Probability: Why good projects often go bad

Did you know that on complex projects, even when each of your individual project components/tasks is green, that your overall project is probably red.

How can this paradox be true?

Because basic probability says so. In order to understand the actual probability of a project being completed on time, budget and with quality, you need to assign a confidence of each  dependent task being done and then multiply the confidence levels together.

What you will see is that your probability of success decays exponentially. In other words, if you have a project with two dependent tasks and each task has an 80% probability of success, the overall probability of success is 64% because you don’t average probabilities, you multiply them together.

So even if you are 90% confident that each task will be completed on time, budget and quality, the overall probability of success when you have 9 tasks is 48% and for 20 tasks is 12%.

As a result,

  1. Just because each individual task is green doesn’t mean overall project is green.

  2. Stop averaging your confidence scores and start multiplying them together. Simply convert your colors into percentages, e.g. green = 80%, yellow = 60% & red = 40%, so you can multiply.

  3. All complex projects should start out as red and only go yellow and green after you eliminate your uncertainties, not before. This would eliminate the game of corporate chicken, where different people and teams wait as long as possible before admitting a timeline is at risk, hoping that someone else or another team will fess up first and take the blame. ;=(

  4. Think of forecasting projects as like predicting the weather. It’s easy to do in small chunks for the immediate future, but becomes difficult and ultimately impossible the further out you try, e.g. It’s fairly easy to accurately predict the weather accurately an hour from now in a single location but impossible for either the entire globe or three months from now at a single location with any fidelity.

  5. Plan iteratively, knowing that the more complex/ larger the project, the lower the ability to accurately forecast.

  6. Microsoft Project and other project management tools should have a probability calculator integrated into the tools, so when you looking at the Gannt chart, you are forced to see what the actual probability of meeting the timelines.

  7. Build processes, funding, rewards, etc. that embrace uncertainty as a fact of life rather than something that can magically eliminated through enough hard work and to make executives happy. In Romania, just because they outlawed traffic accidents, didn’t mean that traffic accidents went away, it just meant people avoided reporting them or fled the scene.

  8. Eliminating complexity is a key to success, so question is how to best eliminate complexity? Or structure your complexity in such a way that it minimizes the impact of delays.

  9. Stop thinking linearly and arithmetically. Embrace probability and Complex Systems as core to understanding how the world actually operates.

Leave a Comment

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s