productmanager

I find it helpful to keep in mind that the traditional statistical significance test is a statement about a conditional probability. i.e. it's the probability of the data given the hypothesis (the null hypothesis). But what many actually want is the probability of the hypothesis given the data. Sometimes these are referred to as the frequentist vs. bayesian approach. There's a helpful recent podcast here by with author of Trustworthy Online Controlled Experiments https://music.youtube.com/podcast/hEzpiDuYFoE

agree.. technically the velocity should be constant and more or less is.. but the story estimates still establish the building blocks for this.. a story seems to need tasks and task estimates to make the overall story estimate more accurate. Tasks help to break up activities. alternative is to groom the story of course, then you're left with a larger list of stories to manage however, and taking care of that in turn eats into our PM team's velocity :)

We've been using Pivotal, but are about to switch over to Jira completely. Tried Trello. At end of day Trello and Pivotal are lacking 2 key features we need for better predictability in scrum, particularly on enterprise software with many projects: tasks with t-shirt/time estimates, and burndown charts. Without these, the PM team has a hard time estimating a few months out what is possible.