In Black Swans, White Swans, They Are All Swans, I used a figure for parts per million (PPM) at six sigma z=6.0. My excel formulas produced this value at 3.4 PPM. Pretty low by not zero. William Nicholas (Software Engineering Institute) commented that this number is common but wrong. See the comments section in the post.
William provides some history of the 3.4 and links to better calculators.
- Four Mile Labs Z Calculator
- Curious Cat's Management Improvement site
- Balanced Enterprise Solution's The Six Sigma System
These are now in my "reference list," and I'm deleting my lame attempt to solve the closed form integral equations in Excel.
In the domain I work, the Black Swan is still an excuse for not doing your homework as a risk manager. I'm not "The" risk manager, but there are Risk Managers on our manned space flight, nuke power, and biopharma programs that we work. In the non-mission critical, "we don't kill people by accident or on purpose" world of software development, Black Swans represent real threats, since it likely takes too much time and costs too much money to look for them. Be rest assured they are there, hiding in the bushes, blending in with the foliage, just waiting to jump out a derail those types of projects, or the financial plans motivated by greed. They are there, they are not Unknown Unknowns. No matter Rumsfled says. Taleb has changed his approach and is now using Bayesian Networks as the mechanism for dealing with being "fooled by randomness," after he got hammered by the mathematical economist - see Signal Extraction and Non-Certainty Equivalence in Optimal Monetary Policy as an example..
I've been in two start-ups. One failure the other went public and is still in business - Triconex. I still remember the words of the VC guy - Chuck Cole. "You boys will be funded for production when greed overcomes fear." That's the 3rd definition of a Black Swan in the Black Swans, White Swans, They Are All Swans post. Thanks William.
