Candidate Cap¶
Protocol — instantiates Solution Space Bounding
Fixes a maximum number of candidates — options, vendors, hypotheses, designs — carried into a cycle, filling the slots by ranking or sampling.
A Candidate Cap bounds the working set by count: at most N candidates advance, no matter how many exist. The single idea that makes it this mechanism is that admission is relative and numeric — a candidate is in because it ranks in the top N against a fixed ceiling, not because it clears an absolute pass/fail bar (that is an eligibility screen) and not because of what category it belongs to (that is a domain restriction). The cap converts "too many to evaluate" into "the best N we can afford to evaluate," which means its behavior is governed jointly by the ceiling and by whatever ranking or sampling rule fills the slots.
Example¶
A single engineering role draws roughly 400 applications. The hiring committee cannot interview 400 people, so it caps onsite interviews at six. It scores every application against a shared rubric, sorts, and admits the top six; everyone below the line is held on a waitlist rather than rejected outright. The cap is explicit and its basis is stated — "top six by rubric score" — so the boundary can be audited, and if two of the six drop out the waitlist tops the set back up. The evaluation load is now bounded and legible, and the committee argues about the ranking rule rather than about how many people to see.
How it works¶
The protocol has three moves. First, size the space enough to know a cap is even needed and roughly where to set it — dozens, hundreds, thousands? Second, fix N from the evaluation budget: how many candidates the downstream process can actually give real attention. Third, fill the N slots with a ranking or sampling rule, because a cap without a fill rule is arbitrary. The defining feature is that the cutoff is a rank position, not a criterion value — the marginal candidate is the (N+1)th best, whoever that turns out to be, which is exactly why the quality of the ranker determines the quality of the cap.
Tuning parameters¶
- Cap size N — a smaller cap is cheaper but raises the chance a good candidate falls just below the line; a larger cap is safer but costlier. Set it against the per-candidate evaluation cost.
- Fill rule — rank the whole pool, sample it, or impose a diversity quota across slots. Ranking exploits a good scorer; sampling hedges against a bad one.
- Cutoff handling — how ties and near-line candidates at position N/N+1 are resolved, and whether a waitlist is retained.
- Refresh policy — a one-shot fill versus topping up the cap as candidates drop out or new ones arrive.
When it helps, and when it misleads¶
Its strength is a hard, defensible bound on evaluation cost that forces explicit prioritization: the team commits to spending its scarce attention on a deliberately constructed short set — a consideration set[n1] — rather than skimming everything.
Its failure mode is that a rank-based cap inherits the ranker's blind spots. If the scoring rule is silent on some dimension of value, candidates strong on that dimension but weak on the scored ones are cut without anyone noticing, and the loss concentrates just below the line where good and excluded candidates are nearly tied. The classic misuse is treating "top six by our score" as "the six best candidates," laundering a noisy ranking into a confident claim. The guarding discipline is to keep a waitlist and periodically inspect near-the-line exclusions, and to widen N when the ranker is known to be noisy rather than trusting a razor-thin cutoff.
How it implements the components¶
cardinality_estimate— the cap presupposes and encodes a target count judged against the space's size; picking N is a cardinality decision.inclusion_exclusion_rule— the operating rule is "rank at or above N is in, below is held," admitting or deferring each candidate.search_strategy— filling the fixed slots requires ranking or sampling; the cap embeds a search strategy for the bounded set.
It does not draw a temporal window (scope_boundary, reopening_trigger) — that is Bounded Planning Window; and it does not justify a category boundary or review category exclusions (boundary_rationale, false_exclusion_review) — that is Domain Restriction, which bounds by kind rather than by count.
Related¶
- Instantiates: Solution Space Bounding — turns an oversized candidate set into a fixed-size working shortlist.
- Consumes: Search Filter — a filter can supply the reduced, ranked pool that the cap then truncates to N.
- Sibling mechanisms: Bounded Planning Window · Domain Restriction · Eligibility Screen · Finite Horizon Assumption · Sampling Frame Definition · Scope Statement · Search Filter
Editorial Notes¶
Form Classification¶
Form family: Rule, Policy & Commitment
Rationale: Fixes a maximum number of candidates — options, vendors, hypotheses, designs — carried into a cycle, filling the slots by ranking or sampling, making its operative form a standing rule, threshold, contractual commitment, or policy constraint governing future conduct.
Independent corroboration: The frozen evidence defines Candidate Cap as 'Fixes a maximum number of candidates — options, vendors, hypotheses, designs — carried into a cycle, filling the slots by ranking or sampling', so its operative form is Rule, Policy & Commitment.
Review outcome: Independent reviewer agreement; high confidence.
Origin Attribution¶
Primary origin: Operations Research
Origin pattern: Single lineage
Present-day reach: Multi-domain
Rationale: Decision analysis and bounded optimization formalized limiting a consideration set by count and filling it through ranking or sampling under evaluation constraints.
Related originating lineages:
- Behavioral Economics — Consumer-choice research supplies the distinction between the full option universe and the smaller consideration set actually evaluated.
- Organizational & Management Science — Portfolio and stage-gate processes operationalize the cap across repeated selection cycles.
Review resolution: Operations research is primary because the mechanism fixes a cardinality bound and fills it through ranking or sampling under limited evaluation capacity. Organizational shortlisting and behavioral consideration-set research are genuine contributing traditions, but the method is established rather than Encyclopedia-authored.
Review outcome: Reconciled after independent review; high confidence.
Notes¶
[n1] The consideration set — in consumer-choice research, the small subset of available alternatives a decision-maker actually evaluates, as distinct from the full universe of options. A candidate cap is a deliberately constructed consideration set with an explicit size and fill rule. ↩