The Defense Contract Management Agency's 14-Point Schedule Assessment is the most widely cited schedule health evaluation methodology in federal project controls. Program managers invoke it. Contractors dread it. And yet a surprising number of project controls professionals who live inside EVMS programs cannot explain what each metric is actually testing — or why the thresholds are set where they are.
This post goes through each assessment point, what it measures, and what a violation actually means for your schedule.
What the 14-Point Assessment Is
The DCMA 14-Point Assessment is a standardized surveillance methodology used to evaluate the health and integrity of an Integrated Master Schedule (IMS). It does not assess whether the project will finish on time or on budget. It assesses whether the schedule is structurally sound enough to be trusted as a management tool. A schedule that fails multiple 14-Point checks is not necessarily a late project — it is a project whose schedule cannot be relied upon to give an accurate picture of where the project stands.
DCMA applies this assessment to contractor-submitted schedules on programs subject to government surveillance, typically those governed by DFARS clauses or DOE Order 413.3B.
The 16 Metrics (Yes, 16)
Despite the name, the current assessment involves 16 discrete metrics across 14 categories. Here is what each is measuring:
1. Logic (Missing Logic) — Tasks with no predecessor or no successor relationships. Threshold: less than 5% of activities. A task with no predecessor can start at any time without constraint. A task with no successor produces no downstream impact on the network. Either condition means the schedule is not a true network — it is a collection of disconnected activities that cannot accurately model the program.
2. Leads — Tasks with negative lag on a predecessor relationship (a finish-to-start relationship where the successor starts before the predecessor finishes). Threshold: 0%. Leads introduce optimism into the network that is not grounded in real physical work sequencing.
3. Lags — Positive lag durations on predecessor relationships. Threshold: less than 5% of relationships. Excessive lags are often used to create artificial schedule buffers that obscure the true critical path.
4. Relationship Types — Percentage of finish-to-start relationships versus start-to-start and finish-to-finish. Threshold: greater than 90% finish-to-start. Non-finish-to-start relationships are not inherently wrong, but high concentrations indicate a schedule that has been tuned rather than built.
5. Hardcoded Dates (Date Constraints) — Tasks with imposed date constraints that override network logic. Threshold: less than 5% of activities. Hardcoded dates break the network's ability to propagate delay. When a constrained task's predecessor slips, the schedule does not reflect the impact downstream.
6. High Float — Tasks with total float greater than 44 working days. Threshold: less than 5% of activities. High float often indicates missing logic, disconnected network segments, or tasks that have been isolated from the critical path to avoid scrutiny.
7. Negative Float — Tasks with negative total float. Threshold: less than 5% of activities. Negative float means the task is already projected to finish later than its imposed end date. Any program with widespread negative float is reporting that it cannot meet its contractual schedule under the current baseline.
8. High Duration — Activities with original durations greater than 44 working days. Threshold: less than 5% of activities. Long-duration tasks are difficult to status accurately. A 200-day task reported at 50% complete has provided almost no useful information about whether the work is actually on track.
9. Invalid Dates — Actual dates that fall outside logical bounds (actuals in the future, forecast dates in the past). Threshold: 0%. Invalid dates indicate poor schedule maintenance discipline or deliberate manipulation of progress reporting.
10. Resources — Percentage of activities with resources assigned. Threshold: greater than 90%. An unresourced schedule cannot accurately model resource constraints, contention, or loading. It cannot produce a reliable critical path for resource-limited programs.
11. Missed Tasks — In-progress tasks with an actual start but no actual finish that are currently past their baseline finish date. Threshold: less than 5%. Missed tasks directly measure slippage that has already occurred but has not been formally acknowledged in the schedule.
12. Critical Path Length Index (CPLI) — The ratio of the critical path length plus total float to the critical path length. Threshold: greater than or equal to 0.95. CPLI measures schedule health relative to the contractual end date. A value below 0.95 indicates insufficient float to absorb further slippage and still meet the milestone.
13. Baseline Execution Index (BEI) — The ratio of activities completed to activities that were baselined to be complete by the data date. Threshold: greater than or equal to 0.95. BEI measures whether the team is executing against the plan. Consistent BEI degradation is one of the clearest early warning signals of program underperformance.
14/15/16. Schedule Performance Index (SPI/SPI(t)/TCPI) — EVM-derived performance metrics that tie schedule health to earned value performance. These connect the 14-Point structural checks back to the financial performance baseline.
What Passing Actually Means
A schedule that passes all 14-Point checks has been demonstrated to be structurally sound. It has network integrity, reasonable logic coverage, no widespread constraint violations, and performance metrics that are within acceptable bounds. What it has not been demonstrated to have is accurate data. A schedule with perfectly clean DCMA metrics, built on manipulated float values, fabricated milestone completions, or retroactively adjusted logic, will pass every 14-Point check while being fundamentally misleading.
This is the limitation of any point-in-time compliance check. It validates structure. It cannot validate truth.
Peveka Solutions' Forensic Intelligence Engine runs deterministic DCMA 14-Point assessments and cross-validates findings against contract language and cost data — because structural compliance and data integrity are different problems.