Three-Point Estimate with Base Rates¶
Estimation model — instantiates Reference-Class Planning Calibration
Replaces a single-number estimate with an optimistic, most-likely, and pessimistic triad in which the likely and pessimistic legs are pulled to comparable-case base rates.
A single-number estimate hides everything that matters about a task's uncertainty; a three-point estimate — optimistic, most-likely, pessimistic — exposes the spread. Three-Point Estimate with Base Rates is the small estimation model that improves the classic triad by anchoring its legs to comparable-case history rather than to feeling: the optimistic point is the team's inside-view best case, but the most-likely and — especially — the pessimistic points are dragged toward what similar tasks actually did, so the triad's width reflects real base rates instead of the estimator's imagination. Its defining move is producing a per-task distribution whose shape is disciplined by outside-view evidence, then combining it (via a weighted formula like PERT) into an expected value and a spread that can roll up across a plan. It is a task-level estimator; it does not maintain the case library, run a project-wide governance gate, or catalog discrete named risks.
Example¶
A publisher is scheduling the localization of a technical book into six languages and needs a defensible timeline per language before committing to a launch. For each language the managing editor builds a three-point estimate. Take the translation-and-review step: the inside-view optimistic case, from the lead translator, is 4 weeks. Left alone, an ordinary triad would set "most likely" at 5 and "pessimistic" at 6 — all three anchored to the same optimism. The base-rate discipline breaks that: pulling from the publisher's records of comparable technical-book localizations, the most-likely lands at 6 weeks and the pessimistic tail — where terminology disputes and a second review cycle recur — stretches to 11, because that is what actually happened in similar past titles. The three points now span 4 / 6 / 11, and a PERT weighting yields an expected duration near 6.5 weeks with a visibly right-skewed spread.[n1] The optimistic 4 is still on the page — the inside view is preserved, not deleted — but it no longer is the estimate, and the honest tail is what feeds the schedule.
How it works¶
- Capture the inside-view optimistic point. The team's own best case is recorded as the optimistic leg — preserving local expertise as an explicit, inspectable assumption rather than discarding it.
- Pull the likely and pessimistic legs to base rates. The most-likely and pessimistic points are set (or moved) toward the actual outcomes of comparable past tasks, so the triad's width is evidence-anchored, not imagined.
- Weight into an expected value and spread. A formula such as PERT's
(O + 4M + P) / 6combines the three points into an expected value, with the pessimistic leg giving the distribution its skew and tail.[n1] - Roll up across tasks. Because each estimate is a small distribution, per-task estimates aggregate into a plan-level range instead of a single summed date.
Tuning parameters¶
- Base-rate pull strength — how hard the likely and pessimistic legs are dragged toward class history versus the estimator's feel. A strong pull corrects optimism but can override a genuinely faster local method; a weak pull respects local knowledge but lets optimism back in.
- Weighting formula — classic PERT weighting versus a more tail-heavy scheme. Heavier pessimistic weighting produces more conservative, tail-aware expected values but can over-pad routine tasks.
- Spread width — how far apart optimistic and pessimistic are set. A wide spread is honest about uncertainty but weakens the estimate's usefulness for tight scheduling; a narrow one is actionable but risks false confidence.
- Elicitation of the three points — whether the triad comes from one estimator or several. Multiple estimators widen and de-bias the points but cost coordination.
When it helps, and when it misleads¶
Its strength is being the lightweight, task-level entry point to reference-class discipline: it needs no formal database or governance gate, just comparable-case memory and a formula, and it fixes the single worst habit — the point estimate — while still keeping the inside view visible as the optimistic leg. Because each estimate is a distribution, it composes naturally up a work-breakdown structure into a plan-level range.
It misleads when the three points are all sourced from the same optimism: a triad whose "pessimistic" leg is just the optimistic guess plus a token margin looks like uncertainty quantification but is inside-view estimation in disguise — the base-rate anchoring is the entire value, and skipping it hollows out the method. It also invites false precision, since a tidy expected value can imply more rigor than three soft guesses deserve. The classic misuse is generating a precise PERT roll-up from optimistic inputs — the archetype's own non-example of a confident simulation built on a flattering base. The guarding discipline is to source the pessimistic leg explicitly from comparable-case outcomes rather than from imagination, and to treat the expected value as a spread with a real tail, not as a new single number.
How it implements the components¶
inside_view_assumption_trace— the optimistic leg preserves the team's own best-case reasoning as an explicit, recorded assumption rather than erasing it.outside_view_adjustment_rule— the likely and pessimistic legs are pulled toward comparable-case base rates, moving the estimate off pure optimism by a defined rule.uncertainty_interval_and_tail_frame— its output is a three-point spread with a skewed tail, so the task carries a range rather than a point.
It does not select or bound the comparable class it borrows base rates from (reference_class_boundary), nor maintain the case library — those belong to the Reference-Class Forecasting Workbook and the Historical Project Outcome Database. Where the workbook runs the full class-selection procedure for a headline forecast, this model just spreads a single task's estimate into three base-rate-anchored points.
Related¶
- Instantiates: Reference-Class Planning Calibration — a lightweight task-level estimator that carries uncertainty as a base-rate-anchored triad.
- Consumes: Historical Project Outcome Database — supplies the comparable-case base rates that anchor the likely and pessimistic legs.
- Sibling mechanisms: Reference-Class Forecasting Workbook · Historical Project Outcome Database · Schedule and Cost Risk Register · Contingency Reserve Formula · Independent Estimate Round · Launch or Commitment Readiness Gate · Premortem as Auxiliary Probe · Forecast Backtesting Review
Editorial Notes¶
Form Classification¶
Form family: Analysis, Modeling & Optimization
Rationale: Three-Point Estimate with Base Rates operates as an analytical, modeling, inference, comparison, or optimization procedure that derives insight or a solution because it replaces a single-number estimate with an optimistic, most-likely, and pessimistic triad in which the likely and pessimistic legs are pulled to comparable-case base rates.
Independent corroboration: The frozen evidence defines Three-Point Estimate with Base Rates as 'Replaces a single-number estimate with an optimistic, most-likely, and pessimistic triad in which the likely and pessimistic legs are pulled to comparable-case base rates', so its operative form is Analysis, Modeling & Optimization.
Nearest alternative: Representation, Specification & Plan — Three-Point Estimate with Base Rates includes features of a static representation, map, specification, schema, or prospective plan that externalizes information, but its defining operation is an analytical, modeling, inference, comparison, or optimization procedure that derives insight or a solution.
Review outcome: Independent reviewer agreement; medium confidence.
Origin Attribution¶
Primary origin: Statistics & Experimental Design
Origin pattern: Cross-disciplinary synthesis
Present-day reach: Universal
Rationale: Three point estimate with base rates derives most directly from statistics' measurement, sampling, inference, and experimental-design tradition; its defining operation is to replaces a single-number estimate with an optimistic, most-likely, and pessimistic triad in which the likely and pessimistic legs are pulled to comparable-case base rates.
Related originating lineages:
- Data Science & Analytics — Data science's telemetry, modeling, profiling, and monitoring tradition provides a formative adjacent lineage for the same three point estimate with base rates operation.
- Economics & Finance — Economics, finance, and mechanism-design practice supplies a parallel or contributing lineage for the mechanism's defining operation: replaces a single-number estimate with an optimistic, most-likely, and pessimistic triad in which the likely and pessimistic legs are pulled to comparable-case base rates.
- Mathematics — Mathematical modeling, proof, and abstract-structure practice supplies a parallel or contributing lineage for the mechanism's defining operation: replaces a single-number estimate with an optimistic, most-likely, and pessimistic triad in which the likely and pessimistic legs are pulled to comparable-case base rates.
- Operations Research — Operations research, optimization, and queueing analysis supplies a parallel or contributing lineage for the mechanism's defining operation: replaces a single-number estimate with an optimistic, most-likely, and pessimistic triad in which the likely and pessimistic legs are pulled to comparable-case base rates.
Review resolution: Both blind reviewers independently select statistics_experimental_design as the primary historical origin for the concrete operation—Replaces a single-number estimate with an optimistic, most-likely, and pessimistic triad in which the likely and pessimistic legs are pulled to comparable-case base rates. The queued differences concern alternate origin disagreement, origin mode disagreement, domain reach disagreement, encyclopedia synthesis disagreement, not the primary lineage. I retain every alternate that either reviewer explains, without a numeric cap, and choose origin_mode=cross_disciplinary_synthesis because the reviewers' combined evidence identifies material construction from multiple disciplines. domain_reach=universal records later portability rather than multiplying historical origins; confidence=high is the conservative shared evidentiary level, and encyclopedia_synthesis=true preserves either reviewer's affirmative synthesis finding.
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] The PERT (Program Evaluation and Review Technique) three-point estimate combines an optimistic (O), most-likely (M), and pessimistic (P) value into an expected duration, classically (O + 4M + P) / 6, approximating a beta distribution whose spread and skew come from the gap between the optimistic and pessimistic legs. ↩a ↩b