97 Agile Ideas - Idea Number 18
The Planning Fallacy
If we make a list of tasks and estimate how long that list of tasks is going to take, our estimates will be near the most optimistic amount of time possible. It's very difficult to cure ourselves of this "bias." Expertise and prior knowledge on its own isn't likely to cure us of the optimism.
There are ways of thinking that can cure use of this optimism: we can imagine that the individual, or team that's doing the work isn't us, but some other individual or team.When it isn't us that's doing the work we seem to be able to use the distance to give a more realistic estimate. Another way is to focus on what Daniel Kahneman, who discusses this idea in his book "Thinking Fast and Slow," calls "The outside view." The outside view is a view that sees how long a similar project actually takes.
Kahneman tells the story of working on a text book project. Four years into the project the team working on the book met up. There was some concern about how long the project was taking so Kahneman asked everyone on the team to estimate how much longer the project would take (in years) and write the number down secretly. When all the numbers were collected, the average score was three. Meaning that the total length of the project would be seven years.
Most of the team, including Kahneman found this quite depressing. Until Kahneman had an idea - why not ask the most experience member of the team if that estimate seemed realistic. The most senior member of the team (I think he was called Seymore) thought for a moment and then said, "No, I don't think it's realistic, most projects like this actually take about ten years. But then again half of them stop without finishing." This wasn't quite the uplifting answer that Kahneman had been looking for so he dug further. "Is our team better than the average team?" Seymour thought for a while before he grudgingly admitted, "No, this team is slightly below average." As Kahneman comments. This was maybe the time, having found out that the project would likely take another five years and had a fifty percent chance of failure, to just get up and leave. But he didn't and in due course, a few years later the project was cancelled.
What's perhaps most remarkable about this story is the estimate that Seymore gave for how long the project would take, when he'd been asked in isolation, rather than having to focus on the outside view. Seymour had estimated another four years. Near the average for the group and nowhere near his estimate when forced to consider the "outside view." The planning fallacy is strong, and it's always with us.