Thursday, May 17, 2018

Gambling on Projects


Originally Published May 14, 2018 on LinkedIn
Tom Sheppard
5/7/2018
When most people hear the name Monte Carlo, they think of casinos, gambling, secret agents (James Bond), and royalty. For project managers, however, Monte Carlo is all about a mathematical simulation used to to get estimates for the completion of project tasks, and estimate probabilities in the face of uncertainty.
Monte Carlo simulation leverages principle of the law of large numbers to improve the strength of what amounts to a guess. Applying known constraints, or limits, and averaging results across many scenarios allows a project manager to use a Monte Carlo simulation to get numbers which are very likely to be close to what will actually happen.
TechTarget.com defines the law of large numbers as, "given enough trials or instances. As the number of experiments increases, the actual ratio of outcomes will converge on the theoretical, or expected, ratio of outcomes."
So, a lot of runs gets us closer to the most probable actual results. 
In project management, the Monte Carlo simulation is one way a PM can use to predict the probable duration of a given task, when the actual time (or effort) needed is not known.  
For most projects, using this sophisticated approach is overkill. Most often using a weighted average of estimates provided by experts is a very fast, practical and accurate means.  
The weighted average is based on the best-case, (Be) worst-case (We), and probable-case (Pe) estimates provided directly from subject matter experts (SMEs) who will be actually performing the work and who have performed similar work in the past. This produces a task duration (or effort) (D or E), which, I have found, is highly likely to be meet or exceed the actual duration.
D=(Be+4∗Pe+We)÷6
This approach works because the uncertainty is reduced dramatically by the experience of the SMEs. If they have not actually performed the same sort of work, then you may be better off using their Be and We as upper and lower bounds for a Monte Carlo analysis because, in that case, the experts and the computer both have the same amount of actual experience, but the computer can run many iterations and give you a more statistically probable answer.
About the Author: Tom Sheppard specializes in managing large ($10mm+), high-risk, high-profile projects for the US financial services sector. He is the author of The Art of Project Management, available in hardback from Barnes and Noble, or in paperback and ebook format from Amazon
(c) Copyright 2018 Thomas K Sheppard. May be used with attribution.

4 Basics Every Project Management Entrepreneur Needs to Get Right

Photo Credit: Unsplash 4 Basics Every Project Management Entrepreneur Needs to Get Right By Gloria Martinez of WomenLed.org   When y...