Progressive Option Screening¶
Procedure — instantiates Bounded-Rationality Decision Design
Uses cheap broad filters before costly deep evaluation while retaining false-negative review and reentry.
Progressive Option Screening runs a large pool of candidates through a sequence of filters ordered cheapest-first, so that the expensive evaluation is spent only on the few that survive the cheap ones. The idea that makes it this mechanism and not a sibling is that it manages a whole population of options through staged filtering, not a single acceptance decision: each stage discards the bulk of what remains using a coarse, low-cost test, and only the survivors advance to a finer, costlier one. Its defining safeguard is a memory of what it threw away — a false-negative audit and a reentry path — because the risk of a cheap early filter is that it discards a viable option before anyone looked hard enough to see it.
Example¶
A pharmaceutical team searching for a drug candidate begins with a library of a million compounds — far too many to test properly. Progressive Option Screening is the funnel that makes the search tractable. Stage one is high-throughput screening: an automated assay tests all million compounds cheaply and fast, keeping only the few thousand that show any activity against the target. Stage two applies a costlier battery — potency, selectivity, basic toxicity — narrowing thousands to dozens. Stage three, the most expensive by far, puts a handful of "leads" through animal studies. Each stage is cheap relative to the next and brutal about discarding, so the million-compound haystack collapses to a few serious candidates without the lab ever running its most expensive test a million times. And because a promising compound can be knocked out early by a quirk of the crude first assay, the team samples the rejected pool and keeps a path to pull a discarded compound back in if a later insight warrants — the false-negative[n1] check that keeps the funnel from silently throwing away the winner.
How it works¶
The procedure works by (1) ordering the available filters from cheapest-and-broadest to most-expensive-and-narrowest, matched to what information is actually obtainable at each stage; (2) applying each filter to the survivors of the previous one, discarding the majority at low cost; (3) advancing only the survivors to the next, costlier method; and (4) auditing the discarded pool — sampling false negatives and preserving a reentry path — so a cheap early cut does not become an irreversible verdict. Its distinctive move is spending evaluation cost in proportion to how far a candidate has already come: the field is narrowed by many cheap passes rather than one exhaustive comparison, which is what makes screening a huge pool affordable.
Tuning parameters¶
- Filter ordering and thresholds — which test runs first and how strict each cut is. Aggressive early cuts save the most money but raise the false-negative rate; lenient early cuts pass too much to the costly stages.
- Stage count — how many filters in the funnel; more stages narrow more finely but add handoffs and chances to wrongly discard.
- False-negative sampling rate — how much of the rejected pool is re-examined; higher sampling catches more wrongly-cut options but costs the very effort screening was meant to save.
- Reentry policy — whether and how a discarded option can re-enter at a later stage when new information favors it; generous reentry protects against premature loss but reopens closed work.
- Diversity preservation — whether a holdout of varied candidates is carried through regardless of early scores, guarding against a filter that is systematically blind to a whole class.
When it helps, and when it misleads¶
Its strength is that it makes an otherwise unaffordable search feasible: by refusing to run the expensive test on everything, it evaluates a vast field within a real budget while still ending with deeply-vetted finalists.
Its characteristic failure is premature pruning — a cheap early filter, chosen for speed, discards viable or underrepresented options that a closer look would have kept, and because they are gone, nobody learns they were wrong to cut. That is the archetype's premature-pruning failure mode, and it bites hardest when the early filter correlates with something irrelevant to real quality. A related misuse is treating the funnel's survivors as "the best options" when they are only "the options the early filters happened to favor." The guarding discipline is exactly the false-negative audit and reentry path built into the procedure: sample the discards, preserve a diverse holdout, and treat any early cut as reversible on new evidence rather than final.
How it implements the components¶
search_and_option_budget— the staged funnel is an option budget: it caps how much costly evaluation is spent by letting only survivors advance, so total search cost stays bounded.information_availability_map— the filter order tracks what information is cheap versus expensive to obtain, applying the tests whose data is already at hand before those that must be gathered.method_portfolio— each stage assigns a different method, from a crude broad screen to a rigorous narrow one, sequenced by cost.
It does NOT stop at the first candidate that clears a good-enough bar via aspiration_and_sufficiency_criterion — that is Satisficing Threshold Rule — and it does NOT bound the search by an elapsed-time budget via resource_limit_profile — that is Timeboxed Search. Progressive screening filters a whole pool in cost-ordered stages rather than stopping on quality or on the clock.
Related¶
- Instantiates: Bounded-Rationality Decision Design — the procedure is the archetype's staged search-and-pruning move, with the false-negative safeguards that keep pruning honest.
- Sibling mechanisms: Satisficing Threshold Rule · Timeboxed Search · Algorithmic Escalation Gate · Default and Delegation Protocol · Two-Stage Review · Post-Decision Calibration Review · Cognitive Offloading Aid · Choice Architecture Simplification · Decision Method Triage Matrix
Editorial Notes¶
Form Classification¶
Form family: Decision, Gate & Allocation
Rationale: Progressive Option Screening operates as a case-specific gate, selection, routing, prioritization, or resource disposition because it uses cheap broad filters before costly deep evaluation while retaining false-negative review and reentry.
Independent corroboration: The frozen evidence defines Progressive Option Screening as 'Uses cheap broad filters before costly deep evaluation while retaining false-negative review and reentry', so its operative form is Decision, Gate & Allocation.
Nearest alternative: Protocol, Workflow & Routine — Progressive Option Screening includes features of a repeatable ordered procedure or handoff sequence that coordinates action, but its defining operation is a case-specific gate, selection, routing, prioritization, or resource disposition.
Review outcome: Independent reviewer agreement; medium confidence.
Origin Attribution¶
Primary origin: Operations Research
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Multi-domain
Rationale: Progressive Option Screening is most plausibly rooted in the operations_research tradition because its characteristic form depends on queueing, optimization, scheduling, prioritization, and constrained allocation. The assignment tracks that formative lineage, not the many settings in which the mechanism can now be applied.
Related originating lineages:
- Organizational & Management Science — The organizational_management tradition materially shaped Progressive Option Screening through its own practice of the coordination, governance, learning, and redesign of organized work.
- Statistics & Experimental Design — The statistics_experimental_design tradition materially shaped Progressive Option Screening through its own practice of probability, calibrated inference, experimental design, and uncertainty analysis.
Review resolution: Both blind reviewers agree that operations research is the primary origin. Explicit reconciliation resolves alternate origin disagreement, origin mode disagreement. Formative alternate lineages are retained as organizational_management, statistics_experimental_design; later breadth of use is recorded separately as domain_reach=multi_domain, while origin_mode=cross_disciplinary_synthesis describes the relationship among origin lineages.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] In screening pipelines, a false negative is a viable candidate wrongly discarded by an early, coarse filter — the costliest error a cheap-first funnel can make, because the discard is invisible unless the rejected pool is deliberately sampled. Preserving reentry and a diverse holdout is the standard defense. ↩