Working on a chicken: using animals for relative estimation

Some of you might know about this training method I use when I try to explain relative estimation. I start by explaining that it is really hard for human beings to do actual estimation. As an example I ask people to come up with the exact weight of a lion. We really suck at estimating this. But when I ask people whether a lion weighs more or less than a goat, we all know that a lion weighs more. I repeat this several times with different kind of animals. I also tell people that this is the whole reason why we decided to also use this relative estimation within estimating Product Backlog Items.

In training I also explain the (almost) Fibonacci sequence which we use for this relative estimation: 0, 1/2, 1, 2, 3, 5, 8, 13, 20, 40 , 100 so called story points.

This week somebody from the client I currently work for really had a great idea: Let’s just keep using the animal kinds, instead of the story points. That really is a brilliant idea. So now we have a backlog full of animal-sized stories. We have chickens, goats, lions, elephants, mice, whales and we use this to explain how complex we think a certain Product Backlog Item is.

Next time I train people in the use of relative estimation, I will not even tell them about the fibonacci sequence; We will just start off with working on some chickens…..