One of the books used by a vocal No Estimates advocate is Black Swans by Nassim Taleb. This advocate continually confuses Macroeconomics of financial markets and sovereign finance with Microeconomics of software development. They are not the same, the processes of decision making are not the same.
Here's a book review from 2008, about Black Swans and Fooled By Randomness in the context of managing software development in the presence of uncertainty
Nassim Taleb’s books make use of the concept of a “Black Swan.” A Black Swan is a large-impact, hard-to-predict, and rare event beyond the realm of normal expectations. The notion of a Black Swan comes from the conception that all Swan’s are white. Once actual Black Swans were discovered in Australia in the 17th century the concept was extended to an “impossibility that has come to pass.”
Taleb speaks in his two books Fooled by Randomness and Black Swans mostly about financial markets, global political events like 9/11. He also uses terms like “long tails,” to describe events with low probabilities of occurrence, but high consequences.
All is good so far until he mentions these concepts are applicable to the management of projects. While watching Brain Lamb’s book discussion show this morning (Tuesday, 1.1.08) Taleb mentioned this again. What struck me was how disconnected Taleb is from the profession of project management. The concepts of Taleb and its application to project management is shared by some in the project management community (and now in the No Estimates community). But as a professional in this community, I’ve come to understand an important flaw in the thinking about Black Swans, Long Tails, and the supposed “randomness” of projects.
Just as a short aside, Taleb is beginning to qualify...
In Martin Gardner’s Fad and Fallacies, he states, "Although we are amused, we may also be embarrassed to find our friends or even ourselves among the gullible advocates of plausible-sounding doubletalk," and establishes the definition of a crank, where such a person possesses:
- A profound intellectual superiority complex.
- Regards other researchers as idiotic, and always operates outside the peer review system.
- Believes there is a campaign against their ideas, a campaign compared with the persecution of Galileo or Pasteur.
- Attacks only the biggest theories and scientific figures.
- Coins neologisms.
- And for those in our agile development community who are advocates of #NoEstimates, a 6th category can be added to Gardner's description - Has a huge dedicated following of “believers.”
Taleb describes himself – in the 3rd person – as an essayist, belletrist, and literary-philosophical-mathematical flâneur. (The No Estimates advocates have a similar set of behaviors.)
One critical concept missed by Taleb and those who follow his concepts while applying them to project management is All Models are Wrong, Some are Useful. The critical success factor of managing projects in the presence of uncertainty is to determine which models are useful.
The fundamental flaw in Taleb’s thesis and those who apply it to project management, (and the conjecture that estimates are not needed in the presence of uncertainty,) is the use of the “Bell Curve,” or Gaussian curve. In the modeling of cost and schedule variables, a symmetric probability distribution function (pdf) belays a lack of knowledge of the underlying behaviors of cost and schedule. Such statistical behaviors – cost and schedule – are never symmetric. This lack of knowledge is different from the lack of data.
Lack of Knowledge is Different than the Lack of Data
When Taleb speaks about “Black Swan” he is speaking (in his own words) about …"large-impact, hard-to-predict, and rare event beyond the realm of normal expectations." These behaviors are by their nature intractable to standard project management processes. They are as described in the papers “On Uncertainty, Ambiguity, and Complexity in Project Management,” Michael T. Pich, Christoph H. Loch and Arnoud De Meyer, Management Science, INFORMS, Volume 48, Number 8, August 2002 pp. 1008–1023 and “Managing Project Uncertainty: From Variation to Chaos,” MIT Sloan Management Review, Winter 2002, in the category of “Chaos.”
Three other categories of uncertainty plus Chaos are: (1) Variation, (2) Foresee Uncertainty, (3) Unseen Uncertainty, and (4) Chaos. The role of the Project Manager in each category is dramatically different.
In the last category – Chaos – and for Taleb’s Black Swans, the Lack of Knowledge is the source of the uncertainty. Any project in which progress is assumed to be made and the “lack of knowledge” is in place is doomed to fail.
For two categories of uncertainty Foreseen Uncertainty and Unforeseen Uncertainty, some form of “mitigation” must be present in order to improve the chances of success. For the Variance that occurs naturally in the execution of a project (cost, schedule, and technical performance variances), attempting to manage these inside their variance boundaries is a simple waste of time. The only solution for the naturally occurring variances is margin.
Acquiring Data Retires Risk
One role of the project manager is to reduce or retire risk. This supposes a risk management system of course. For the moment let’s assume one is in place. One that follows the standard Risk Management process model in use today. With the acquisition of data, the application of that data to the project at hand, the proper categorization of the risks into the four categories above – a risk management process can be put in place that is not driven by the notion of Black Swans driving our behavior.