NoEstimates
--
What’s the problem just discussing how big a story is, the problem is the focus, why is valuable to give a number to a user story?
Story points are an invention of XP (extreme programming), in XP, stories were originally estimated in time. They used to use ideal days, but no day is ideal, so they go a step further, and they abstract it by points.
Managers
Managers are interested in story points because they are ultimately used to create a forecast, managers want to give a date to their bosses. This will be done the next week, they want to predict the future, but in some sense this is normal, we are in our society obsessed with the future, to anticipate it.
So for managers they represent a tool to have that forecast, to answer the question, to report, sometimes to decide if we do this now or later.
So at the end they will translate in someway what means a story point in days for that team, using any statistical methods. Story points will be converted to working days, that’s it. Some people will be tempted to use story points per developer or even calculate how many story points are done by me or you in that iteration.
We usually say, can we commit to this number of story points this iteration?. Commit is a hard work, it is scary and make devs feel pressure.
Do we really want to create this environment just because we want to hit a forecast?.
But if we need a forecast, why not using any tool as good as estimations?. For example, why not using the user stories done per week as a way of forecasting the future?. You could say you cannot because user stories are not of the same size, and yes, this is the point.
Developers
The problem I have seen with story points is that we focus too much the conversation in the number, not in splitting the stories. They have converted in a tool to communicate two worlds, the world of the thinkers the ones that create the stories and the world of the doers the ones that write the code. They tend to avoid collaboration.
If a story is too big, we need to split it in smaller user stories, if something is smaller enough let’s do it if this is the next more important thing to do. But splitting it in smaller stories is a game for everyone…