Understanding the Risk Analysis Results

After the risk analysis process completes, you can view a variety of statistical results. Available options may vary depending on the settings you selected before running the analysis. To ensure you have the most current analysis information, view the analysis start date, end date, and duration in the Risk Analysis Run Settings section. For information on Risk Iteration Analysis, see Risk Iteration Analysis Overview.

Note: When Use Convergence is selected, the risk analysis statics panel displays Maximum Iterations and the risk analysis runs until the specified convergence criteria are met, or until the analysis reaches the specified maximum number of iterations. For more information, see Understanding Convergence. You can view the Risk Analysis Log to see iteration details.

Distribution Results

When a risk analysis is run, the Remaining Cost, At-Completion Cost, Duration, Start Date, and Finish Date are recorded on both the activity and the project as a whole. A visualization of Schedule & Cost data together is recorded for the project. For each iteration, the risk may or may not occur according to its probability. If the risk does occur, the schedule impact and cost impact are sampled from their thresholds and applied to the project or activity. If an activity has uncertainty applied, the remaining duration is also sampled from within the uncertainty distribution before risk impacts are calculated. Non-working time due to weather risks will only affect the finish date of an activity, not the duration.

Each tab, excluding the Schedule & Cost tab, displays a probability distribution that represents the various cost, duration, and date values the project or activity could have. For each cost, duration, or date range, the graph displays its frequency and cumulative frequency, with horizontal lines aligned with the cumulative frequency. From the distribution, you can determine the likelihood of completing the project or activity within a certain finish date or cost. A color-coded legend at the top of the page identifies key statistical values based on Response Context, and you can hover over data points, bars, and indicator lines for more specific information. Details next to the graph display additional key statistics and analysis settings. The distribution graph displays durations, costs, and dates as configured in user preferences.

The Schedule & Cost tab displays a scatter plot graph to better visualize the likelihood of completing a project within a certain cost and finish date. See Working with the Risk Analysis Distribution Results Scatter Plot for more information.

Mean Impact

A risk mean impact analysis is useful to determine which risks and activities have the greatest average impact on your project duration and cost. The results of the mean impact analysis are presented in tornado charts, which rank each activity or risk according to their potential impact on the project duration or cost.

There are four views in the chart that allow you to focus on either the Contribution towards Project Cost or Project Delay, and then breakdown either by Activity or by Risk. This information can help you determine which risks you may want to address first, or which activities whose estimates you may want to revise in terms of duration or costs. The chart will also include other sources of delay and cost, so that all contributions to the Project Mean Cost and Project Mean Delay can be seen in the chart. For example, in the Activity Cost Contribution and Activity Delay Contribution, the contribution from Risks that are not assigned to activities is included as a single bar called 'Unassigned Risks'. In the Risk Cost Contribution and Risk Delay Contribution, you can see contributions from Activity Uncertainty and Other Cost increases such as WBS Summary and Level of Effort Activities.

You can choose to view results as values, percentages, or both. Hover over a bar to display impact information including activity name, impact value, and risk name.

Risk Removal Impact

A risk removal impact analysis, also called a sensitivity analysis, is useful to determine which risks can have the greatest impact on your project. During the impact analysis, the system simulates the complete removal of each risk to measure and rank the effect the risk has on the Pessimistic Finish and Pessimistic Remaining Cost of the project. The analysis also simulates the removal of all activity uncertainty values and measures the potential Pessimistic Finish impact on the project. The results of the risk removal impact analysis are presented as a tornado chart, which ranks the risks that have the highest potential negative or positive impact on the project.

Each risk is represented in the chart by a bar stretching between the lowest and highest time or cost estimates for the risk. Each bar indicates the degree to which the removal of a risk can positively or negatively affect the project Pessimistic Remaining Cost or Pessimistic Finish date. A bar representing the removal of all project activity uncertainty values is also included in the Finish Date chart. This information can help you quickly visualize which risks can have the most impact on your project's cost and schedule targets.

By default, every risk in the register is included in the risk removal impact analysis. You can select risks to exclude from the risk removal impact analysis on the Risk Register page by selecting Exclude from Risk Removal.

For Level of Effort and WBS Summary type activities, weather risks are ignored and not included in the risk analysis. The schedule impact of a risk mapped to Level of Effort and WBS Summary type activities is ignored and not included in the risk analysis. However, the cost impact of weather risks and risks with a schedule impact is still included in the risk analysis for Level of Effort and WBS Summary type activities.

Threat, opportunity, and weather type risks are all included in the risk removal impact analysis. While threat and opportunity risks support both pre-response and post-response analysis contexts, weather risks do not. The same weather data is used when the application runs simulations for both contexts. However, due to randomly seeded numbers used during the analysis calculations, the pre-response and post-response results for weather risks may differ. If this difference is significant, consider increasing the number of iterations run during the analysis.