Assessing the Likelihood of a Risk in Your Project
4 of 9 in Series: The Essentials of Managing Risk in Your Project
As you refine your project’s schedule, you need to assess the various risks that you have identified. Determining the likelihood of each risk will help you manage risk if the worst happens. The first step in deciding whether to deal proactively with a risk is assessing the likelihood that it will occur.
Use one of the following schemes to describe the chances that a risk will occur:
Probability of occurrence: You can express the likelihood that a risk will occur as probability. Probability is a number between 0 and 1, with 0.0 signifying that a situation will never happen, and 1.0 signifying that it will always occur. (You may also express probability as a percentage, with 100 percent meaning the situation will always occur.)
Category ranking: Classify risks into categories that represent their likelihood. You may use high, medium, and low, or always, often, sometimes, rarely, and never.
Ordinal ranking: Order the risks so the first is the most likely to occur, the second is the next most likely, and so on.
Relative likelihood of occurrence: If you have two possible risks, you can express how much more likely one is to occur than the other. For example, you can declare that the first risk is twice as likely to occur as the second.
If you have objective data on the number of times a risk has occurred in similar situations in the past, use the first scheme in the preceding list to determine the likelihood that the risk will occur again in the future. If you don’t have objective data available, use one of the other three schemes that are based on personal opinion to describe the likelihood particular risks will occur.
Relying on objective info
You can estimate the probability of a risk occurring by considering the number of times the risk actually occurred on similar projects. Suppose, for example, that you designed 20 computer-generated reports over the past year for new clients. Eight times, when you submitted your design for final approval, new clients wanted at least one change. If you’re planning to design a computer-generated report for another new client, you may conclude the chances are 40 percent that you’ll have to make a change in the design you submit (8 divided by 20, then multiplied by 100).
When using objective information, such as past project reports, to determine the likelihood of different risks,
Consider previous experience with similar projects.
Consider as many similar situations as possible.
Keep in mind that the more similar situations you consider, the more confidence you can have in your conclusions.
Counting on personal opinions
In the absence of objective data, solicit the opinions of experts and people who have worked on similar projects in the past. You can estimate the likelihood of a particular risk by soliciting the opinions of ten people who have worked on projects similar to yours.
For example, ask them to rate the likelihood of a specific risk as high, medium, or low. Suppose six people choose high, two choose medium, and two choose low. You may then develop your estimate of the likelihood by assigning values of 3, 2, and 1, to high, medium, and low, respectively, and determining the weighted average of the responses as follows:
(6 × 3) + (2 × 2) + (2 × 1) = (18 + 4 + 2) / 10 = 2.4
This formula suggests the risk has medium to high likelihood of occurring (because 2.4 is between 2 and 3).
To increase the accuracy of the likelihood estimates you make based on personal opinions, try the following:
Define the category name as clearly as possible.
Consider the opinions of as many people as possible.
Be sure the projects your respondents have worked on are truly similar to yours.
Don’t allow people to discuss their estimates with each other before they share them with you.
After they’ve submitted their initial estimates to you, consider having the people discuss their reasons for their estimates with one another and then asking them whether they want to revise their estimates.