The DCMA 14-Point Assessment has done more than any other tool to give owners and analysts a common vocabulary for schedule quality. Logic, leads, lags, relationship types, hard constraints, high float, negative float, high duration, invalid dates, resources, missed tasks, the critical path test, the critical path length index and the baseline execution index: each is a useful check, and together they catch a great deal of careless work.
They also produce a dangerous kind of comfort. A schedule can pass every one of the fourteen tests and still be unable to forecast anything. Understanding why is the difference between accepting a baseline and owning one.
Structure is not realism
DCMA tests the shape of the network. It can tell you whether activities have predecessors and successors, whether constraints are excessive and whether durations are long. It cannot tell you whether forty days is enough to install the ductwork on level six, whether the pour sequence is buildable, or whether the commissioning window reflects how many systems actually need to be tested.
A schedule can be perfectly linked and entirely fictional.
Five dimensions of a defensible schedule
When we assess a schedule, structure is one test out of five:
- Integrity: complete logic, appropriate relationships, minimal and justified constraints
- Realism: durations derived from quantities, crew sizes and achievable productivity
- Completeness: all scope, including permits, utilities, procurement, approvals and commissioning
- Traceability: alignment with the contract, the estimate and any higher-level program schedule
- Usability: a structure and coding that the field can work from and the owner can report from
A schedule that performs well on all five is one you can manage by and defend later. A schedule that performs well only on the first is a well-formatted estimate of hope.
Before you trust a date, test the schedule that produced it, and then test the people who built it.
Realism takes people who have built the work
Testing realism is a judgment exercise. An experienced planner will notice when slab durations ignore cure and stressing, when MEP rough-in ignores ceiling density, or when a hospital's medical gas certification has been given the same duration as a standard plumbing test. Benchmarks and historical data help; they do not replace someone who has seen the work go wrong.
Omissions are where programs fail
The most expensive defects are usually not errors in what is shown but gaps in what is not: the permit that was assumed, the utility work that belongs to a third party, the owner-furnished equipment with no delivery logic, commissioning drawn as a single bar at the end. A rigorous review asks what is missing as persistently as it asks what is wrong.
Float tells you where to look
High total float across large parts of the network often means missing successors, not genuine flexibility. A critical path that jumps between unrelated areas from update to update usually signals logic being adjusted to suit a date. Negative float that persists without a recovery plan is a forecast that has already been rejected and not yet admitted.
A practical standard for owners
We recommend that owners require an independent health assessment before accepting any baseline, and at defined intervals afterward. Combine automated metrics with expert review of realism and completeness. Rank findings by consequence to the completion date, not by count, and track each one to closure. Reject the baseline, politely and in writing, if the material findings are not resolved.
DCMA tells you whether a schedule is built properly. Experience tells you whether it is true. Owners need both before they announce a date.
