The Lego Exercise
June 12th, 2008It’s almost over, FY09 planning season, that is.
It begins in February and is running long this year compared to the last two. In the past 5 months, I’ve been in a variety of teams that have estimated several million dollars worth of projects when you add them together. Because of this work load, I had the good fortune to be invited to a teleconference with 50+ project managers and talk about the process of estimating. Since I only had 15 minutes to talk about a topic that you could spend well over an hour on, I fell back on my old reliable: The Lego Exercise.
I was first introduced to it by a manager 10 years ago and the concept is pretty simple. He took our 8 person team and broke us into pairs. Each pair was given a 50 piece Lego kit whose instructions had been removed. Upon first seeing the box and the pieces inside, each pair was given 30 seconds to estimate how long it would take to assemble the kit with a cap of 20 minutes. If your group came within 2 minutes either way, you got a bonus (I believe it was candy).
How do you accurately estimate how long it takes to complete such an ambiguous task that has an artificial top-down limit to it?
In other words, all the same things that happen when you estimate software projects!
It’s a great exercise and the results are predictable. None of the creations look exactly like the picture on the box, so quality is less than perfect. Some groups complain that their task was harder than others. People cheat when it becomes clear they may finish too early to get the bonus and drag construction.
But then the interesting part happens: What if you did it again?
Take all the kits apart again and swap them with another group. Do the estimation again and you’ll get a much more accurate prediction. Why? The repetition. If you do something more than once, even if it is ambiguously defined, you get better at guessing how long it will take.
This concept, and the suggestion to record the duration of your tasks as you perform them so you have a historical record to look back on later for future tasks, are keys to getting good at operating within the confines of a project schedule.
Are you kidding me? Take a look at my picture. If I’m not a genuine, bona fide nerd I’m not sure who is. I'm currently employed as Portals and Marketing Solutions IT Chief Architect at Hewlett Packard and write here about career best practices for techies. Why? Because I wish I'd had this kind of free advice earlier in my own career and now I'm trying to "pay it forward". See more in
June 15th, 2008 at 6:40 pm
Hi Pete
Just stopping by to say Hi and to let you know that I check your blog every few days to see what you are up to. Your blog and your writing just get better and better.
Have a great week.
Liz
June 16th, 2008 at 9:14 am
As always, thanks for the kind words, Liz.
—Pete