Schedule Risk Analysis: From Risk Register to Finish Date

A risk register tells a sponsor what could go wrong. It does not tell them whether the go-live date survives it. That gap is where schedule risk analysis sits, and it is why a project with a tidy register can still be handed a finish date that had a one-in-five chance of holding.

The direct answer: schedule risk analysis is the step that turns a risk register and a set of duration ranges into a probability of finishing by a date. Each uncertain task gets an optimistic, most likely and pessimistic duration, each risk is mapped to the tasks it touches, and the network is run many times to produce a distribution of finish dates. In Onplana, per the published plan feature list (October 2026), the Monte Carlo version of this is on Business and above.

In short. Schedule risk analysis answers "what is the chance we finish by this date?" Its three inputs are a clean logic network, three-point durations and risks mapped to tasks. Its output is a P50 and a P80 date. A team without simulation software can approximate it by hand on the critical path, as the worked example below shows.

What does schedule risk analysis add to a risk register?

A register ranks risks by probability and impact on a scale. Schedule risk analysis goes one step further and asks which tasks each risk lengthens, by how much, and what that does to the finish date once dependencies pass the delay along.

The two work as a pair. Project risk management covers building the register; this post picks up where the register ends.

What are the three inputs?

1. A clean logic network. Missing predecessors, dangling tasks and hard constraints make the finish date a product of the mistakes rather than the risk. Run the Schedule Health Check first; a network that fails it cannot be risked honestly.
2. Three-point durations. Put an optimistic, most likely and pessimistic estimate on every task that carries real uncertainty, as described in PERT estimation. Leave stable tasks at a single value.
3. Risks mapped to tasks. A risk such as "vendor delivers late" becomes a longer pessimistic value on the specific tasks that wait on the vendor, not a line in a separate document.

The diagram below shows the flow from inputs to the date a sponsor can use.

What does a P80 date mean to a sponsor?

A P80 is the date by which 80 percent of the simulated runs had finished. It does not promise the project will land there; it says the chance of missing it is one in five. Sponsors tend to grasp this faster than a range: the plan date might sit at P20, the P50 is the coin-flip, and the P80 is the date you can sign without hoping. The full mechanics of the S-curve are in Monte Carlo schedule simulation, which owns the simulation itself.

How can you run schedule risk analysis by hand?

You can approximate it on the critical path with arithmetic. This is a screen, not a replacement for a simulation, because it ignores near-critical paths that can overtake the critical one.

Take a four-task critical path, in working days:

1. Compute each mean as (O + 4M + P) / 6, and each standard deviation as (P - O) / 6.
2. Sum the means: 53.8 days. The most likely values alone sum to 50, so the plan date is optimistic.
3. Combine the standard deviations by summing their squares (1.8 + 11.1 + 2.3 + 2.8 = 18.0) and taking the root: about 4.2 days.
4. Read P80 as the mean plus 0.84 standard deviations: about 57.4 days.
5. Check the plan date: 50 days is 0.9 standard deviations below the mean, which is roughly an 18 percent chance of finishing by then.

The honest message to the sponsor is that the 50-day plan has an 18 percent chance, and 57 or 58 days is the date to commit to. Task B alone supplies 62 percent of the variance, so a conversation about B's pessimistic case is worth more than any other change. If your simulation tool also reports how often each task is critical, it ranks the same way.

What does Onplana provide for this?

Onplana runs the Monte Carlo version on its schedule engine: give each task optimistic, most likely and pessimistic durations, run the plan a thousand times, and read the P50 and P80 finish, the chance of your plan date, which milestones slip and which tasks move the finish most. Per the plan feature list (October 2026) that is on Business and above, while the Gantt, critical path and baselines it builds on are on every plan. A schedule you are bringing in from Microsoft Project comes across with its dependencies and constraints through import, so the logic input is already there.

Before any of this, check the network. A model built on a broken schedule produces a confident answer to the wrong question.

Check the logic before you risk it The free Schedule Health Check reads your schedule file and flags the missing links, constraints and dangling tasks that make a risk analysis meaningless. No signup required. → Open the Schedule Health Check