Using Probabilistic Linking in Risk Management

Probabilistic branching and probabilistic linking are two completely different concepts in risk management. Many risk management practitioners confuse these two and use them interchangeably. This is not correct. To understand probabilistic linking, first you need to understand probabilistic branching. You can refer to this detailed article to understand probabilistic branching.

I hope you have gone through the linked article. Probabilistic branching informs the probability of an activity existing or not. On the other hand, probabilistic linking allow you to model the probability of a link between two tasks existing or not existing. 

In this article, we will use a simple example and will follow some steps to understand how probabilistic linking can be there in a schedule. The content of this article is taken from my new course and book on Risk Management Professional (RMP) exam: 

RMP Live Lessons, Guaranteed Pass or Your Money Back

Now, let’s see the steps to perform a probabilistic linking and analysis.

Step – 1: Create the tasks and Link the tasks

Let’s create two tasks: 

• Task A. It’s of 3 days duration.
• Task B. It’s of 6 days duration.

I’m going to use the Primavera Risk Analysis (PRA) to build this simple plan. Next, we are going to select both these tasks and right click, which will give us the option to link. We are going to use the finish-to-start (FS) link. This is depicted below. 

The linked tasks will come as shown below. 

Step – 2: Add probability values to the links

Next, we are going to select Task B and go to the Risk and Uncertainty tab under the Task Details pane. Under this tab, we will select the “Probabilistic Links” sub-tab. As shown below, for Task B, the link is with respect to Task A. It’s mentioned as the preceding link. 

Here, we will add linking probability with Task A as 75%. In other words, the probability or chance of Task A being linked with Task B is 75%. Obviously, we will have 25% left to allocate. This is shown below.

Do note that there is another tab called Probabilistic Branch tab, just left of the Probabilistic Links tab. As noted in the beginning of the article, these two are completely different concepts.

Step – 3: Add another probabilistic link

Next, we are going to add another task of Task C. For Task C, we have the following operations:

• Task C is of 5 days duration. 
• Task C is linked to Task B and it’ll be the predecessor to Task B. The dependency will be again finish-to-start (FS). 
• In other words, Task B has two predecessors now: Task A and Task C.

This is represented in the below figure.

Next, select Task B (not Task C) and click on the Probabilistic Links tab. Here, give the probability or chance of link between Task B and Task C as 25%. The linked values given are shown below.

As shown above:

• Task B now has two preceding links: Task A and Task C.
• Task A’s linking chance is 75%, whereas Task B’s linking chance is 25%.
• Together it equals 100% or as noted in the above figure, total % left to allocate for probabilistic linking is now 0%.

Step – 4: Run a risk analysis

Next, we are going to run a risk analysis with the PRA tool. This can be done by simply using the command Run Risk Analysis from Risk menu list in the toolbar.

As you step through the analysis you will see the links disappearing and the effect on the plan. The below figure shows Iteration 1 out of 1000 and with it the link between Task B and Task C is not considered. 

As shown above, while Task B is actually connected to both Task A and Task C, in Iteration 1, only the link between Task A and Task B are considered! 

As you proceed with the Risk Analysis step after step, the links will keep on changing, i.e., one link will disappear and another one will appear. 

The above figure shows for Iteration 3, the link between Task B and Task C is considered, whereas the link between Task A and Task B has been removed. 

Step – 5: Complete the risk analysis

As your plan moves through multiple iterations, the probabilistic linking and its value will keep on changing for the tasks in the schedule. If you have applied probabilistic distribution for the activities or tasks in your project, then you run a complete Monte Carlo or Latin Hypercube simulation. You can learn a brief on probability distribution for tasks in this article. 

I’ve intended to keep the example very simple to make you understand the difference between Probabilistic Branching and Probabilistic Linking with a real-world software tool. 

Conclusion

In probabilistic linking, the preceding links for a concerned task can have a probability of 100%. When two or more preceding links are there for a task with less than 100% probability, then all of those probabilities must add up to 100%.

If you are preparing for your Risk Management Professional (RMP) exam, you need to understand all these:

• Probability Distribution
• Probabilistic Branching
• Probabilistic Linking

Predominantly, your questions will be on probability distributions. However, you can expect a few questions on probabilistic branching and linking as well.

References:

[1] New Course: RMP Live Lessons, Guaranteed Pass, by Satya Narayan Dash

[2] New Course: RMP 30 Contact Hours Online, by Satya Narayan Dash

[3] Book: I Want To Be A RMP, The Plain and Simple Way, Second Edition, by Satya Narayan Dash