Question

Looking across the various reports, I notice the following...

Per User Completion Date: 95% User Completion Date for the user with the latest expected delivery is Nov 5.

Per User Timelines: The 95% probability for work completed date for the same user shows as Nov 13, Nov 14, and Nov 15.

Probability Distribution: 95% Probability date is Nov 22

The Problem

• None of the sliders were changed between any of the reports - they're all looking at the same data.
• The dates should be the same across all three reports.
• The dates are nowhere close to each other.
• Per User Timelines shouldn't be showing a range of dates for 95% complete (or any other %). Every other report gives the impression that the 95% (and every other) is a specific date. Inconsistency in the representation of the confidence points will leads our management to believe that FB is broken. (Alternatively, have all of the other reports show the same range - consistency is the thing!)
• There may be a valid reason why the dates vary across each report. However, having them differ undermines the confidence people place in the tool. I can't tell our executive management to trust the tool but not to trust one report vs. the other.

Case FC1856379

Answer 1

OK. I spoke with Brett. Here's the deal:

There is a bug with the inconsistency between the Per User Completion Date and the Per User Timelines reports. Going back to the example above, the Timelines report can legitimately show 95% probability over the dates of Nov 13 - Nov 15. However, the 95% from the Completion report should be one of those dates. It shouldn't be off by 8 days.

Brett has noted this and opened a bug.

---

Now for the discrepancy between the "User" reports and the Probability Distribution report...

The difference between the 95% completion date of the Probability Distribution vs. the other charts requires a bit more understanding of how things work. It's simple once explained so here we go:

• You look at the Per User Completion Date report and notice the "long pole" or person with the furthest out 95% completion probability is Nov 5 (as in the question)
• You draw the conclusion that the 95% probability of the Probability Distribution chart should be the same. You're wrong.
• The easiest way to why is to look at the Per User Timelines report.
  • If there are other users with trails extending beyond the 95% probability of your "long pole" on the Completion Date report, it will impact the date of the 95% probability of your Probability Distribution report.
  • The probabilities are determined by running a series of simulations. By looking at the Timelines report, you can see that there is a user with a tail extending beyond the 95% of your "long pole" (for this example). Well, this essentially affects the simulations for the overall team figured for the 95% probability on the Probability distribution chart.
  • If this doesn't read clearly, post a comment and I'll do my best to clarify.