Re:When Agile jumped the shark
by bradly on Sunday July 15, 2012 @01:35PM
(#4247473)
My big problem with estimating in story points is that it ignores the reality that how long something takes is directly tied to who is doing the work. A junior eng, a senior eng, someone who just started, someone who isn't familiar with that piece of the system, the person who wrote and has always maintained that piece of the system. These people will all do a given task in a different amount of time. Possibly an order of magnitude differently when comparing a junior eng not familiar with the code to the codes author. This is okay. Embrace it. Plan for it.
[ Parent
| on HN ]
I would argue that this is actually an advantage of story points. You don't know who will be working on it at estimation time. But a 2-pointer should take any developer about twice as long as it would take the same developer to do a 1-pointer.
If you have many super fast, very senior developers you get more points done and have a higher velocity. That's where the speed will be reflected and you can see how much your team can get doen in one iteration. This allows you to plan your iteration without having to care who works on any given story.
On a sufficiently large project (more than a few people) the differences in productivity average out, as do to some extent estimation errors. I never expect any individual estimate to be correct; it's more important to get an accurate estimate of how far off your estimates are in aggregate.