I’m fascinated by things that we’re serially terrible at and estimating is one of them. Judging how long something will take is a tricky proposition for us humans.
We overestimate our competence, underestimate the complexity, rarely factor in the inevitable curve balls and often forget to consider any clean up.
If we repeat something enough, we can get a good general idea of the time it will take, but if it’s a first time, we’re hopeless.
And we fool ourselves about how bad we are. We will often compensate for our poor estimates by working harder, burning the midnight oil, throwing money (or resources) at the situation and still not taking that into account next time we come around to estimating again.
My proposition is that we don’t give up estimating, but that we recognise how bad we are and build some buffer.
And I’d suggest a round number. And by round, I mean Pi.
I like Pi for both the metaphor and the magnitude. The metaphor because our instincts are that the distance we’re travelling is the diameter of a circle, but in reality it’s probably closer to the circumference.
And the magnitude because multiplying my estimates by three (and a bit) has gotten me out of more trouble than it’s gotten me into.
This doesn’t mean that (my estimate x 3.14) is what I tell the person asking, but it’s what I factor in for myself.
The task might only take an hour, and maybe I can delivery it in 90mins by the time I’m set up to tackle it, but I’m probably going to have 90mins of catch-up, dealing with the fallout of dropping everything to pick up this one hour task right now.
As always with heuristics, there are plenty of exceptions and we shouldn’t be dogmatic about its application. But when we’re faced with something novel and have to estimate its scale, we can do a lot worse than taking a slice of Pi.
Leave a Reply