Anti-Pattern Review¶
Catalog-based review — instantiates Archetype Overmatching Guardrail
Checks the case against a catalog of the archetype's known look-alikes — situations that show the surface markers but carry a disqualifying feature — so a familiar-looking failure isn't mistaken for the pattern it mimics.
Anti-Pattern Review works the guardrail from the failure side. It maintains a catalog of the archetype's known anti-patterns — cases that wear the pattern's surface badge but contain a structural disqualifier that makes applying the pattern actively harmful. The review separates the tempting surface markers (logged and explicitly discounted as non-probative) from the specific disqualifying features (tested for). Its defining move is that a hit on a catalogued disqualifier flips the label from pattern to anti-pattern — the resemblance was real, the fit was not.
Example¶
A team proposes a shared Singleton configuration object: "this is the classic case for a Singleton — one config for the whole app." An anti-pattern review pulls the catalog of Singleton anti-patterns. First it logs the surface markers that make Singleton tempting — there is conceptually one of the thing, and global access is convenient — and marks them non-probative, because they are exactly what fools people. Then it tests for the disqualifying features that turn Singleton into an anti-pattern here: does the "one" instance actually vary by tenant, test, or thread? Does global access hide dependencies and block testability? Both fire — the app is multi-tenant and the team already fakes the singleton in tests. The review flags the case as the Singleton anti-pattern and the design moves to injected configuration. Nothing about the surface resemblance was wrong; it simply wasn't structural.
How it works¶
The review starts from a library of the archetype's known failure look-alikes rather than from the pattern's positive requirements. For each, it records the surface markers that make it resemble the good pattern — and sets them aside as evidence of nothing — then tests the case for the disqualifying structural feature whose presence defines that anti-pattern. Detection of a named disqualifier converts the match into its failure twin.
Tuning parameters¶
- Library scope and freshness — how many look-alikes are catalogued and how current; a stale library silently misses newly recognized failure modes.
- Disqualifier strictness — how much of a disqualifying feature must be present to trigger the flag.
- Surface-marker discounting — how aggressively the tempting resemblances are set aside; over-discounting can wave away a genuine fit as if it were a look-alike.
When it helps, and when it misleads¶
It is strongest in fields that have accumulated named failure modes masquerading as good solutions — software engineering's "anti-pattern" is the canonical case.[1] It fails when the case's failure mode is novel and simply not in the catalog yet (it passes unflagged), or when zeal turns every proposed pattern into a suspected anti-pattern (chronic over-rejection). The guarding disciplines are to keep the library evidence-based and additive, and to require a named, specific disqualifier before flagging — not a general unease.
How it implements the components¶
surface_resemblance_log— it explicitly catalogs the tempting surface markers that make an anti-pattern look like the real pattern, and records them as non-probative.structural_fit_check— it tests the case for the specific disqualifying structural features whose presence converts the pattern into its anti-pattern (a fit check run in the negative).
It does NOT generate fresh counterexamples from history — that's Counterexample Search Session; it does not weigh the candidate against neighbour archetypes — that's Differential Pattern Review; and it does not set the necessary-feature gate for the *positive pattern — that's Pattern Fit Checklist.*
Related¶
- Instantiates: Archetype Overmatching Guardrail — the review guards specifically against the case where the pattern's own famous failure mode is mistaken for the pattern.
- Sibling mechanisms: Counterexample Search Session · Pattern Fit Checklist · Differential Pattern Review · Case Comparison Matrix · Pattern Fit Scoring Rubric · Precedent Distinction Memo · Red-Team Pattern Match Review · Review Queue · Decision Confidence Label
Editorial Notes¶
Form Classification¶
Form family: Assessment, Review & Assurance
Rationale: Checks the case against a catalog of the archetype's known look-alikes — situations that show the surface markers but carry a disqualifying feature — so a familiar-looking failure isn't mistaken for the pattern it mimics, making its operative form a bounded evaluation of existing evidence or work that produces a finding or disposition.
Independent corroboration: The frozen evidence defines Anti-Pattern Review as 'Checks the case against a catalog of the archetype's known look-alikes — situations that show the surface markers but carry a disqualifying feature — so a familiar-looking failure isn't mistaken for the pattern it mimics', so its operative form is Assessment, Review & Assurance.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Computer Science & Software Engineering
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Software anti-pattern practice supplies catalog-based recognition of superficially attractive structures with disqualifying failure conditions.
Related originating lineages:
- Architecture & Urban Planning — Pattern-language thinking supplies the positive-pattern counterpart.
- Engineering & Design — Failure-mode review compares designs to known unsafe configurations.
- Organizational & Management Science — Design and project reviews generalized anti-pattern checks to institutions.
- Systems Thinking & Cybernetics — System-archetype diagnosis contributes recurring structural patterns.
Review resolution: Software anti-pattern recognition is primary. Architectural pattern language, engineering review, organizational failure analysis, and systems structure materially shape the review; separating surface markers from disqualifying structure is synthesized here.
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.
References¶
[1] An anti-pattern is a commonly-reached-for "solution" that looks appropriate but reliably makes things worse; the term was popularized in software engineering by Brown et al., AntiPatterns (1998). The mechanism generalizes the idea — a catalog of a pattern's look-alike failures — to any archetype. registry ↩