Like Santa in the North Pole, I’ve been working hard in my workshop, preparing the new release of Statistical PERT which I’ll be tentatively releasing on Monday, January 2, 2017.
Initially, Version 2 of Statistical PERT (Normal Edition) was just going to offer users an easy way to estimate using just a single point-estimate and heuristics to generate the minimum and maximum point-estimates, which are necessary to create an implied, bell-shaped probability curve. But as I worked to prepare for Statistical PERT – Beta Edition, which uses Excel’s two beta distribution functions (BETA.DIST and BETA.INV), I found myself doing a lot more to the original Statistical PERT spreadsheet (which is now dubbed Statistical PERT – Normal Edition, since it uses Excel’s two normal distribution functions, NORM.DIST and NORM.INV).
The biggest change, philosophically, is that when I release Version 2 of SPERT-Normal, there will only be one example workbook and one template, rather than the six example workbooks and six templates that are available under the current 1.4 version. Instead of creating special “flavors” of Statistical PERT for task duration, expenses, revenue, agile sprints, event attendance and portfolios, there will only be one “flavor.” This will make it easier for me to release new versions of Statistical PERT without the added effort of updating multiple example workbooks and templates. But the good news is that Version 2 will make it easy to estimate any bell-shaped uncertainty by letting estimators choose whether they want to view the left-side or right-side area of the implied probability curve, and whether they want to see currency formatting or not in all the different cells.
So for task duration, for instance, most estimators will want a high-probability estimate with 70% confidence or higher. Version 2 of SPERT-Normal lets you obtain cumulative probability estimates from the left-side area of the implied distribution curve. But if you wanted to estimate revenue, and you wanted a conservative revenue estimate, you might want to choose an estimate that has a very high probability of being exceeded. In that case, you want would want to see the cumulative probability from the right-side area. In Version 2, you can toggle between the left-side and right-side area, so you can either see the probability of a SPERT estimate being equal to or greater than an uncertainty (the left-side area), or you can choose to see the likelihood of a SPERT estimate being exceeded by the uncertainty (the right-side area).
Aside from that change, here’s what else is new in Version 2 of Statistical PERT – Normal Edition:
- A separate worksheet for entering 1-point estimates that will use heuristics to auto-create a 3-point estimate
- A new toggle that lets you choose whether to use currency formatting or not
- New conditional formatting for the sum of all uncertainties
- Removal of hidden columns that aren’t needed
- Improved visual layout and formatting throughout
- New hover comments for each column’s heading (some worksheets)
Statistical PERT – Beta Edition won’t have its Version 1 release until later in 1Q-2017, but the January release of Version 0.6 will be vastly improved over all preceding development versions, and will look very similar to SPERT-Normal Version 2. And I’ll make SPERT-Beta available on the main site instead of only being available on the Beta Blog.
So, look forward to a great new year with great new versions of Statistical PERT!