being a generator of words, not numbers. It can produce a vivid list of fears that never actually moves the bid — anxiety theater. It can be run backwards, staged after the team has already decided to bid so the hazards are noted and waved through. And a homogeneous room will imagine a homogeneous, comfortable failure. The discipline that keeps it honest is to require every surfaced hazard to resolve into a specific bid change or an explicitly owned acceptance, and to seat someone whose job is to disagree.
How it implements the components¶
This ritual realizes the hazard-surfacing side of the archetype — the qualitative reasons winning is bad news that a model cannot generate on its own:
common_value_exposure_map— imagining the failure maps where the shared, uncertain value could be overstated, marking the exposures the bid must survive.seller_incentive_scan— it asks why the seller is transacting now, treating the sale's timing as joint evidence, alongside the win, about the common value.correlated_signal_warning— it flags where the team's ostensibly independent estimates or subquotes trace back to one source, the condition under which the winner's curse bites hardest.
It surfaces hazards but does not price them: the debiased base-rate estimate is Reference-Class Bid Review's output and the winning-conditional correction is Winner's-Curse-Adjusted Bid Model's — the premortem only supplies the flags those must account for.
Related¶
- Instantiates: Winner-Conditioned Valuation Correction — this ritual is the qualitative foresight step that surfaces the winner's-curse hazards a sealed, one-shot bid must price in advance.
- Sibling mechanisms: Winner's-Curse-Adjusted Bid Model · Reference-Class Bid Review · Independent Valuation Panel · Competing Estimate Simulation · Bid/No-Bid Gate · Post-Auction Loss Review
Editorial Notes¶
Form Classification¶
Form family: Experiment, Test & Rehearsal
Rationale: Sealed-Bid Premortem operates as an active test, trial, simulation, drill, or rehearsal that generates evidence through a deliberate attempt or perturbation because it just before an irreversible sealed bid goes in, the team imagines it won and the deal went sour, then works backward to surface why — dragging the hidden reasons winning is bad news into view while the number can still change.
Independent corroboration: The frozen evidence defines Sealed-Bid Premortem as 'Just before an irreversible sealed bid goes in, the team imagines it won and the deal went sour, then works backward to surface why — dragging the hidden reasons winning is bad news into view while the number can still change', so its operative form is Experiment, Test & Rehearsal.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Behavioral Economics
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: The mechanism combines auction winner's-curse reasoning with a premortem intended to counter overconfidence before commitment. AEA documents systematic overbidding and the winner's curse; Klein's premortem method elicits prospective failure explanations, making behavioral economics the best primary lineage.
Related originating lineages:
- Economics & Finance — Auction theory independently explains why winning can signal overbidding.
- Futurism & Strategic Foresight — Strategic foresight, scenario planning, and anticipatory governance supplies a parallel or contributing lineage for the mechanism's defining operation: just before an irreversible sealed bid goes in, the team imagines it won and the deal went sour, then works backward to surface why — dragging the hidden reasons winning is bad news….
- Organizational & Management Science — Team premortems materially structure backward causal review before commitment.
- Psychology — psychology contributes human judgment, assessment, learning, and behavioral response to this mechanism's defining operation—Just before an irreversible sealed bid goes in, the team imagines it won and the deal went sour, then works backward to surface why — dragging the hidden reasons winning is bad news into view while the number can still change—without displacing the selected primary historical lineage.
Review resolution: The blind reviewers disagree on primary lineage (behavioral_economics versus economics_finance). Authoritative or primary research supports behavioral_economics as the best historical origin: The mechanism combines auction winner's-curse reasoning with a premortem intended to counter overconfidence before commitment. AEA documents systematic overbidding and the winner's curse; Klein's premortem method elicits prospective failure explanations, making behavioral economics the best primary lineage. The cited American Economic Association, Auctions and the Winner's Curse; Harvard Business Review, Performing a Project Premortem directly supports the mechanism's defining operation. All independently supported contributing domains are retained without an arbitrary cap. origin_mode=cross_disciplinary_synthesis records the lineage relationship, while domain_reach=multi_domain records later applicability separately from provenance.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Researched adjudication after independent review; high confidence.
Sources consulted:
- American Economic Association, Auctions and the Winner's Curse
- Harvard Business Review, Performing a Project Premortem
References¶
[1] Klein, Gary. "Performing a Project Premortem". Harvard Business Review, September 2007. Uses imagined failure to surface threats and suppressed concerns before commitment. registry ↩