Every project that is scheduled using the Critical Path Method is exactly as long as its longest path, its duration determined by the sum of that path's activities, lags and constraints. Therefore:
That path is called the critical path.
Any delays on that path make the project longer. By contrast, a delay on a shorter path will only make the project longer if it is so great as to make a different path the longest path. (The amount a non-critical activity can be delayed without switching critical paths is called its float or slack.)
To shorten a project's duration, you have to shorten the current longest path. And if you get a new critical path, then you have to shorten it to compress the schedule further.
Critical path drag is the amount of time that a scheduled activity or lag is adding to the project's duration. Or, alternatively, it is the amount by which the project's duration would be shortened if the duration of any individual activity or lag were eliminated (i.e., reduced to zero).
Whereas float is always off the critical path, drag is only on the critical path. In other words, only items that are on the critical path delay the end of the project. When looking to compress a schedule, it's very helpful to know which items offer the most opportunity simply because they are adding the most time!
When using the Critical Path Method to schedule complex projects we can compute a quantity known as drag. Drag describes the impact that individual tasks have on the overall project duration due to their durations. For instance, a critical task with a duration of 3 months may have a drag value of (close to) 3 months.
This is often not a simple calculation. A critical path activity with a duration of 15 days may only have drag of 1 day—eliminate it entirely and the project would only be one day shorter! Elsewhere in the schedule, another activity with a duration of 6 days may have drag of 3 days. Find a way to reduce its duration by half, and the schedule will be compressed by 3 times as much!
When looking to shorten a schedule of thousands of activities, with a critical path of several hundred, schedulers need this critical information to point them to opportunities! Some software packages have been computing critical path drag for 15 or more years. But Primavera, the most popular software in the world for large projects, only computes float! Its CPM algorithm doesn't compute critical path drag, to tell the user which activities are adding how much time!
One more note: The tool may sometimes report that two or more paths have neither float nor drag. This is likely because both paths are exactly the same length, and we therefore have parallel critical paths. (This often occurs due to what's called “pacing” the schedule: allowing non-critical paths to use up the full extent of their float. It's usually fairly easy to eliminate that “padding” on the “artificial” critical paths.)
In such cases, shortening any ONE of the critical paths will not shorten the whole project. However, by shortening one at a time, you will eventually have one critical path that has built up quite a bit of drag! Now shorten that one, and the schedule will often be considerably compressed!
The exact value of a task's drag depends on the rest of the schedule. (For the astute reader, the following excludes the possibility of reverse critical paths, which we'll circle back to later.) There are a few different cases:
If the task has positive total float, then it is not on any critical path. If we reduce its duration to zero, the project duration will not change. In this case, the drag value is zero.
If the task has no total float, then it is on at least one critical path.
If it is not on all critical paths, then reducing its duration still does not change the project duration, since some other critical path will remain unchanged in length.
If it is on all critical paths, then reducing its duration will reduce the project duration. However, it may be that setting its duration to 0 causes some other path to become critical. We can calculate whether this is the case, by looking at the total float of parallel tasks (tasks which are neither descendants, nor ancestors, of the task we're calculating drag for). If some parallel task exists with a total float of say, 3, then we know that the drag on this task cannot exceed 3. We therefore can compute drag as the minimum of the task's duration, and the minimum total float of any parallel task.
To gain intuition for how drag is calculated, let's look at the example schedule in figure 1. In the figure, a schedule and its duration are displayed. Below the schedule we see the list of tasks, with a button for each task that sets the task's duration to 0, runs CPM, and computes the resulting duration. We then record those figures into the table, and therefore we can calculate the drag for each task in this tiny example schedule.
Project duration:
10
days
Task Name
Drag Calculator
Project duration w/ normal activity duration
Project duration w/ 0 activity duration
Drag Value
A
10 days
Click button to calculate
Click button to calculate
B
10 days
Click button to calculate
Click button to calculate
C
10 days
Click button to calculate
Click button to calculate
D
10 days
Click button to calculate
Click button to calculate
Figure 1: Our baseline schedule with four activities. The activities A, B, C, and D have durations of 3, 5, 2 and 4 days respectively. There are finish-start constraints between A → B as well as B → C, and a start-start constraint between B → C. The highlighted precedence relationships in the chart are active for the current schedule. The project start date is set to 2024-01-01. Activities can only be dragged to new start dates if there's no precedence relationship that would be violated by such a change.
Why drag matters, and how to use it
Drag is a powerful tool for project managers. It allows us to understand the impact of individual tasks on the overall project duration. By understanding drag, we can identify tasks that, if sped up, can reduce the overall duration of the project. The way we recommend using drag is as follows:
Evaluate the drag values of the activities in your schedule using the calculator above
Inspect the drag values you get, and identify any issues with the schedule that are causing the drag values to be unexpected. There is more information about this in the section on reverse critical paths. The tool helps you identify those by listing a reason for the drag value being what it is. Concretely:
An activity may have no drag because it is not critical (i.e. it has float)
An activity may have no drag, even though it is critical, because some parallel activity exists which has no float
An activity may have a drag value of its duration (or less than its duration), depending on any parallel activities in the schedule
Having evaluated and studied the drag values, you can now use them to look for way to speed up your project. The drag value tells you how much shorter the project can become by reducing the duration of the activity, so you would then focus on ways to speed up exactly those activities with (highest) drag. This can be done by increasing resource limits, finding new technical ways to execute the activity and so forth.
A problem: reverse critical paths
As briefly mentioned before, an issue can arise when computing drag values that has to do with reverse sections of critical paths. An example schedule that has a critical path with a reverse section is shown in figure 2.
Figure 2: A simple schedule which contains a reverse critical path through activity B. Due to the FF relationship between A and B, B becoming shorter makes the overall project completion date later!
Are you in the infrastructure or construction industry? We would love to help with your AI & tech use cases. If there is a good match with our roadmap, we will build a solution to prove our value