Belief Update Review Template¶
Template — instantiates Revision-Readiness Precommitment
A repeatable artifact for recording original claim, update condition, observed evidence, and revision decision.
A Belief Update Review Template is a fixed fill-in-the-blank form that captures one belief's whole life in a single durable record: what you claimed, how sure you were, what you said would change your mind, what you later saw, and how the belief actually moved. Its distinguishing move among its siblings is that it is a record artifact, not a trigger, a gate, or a working session — it does not design a clever update condition or fire when evidence arrives; it preserves the whole story in one place so that, months later, anyone (including your future self) can see whether you genuinely updated or quietly slid the goalposts. The condition and the decision live inside the form as recorded content, but the template's own job is fidelity of the record, filled once at commitment and reopened once at review.
Example¶
A UX researcher believes that shipping a dark-mode option will lift daily active use of a reading app, and rather than just asserting it she fills a Belief Update Review Template. In the claim field: "Dark mode will raise 28-day retention among night readers, without adding a support-ticket burden." In the confidence field she writes a plain rating — "fairly confident, ~65% (illustrative)" — so any later movement is visible against a fixed starting point. She records, at commitment, the update condition she and the team agreed to and the review date, then closes the form.
Ten weeks later she reopens the same sheet. She pastes in the observed evidence (retention flat, tickets up), her honest reading of whether the pre-stated condition was met, and the decision that followed: "belief lowered to ~30%; dark mode kept as preference, retention hypothesis retired." The value is not the prediction — she was wrong — but that the wrongness is now legible: the sheet holds the original 65%, the agreed condition, and the drop, side by side, so the update cannot later be reframed as "we always knew it was marginal."
How it works¶
The template is a small set of fixed fields, filled in two passes:
- Commitment pass (before evidence). Record the revisable claim in testable language, the current confidence baseline, the update condition and review date as agreed, and file the sheet somewhere it cannot be silently edited.
- Review pass (after the window). Reopen the same sheet; append the observed evidence, an honest interpretation of whether the condition was met, and the revision decision. Nothing from the commitment pass is overwritten.
The discipline is that the two passes are separated in time and the early fields are write-once. Because the sheet is a reusable template, the same skeleton is applied to every belief, so a person or team accumulates a comparable archive rather than a drawer of one-off notes.
Tuning parameters¶
- Field granularity — a three-line index card versus a structured form with separate evidence, interpretation, and action fields. More fields capture more, but a heavy form gets skipped; match weight to how much the belief matters.
- Write-once strictness — whether early fields are truly locked (version-controlled, dated) or merely conventionally not-edited. Strict locking is what actually defeats goalpost-shifting; soft convention is lighter but leakier.
- Confidence format — a number, a band, or a word ("hunch / fairly sure / near-certain"). Numbers make movement precise but invite false precision; words are honest about vagueness.
- Retention and review cadence — a one-off sheet versus a standing archive re-read on a schedule. An archive lets you audit your own calibration over many beliefs, at the cost of upkeep.
When it helps, and when it misleads¶
Its strength is legibility. By holding the original claim, the starting confidence, and the eventual outcome in one write-once record, it turns "of course I update my views" into something checkable, and it makes a genuine reversal easy to show rather than merely assert. It is the natural substrate for reviewing your own calibration and for making revision a normal, documented event — the same spirit as a decision journal, where the point is to preserve your reasoning at the moment of choice so hindsight cannot rewrite it.[n1]
Its failure mode is that a template records; it does not enforce. A locked field stops you editing the past, but nothing stops you writing a vague, unfalsifiable claim in the first place, or filling the review pass with a self-flattering interpretation. The classic misuse is the retrospective fill — writing the "before" fields after the evidence is in, so the sheet launders a settled conclusion as a documented update. The guarding discipline is to date and lock the commitment pass independently (a timestamp someone else can see), and to keep the claim concrete enough that the review pass has something real to check against.
How it implements the components¶
The template realizes the record-keeping face of the archetype — the components that must be captured and preserved, not the ones that design or enact the update:
revisable_claim_statement— its first field forces the belief into concrete, testable language, refusing "our strategy is good"–style vagueness.current_confidence_baseline— a dedicated field fixes the starting confidence so any later movement is measured against a recorded anchor rather than a foggy memory.accountability_trace— the artifact is the trace: a single write-once record binding the original claim, the confidence baseline, the evidence later observed, and how the belief was ultimately revised, so goalpost-shifting is visible.
It does not design a discriminating evidence_update_condition (that is the Red-Team Update Condition Session and the Forecast Update Trigger Log), set a numeric revision_threshold or watch an observation_window (the Strategic Revisit Gate), or pre-map and enact the update_action_mapping when a trigger fires (the Forecast Update Trigger Log and the gate). Its nearest twin is the Forecast Update Trigger Log, which is a running ledger that tracks probability movement across many forecasts and enacts each update; the template is a static, single-belief record that preserves one belief's whole story.
Related¶
- Instantiates: Revision-Readiness Precommitment — it supplies the durable record layer the archetype needs so revision can be audited after the fact.
- Sibling mechanisms: Forecast Update Trigger Log · Preregistration Refutation Table · Red-Team Update Condition Session · Strategic Revisit Gate · If-Then Revision Contract
Editorial Notes¶
Form Classification¶
Form family: Representation, Specification & Plan
Rationale: A repeatable artifact for recording original claim, update condition, observed evidence, and revision decision, making its operative form a non-executable information artifact that externalizes static or prospective structure.
Independent corroboration: The frozen evidence defines Belief Update Review Template as 'A repeatable artifact for recording original claim, update condition, observed evidence, and revision decision', so its operative form is Representation, Specification & Plan.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Philosophy
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: The template's claim, defeater, evidence, and revision structure operationalizes epistemological standards for responsible belief change.
Related originating lineages:
- Organizational & Management Science — Review templates and controlled records contribute the durable, reopenable artifact form.
- Statistics & Experimental Design — Precommitment and evidence-conditioned updating contribute the explicit update condition.
Review resolution: Philosophy is the agreed primary lineage through norms of justified belief revision. Organizational learning supplies the review workflow and statistics supplies uncertainty-weighted evidence; broad usefulness is multi-domain rather than universal.
Encyclopedia synthesis: The exact catalogued form synthesizes established practice rather than reproducing a single standard historical label.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] A decision journal is a recorded note, written at the moment of a decision, of the reasoning and expected outcome — a practice popularized in the decision-making literature (e.g. by Michael Mauboussin) precisely to stop hindsight from rewriting what you believed. A belief-update review template is the same idea narrowed to belief revision, with a locked "before" and a later "after." ↩