<!-- argument-hint: [best practice number, critical path, total float, schedule risk analysis (SRA), resource loading, duration, traceability, baseline, statusing, status date, forward/backward pass, EVM, four characteristics, audit, chapter number] -->
GAO Schedule Assessment Guide (GAO-16-89G, 2015)
Source: US Government Accountability Office (GAO) — US Government work, public domain | Chapters: 8
When to use
Use this skill when you need to build, explain, assess, or critique a program's integrated master schedule (IMS) and judge whether its forecast dates can be trusted: applying or auditing the ten best practices, computing or interpreting a critical path and total float, running or reviewing a schedule risk analysis, resource-loading and setting realistic durations, statusing against a controlled baseline, or interrogating someone else's schedule with GAO's audit questions. This is GAO's cross-government best-practice authority on schedule reliability — the time-domain companion to its Cost Estimating and Assessment Guide and to the agency SE handbooks (NASA, DoD).
Prerequisites: none — plain Markdown; no MCP server, API key, or licence tier needed at runtime.
How to Use This Skill
- Without arguments — load the core frameworks below for an overview of the ten best practices, the four characteristics, CPM, and schedule risk analysis.
- With a topic — ask about a best practice, the critical path, total float, a schedule risk analysis, resource loading, durations, traceability, the baseline, statusing, the forward/backward pass, or EVM scheduling.
- With a chapter — ask for
ch02 (capture & sequence), ch03 (resources & durations), ch04 (traceability & critical path), ch05 (float & risk analysis), ch06 (updating, baseline & four characteristics), ch07 (auditing), or ch08 (EVM, CPM arithmetic & reference data).
Supporting files: glossary.md, patterns.md, cheatsheet.md.
Core Frameworks & Mental Models
The reliable schedule (the thesis)
A program's success rests on an integrated master schedule (IMS) — a single, dynamic network covering all effort (government, contractor, subcontractor, vendor, external) from program start to finish. The schedule is the operating instrument of program management: a route map, a progress yardstick, an early-warning system, the basis for a time-phased budget, and the foundation for cost/schedule/scope tradeoffs. The through-line of the whole guide: the reliability of the schedule determines the credibility of the dates the program forecasts — every downstream decision inherits the schedule's reliability (or lack of it).
The Ten Best Practices
GAO's central framework: ten practices "associated with a high-quality and reliable schedule," stated in no particular order and explicitly not a step-by-step build recipe.
- Capturing all activities — every WBS element (all parties) appears as at least one activity.
- Sequencing all activities — justified predecessor/successor logic so the network forecasts dates automatically.
- Assigning resources to all activities — labor, material, equipment, facilities, travel loaded and checked for feasibility.
- Establishing the duration of all activities — durations derived from effort/resources/productivity, not from a desired finish date.
- Verifying horizontal and vertical traceability — logical hand-offs across activities; consistent dates/status/scope across schedule levels.
- Confirming the critical path is valid — the path that truly drives the finish, recomputed each cycle.
- Ensuring reasonable total float — float that is plausible, not a symptom of broken logic.
- Conducting a schedule risk analysis — simulate duration uncertainty (Monte Carlo) to price the completion date and size contingency.
- Updating using actual progress and logic — status the schedule to a real status date so the forecast stays true.
- Maintaining a baseline schedule — a CM-controlled target the current schedule is measured against.
The Four Characteristics (the ten practices roll up into four)
GAO's top-level synthesis — a reliable schedule is:
- Comprehensive — captures all activities, resources, durations (BP1, BP3, BP4).
- Well-constructed — sound sequencing, valid critical path, reasonable float (BP2, BP6, BP7).
- Credible — traceable and risk-informed (BP5, BP8).
- Controlled — statused and measured against a controlled baseline (BP9, BP10).
These four are the audit lens: is everything in it? is the logic sound? does it trace and is it risk-informed? is it kept current and controlled?
Critical Path Method (CPM)
The schedule is a calculable model. The forward pass (EF = ES + duration − 1) and backward pass (LS = LF − duration + 1) deterministically yield early/late dates, total float (TF = LS − ES), and the critical path — the longest continuous chain that sets the earliest finish. Total float is both the measure of slack and a diagnostic: unreasonable float (large positive or any negative) usually means missing or invalid logic, not real room to maneuver. Near-critical activities sit just above critical-path float and can become critical with a small slip. When constraints, multiple calendars, out-of-sequence progress, or leveled resources distort float, trace the driving (longest) path back from the finish milestone instead of trusting the software-flagged path. "Critical" is precise — it means drives the finish date, never "important" or "long."
Schedule Risk Analysis (SRA)
Durations are forecasts, not facts, so the program duration is uncertain — and summing deterministic durations runs optimistic because of merge bias (converging paths multiply, not average). An SRA feeds quantified risks and uncertainties into a Monte Carlo simulation to produce an S-curve of completion dates, from which the program picks a confidence-priced date, sizes schedule contingency, and ranks the risks driving the exposure. Inputs are three-point durations and risk drivers (combine both). Contingency is PM-owned reserve under change control — categorically distinct from total float.
Statusing and the controlled baseline
A schedule earns trust only when it is continuously reconciled with reality and measured against a controlled commitment. Statusing loads actuals to a status date (the actuals/forecast boundary) and updates by remaining duration, not typed-in percent complete; out-of-sequence progress is handled with conservative retained logic. The baseline is the configuration-managed target, backed by a living schedule basis document; performance is the variance between baseline and current schedule — and that variance is only as valid as the schedule is reliable. Limit baseline changes to scope change or formal replanning; frequent rebaselining is a red flag.
Chapter Index
| # | Section | Key content |
|---|
| ch01 | Introduction & the Reliable Schedule | Why schedules matter, the IMS as focal point, schedule reliability thesis, the ten best practices overview, CPM/total float/near-critical, planning vs. scheduling, the cyclic reliability process, rolling-wave planning |
| ch02 | Capturing & Sequencing Activities (BP 1–2) | WBS as cornerstone, capturing all parties' work, milestones vs. detail activities, F–S logic, forbidden logic types, constraints/lags/leads, path convergence, dangling logic |
| ch03 | Resources & Durations (BP 3–4) | Resource loading, effort vs. duration interdependence, BOE reconciliation, resource leveling within float, deriving durations from effort, calendars, inch-stones (QBD), durations sized to the reporting period |
| ch04 | Traceability & Critical Path (BP 5–6) | Horizontal (hand-offs) and vertical (level-to-level) traceability, roll-up into the program schedule, validating the critical path each cycle, driving (longest) path, keeping the path clean, near/risk-critical paths |
| ch05 | Total Float & Risk Analysis (BP 7–8) | Total/free/independent/interfering float, reasonable-float judgment, merge bias, Monte Carlo SRA & S-curve, three-point durations vs. risk drivers, contingency vs. float, remove-one-and-rerun prioritization |
| ch06 | Updating, Baseline & Four Characteristics (BP 9–10) | Statusing vs. revising, status date, remaining-duration updates, retained logic vs. progress override, baseline & basis document, BEI/trend analysis, recovery/acceleration techniques, the four-characteristics mapping |
| ch07 | Auditing a Schedule (App. I–II) | GAO methodology/provenance, the Appendix II audit triad (Key Questions → Key Documentation → Likely Effects), coupling of practices, documentation as proof, constraints as logic-overrides to justify |
| ch08 | EVM, CPM Arithmetic & Reference Data (App. III–IX) | EVM scheduling considerations & ANSI/EIA-748 mapping, forward/backward pass math, date-constraint names & effects, quantitative health measures (no tripwires), industry/agency comparisons, the 15 artifacts to request |
Topic Index
- As-built schedule / fragnet → ch06
- Audit triad (Key Questions / Documentation / Likely Effects) → ch07, patterns, cheatsheet
- Baseline / baseline schedule → ch06, cheatsheet
- Baseline Execution Index (BEI) / trend analysis → ch06
- Best practices (the ten) → ch01, cheatsheet
- Calendars → ch03
- Constraints (date) / lags / leads → ch02, ch08, cheatsheet
- Contingency (schedule reserve) → ch05, cheatsheet
- Critical path / critical path method (CPM) → ch04, ch08, cheatsheet
- Driving path (longest path) → ch04, cheatsheet
- Durations (establishing) → ch03
- Earned value management (EVM) scheduling → ch08
- Float (total / free / independent / interfering) → ch05, cheatsheet
- Forward / backward pass → ch08, cheatsheet
- Four characteristics (comprehensive / well-constructed / credible / controlled) → ch06, ch01, cheatsheet
- Horizontal traceability → ch04
- Integrated Master Schedule (IMS) → ch01, ch02
- Industry/agency comparisons (DCMA 14PA, NDIA PASEG, NASA, DHS) → ch08
- Level of effort (LOE) → ch02, ch04
- Merge bias → ch05
- Milestones → ch02
- Monte Carlo simulation → ch05
- Near-critical activities → ch01, ch04
- Out-of-sequence progress (retained logic vs. progress override) → ch06
- Path convergence → ch02, ch05
- Planning vs. scheduling → ch01
- Quantitative schedule health measures → ch08
- Rebaselining / replanning → ch06
- Recovery & acceleration (crashing, fast tracking) → ch06
- Resource loading / leveling → ch03, cheatsheet
- Rolling-wave planning → ch01, ch03
- Schedule basis document → ch06
- Schedule narrative → ch06
- Schedule risk analysis (SRA) → ch05, cheatsheet
- Sequencing / logic → ch02, cheatsheet
- Status date (data date) → ch06, cheatsheet
- Statusing vs. revising → ch06
- Tells & smells (unreliable schedule) → cheatsheet, ch07
- Total float (reasonable) → ch05, cheatsheet
- Vertical traceability → ch04
- Work breakdown structure (WBS) → ch02, ch01
Supporting Files
- glossary.md — key scheduling terms (IMS, CPM, total float, SRA, the four characteristics, status date, baseline, merge bias), alphabetical, with chapter references
- patterns.md — seven reusable patterns (capture from the WBS, sequence with sound logic, resource-load & set durations, validate the critical/driving path, run an SRA & size contingency, status & control the baseline, audit with the Appendix II triad) with When/How/Trade-offs
- cheatsheet.md — the ten best practices table, four-characteristics mapping, CPM arithmetic, logic/constraint rules, statusing & SRA quick rules, tells & smells, quantitative health & industry comparisons
Scope & Limits
Covers: GAO best practices for schedule reliability per GAO-16-89G — the integrated master schedule as the focal point of program management; the ten best practices (capturing, sequencing, resourcing, durations, traceability, valid critical path, reasonable total float, schedule risk analysis, updating, baseline maintenance); the four characteristics (comprehensive, well-constructed, credible, controlled); critical-path-method arithmetic (forward/backward pass, total float, the driving path); schedule risk analysis (Monte Carlo, merge bias, contingency); statusing and configuration-controlled baselines; EVM scheduling considerations; and the Appendix II audit instrument (Key Questions / Key Documentation / Likely Effects).
Does not cover in depth: cost estimating — the dollar side (see gao-cost / GAO-09-3SP, which this guide defers to for EVM and cost detail); the full systems-engineering process (see dau-se-guidebook, nasa-se-handbook, nasa-npr-7123); agile schedule assessment (see GAO's separate Agile Assessment Guide); risk management as a discipline (the guide consumes a risk register for the SRA but is not a risk-management framework — see nasa-risk); technology readiness (see gao-tra); and detailed operation of any specific scheduling tool (the guide is tool-agnostic and names tool behaviors only as cautions). The guide is also thin on portfolio/multi-program scheduling and on procurement/contracting mechanics beyond what the IMS must integrate.
Jurisdiction: US Government public domain work (GAO). The best practices are cross-government and adoptable by any organisation; they are guidance, not binding requirements outside an agency that adopts them.