Scrum: The Art of Doing Twice the Work in Half the Time

by Jeff Sutherland

This research has been supported in follow-up studies over the years, confirming that, for example, if someone is good-looking, everyone assumes that they’re also smart and trustworthy.3

“halo”

As with the bandwagon effect, people who are focused on the “halo” don’t look at actual data—rather, they gravitate toward something that has a positive sheen to it.

Again, this isn’t a failure of will; this is the nature of people. Fighting it head-on is silly—it’s like fighting gravity.

“An Experimental Application of the Delphi Method to the Use of Experts,”

In the paper they declared their intention to pose questions without having one person’s opinions affect those of another.

the idea was to ask how many nukes the Russians needed to stop us from making our own nukes. This was back when a nuclear conflict was seen as not only possible, but winnable.

So, in the first questionnaire the number of bombs needed to give a 50-percent confidence in the destruction of the US arms industry was estimated to fall in a range from 50 at the low end to 5,000 at the high end.

Now the range was between 89 and 800.

Eventually the range was down to between 167 and 360 nuclear weapons needed.

It allows them to get a general expert consensus without worries of bias.

The outlook, if you’re interested, was not great.

Planning Poker

Fortunately, there is a way of gathering estimates that is fairly quick and accurate.

“Planning Poker.”

Each person has a deck of cards with those oh-so-interesting Fibonacci numbers on them—1, 1, 2, 3, 5, 8, 13, and so on.

If everyone is within two cards of each other (say a five, two eights, and a thirteen), the team just adds them all up and takes the average (in that case 6.6) and moves on to the next item.

If people are more than three cards apart, then the high and the low cards talk about why they think what they do.

Otherwise they just average the estimates, which will approximate the numbers that the statisticians at the Rand Corporation came up with.

two bedrooms:

three.

living room.

six.

kitchen,

eight,

Close enough, they add them all up, average, and move on to the next task.

This incredibly simple method is a way to avoid any kind of anchoring behavior, such as the bandwagon or halo effects, and it allows the whole team...

“ideal”

But GSI had the idea, which seemed like a good idea at the time, that instead of having each individual team do the estimates, they’d assign the task to the best estimators in the company—the smartest guys in the room who really understood the projects and the technology and knew what needed to be done.

wrong way of going about things that they stopped

The estimates were so far off, they were useless.

It was a complete disaster.

What I took away from that was that only the people doing the work know how long and how much effort it will take.

Forcing them into cookie-cutter processes is a recipe for disaster.

There Are No Tasks; There Are Only Stories

The problem is, if you give any of those items to a separate team that isn’t intimately invested in the results of the decisions between white roses and daisies, you might not get the results you’re looking for.

And because of that, you might provide the wrong kind of data, you might misinterpret the question, or you might just get resentful at being given a bunch of what seems like busywork.

People think in narratives, in stories. That’s how we understand the world. We have an intimate grasp of characters, desires, and motivations.

So the first thing you want to think about when you’re considering a task is character or role—for

Who

what—what

Why

Motivation colors everything.

But the key question not answered in that picture is why.

Those are two very different implementations.

The team needs to figure out what he really wants to do, at which point they might think of a wholly different way to do it, with more relevant information that the captain may not have even thought of but that would be really useful.

Now, if you start that sentence with “As a suburban commuter …” versus “As a farmer in the South Dakota Badlands … ,” you are going to end up with a very different interpretation of what the ideal vehicle is.

you need to define the character, the user, the customer—the

And then you need to understand their motivations.

This will also influence how you’ll estimate things.