Differential Pattern Review¶
Differential review — instantiates Archetype Overmatching Guardrail
Holds the proposed archetype against its nearest rivals and asks which features would tell them apart, refusing the match until the evidence favors it over the alternatives.
Differential Pattern Review borrows the logic of differential diagnosis: instead of asking "does this fit archetype X?", it asks "among X and its plausible rivals, which does the evidence actually favor, and what feature would distinguish them?" Its defining move is to name the rivals up front and then hunt specifically for the discriminating evidence — the features that would support a competing pattern — rather than stopping the moment the first candidate tells a coherent story. A match is accepted only when it beats the field, not merely when it is self-consistent.
Example¶
A patient arrives with right-lower-quadrant pain and the on-call reflex is "classic appendicitis." A differential pattern review makes the reflex earn it. First the plausible alternatives that present similarly go on the list — mesenteric adenitis, ovarian torsion, a kidney stone, a Crohn's flare — and for each, the feature that would separate it from appendicitis. Then the review probes for exactly those discriminators: is there an adnexal finding, flank radiation, a different fever curve? The migratory pain and rebound tenderness favour appendicitis, but the review deliberately looks for the counter-features the anchoring diagnosis would gloss over. When the rivals' discriminators come back negative and appendicitis's come back positive, the match is earned; if a rival's discriminator lights up, the diagnosis splits or defers for more evidence. (Illustrative clinical sketch, not medical guidance.)
How it works¶
The rivals are enumerated before the evidence is weighed, and for each pairwise contrast the single most discriminating feature is identified. The review then actively seeks the evidence that would favour a rival — a targeted counterexample probe aimed at the leading candidate — and accepts the candidate only when its margin over the runner-up is decisive. Where a plain fit check asks whether a story holds together, this asks whether a better story is available.
Tuning parameters¶
- Breadth of the differential — how many rivals to entertain. Too few risks premature closure on the obvious pattern; too many wastes effort on remote ones.
- Discriminator selection — choosing the feature that best separates each pair, so high-information checks come first.
- Stopping rule — how large the leader's margin over the runner-up must be before the match is accepted; tighter margins demand more evidence.
- Split-diagnosis tolerance — whether to allow "could be X or Y, act for both" versus forcing a single label.
When it helps, and when it misleads¶
It is a direct antidote to premature closure and anchoring — the reflex match that feels understood too soon — and its discipline is native to medicine as the differential diagnosis.[n1] It fails when the rival set omits the true pattern (you can only choose among the alternatives you listed), when a common pattern's high base rate is ignored in favour of a vivid rare one, or when it is run to confirm the favourite by stocking the list with strawman rivals. The guarding disciplines: include the base-rate-common rivals, not only the interesting ones, and pick discriminators that could actually overturn the leader.
How it implements the components¶
neighbor_comparison— the review is organized around an explicit list of rival archetypes and the pairwise discriminators between them.counterexample_probe— for each rival it actively seeks the evidence that would favour that rival over the candidate, a targeted hunt for disconfirmation of the leading match.
It does NOT lay the comparison out as a static grid — that's Case Comparison Matrix; it does not run an open-ended search for any disconfirming case — that's Counterexample Search Session; and it does not attach the resulting confidence to the decision — that's Decision Confidence Label.
Related¶
- Instantiates: Archetype Overmatching Guardrail — the review forces a candidate pattern to win against its rivals before it is allowed to explain the case.
- Consumes: Case Comparison Matrix is often the worksheet it reasons over.
- Sibling mechanisms: Case Comparison Matrix · Counterexample Search Session · Pattern Fit Checklist · Pattern Fit Scoring Rubric · Anti-Pattern Review · Precedent Distinction Memo · Red-Team Pattern Match Review · Review Queue · Decision Confidence Label
Editorial Notes¶
Form Classification¶
Form family: Assessment, Review & Assurance
Rationale: Differential Pattern Review operates as a bounded evaluation of existing evidence or work that produces a finding or disposition because it holds the proposed archetype against its nearest rivals and asks which features would tell them apart, refusing the match until the evidence favors it over the alternatives.
Independent corroboration: The frozen evidence defines Differential Pattern Review as 'Holds the proposed archetype against its nearest rivals and asks which features would tell them apart, refusing the match until the evidence favors it over the alternatives', so its operative form is Assessment, Review & Assurance.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Medicine & Healthcare
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Clinical differential diagnosis supplied the recognizable method of comparing a proposed match with its nearest look-alikes through discriminating features.
Related originating lineages:
- Engineering & Design — Pattern and design review supplied the target domain in which competing archetype matches are evaluated.
Review resolution: Clinical differential diagnosis supplied the recognizable method of comparing a proposed match with its nearest look-alikes through discriminating features. The source explicitly transposes clinical differential diagnosis into design-pattern classification, making the engineering lineage constitutive of the synthesized review.
Attribution caveat: The review is an explicit transposition of clinical differential diagnosis to design-pattern classification.
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¶
The Matrix and this review divide the same labour: the matrix is the display of case-against-neighbours, this is the reasoning that names rivals, chooses discriminators, and rules among them. Use the matrix to see the alignment; use the review to decide what the alignment means.
[n1] Differential diagnosis — the clinical method of listing every condition that could produce the findings and using discriminating tests to narrow among them — is the direct ancestor of this mechanism, transposed from disease to archetype. ↩