Archive for February 2009

## Hard numbers

Numbers are tricky, especially when you want to use them to understand business problems. A book I read a few years ago and several blog posts more recently have highlighted this for me, so I got the urge to write about the challenges inherent in understanding with numbers and a couple of helpful tips I’ve picked up from the Agile community.

In The Elegant Solution, Matthew E. May devotes a chapter to talking about the importance of picking the right things to measure for your business (Chapter 11, “Run the Numbers”). The Elegant Solution is one of my two favorite business books (together with The Seven-Day Weekend by Ricardo Semler), filled with Lean principles learned from May’s years teaching for Toyota. But while most of the book is quite practical, its practicality kind of unravels in this chapter.

The problem is, the first several examples he gives show just how darned hard it is to get numbers right. It’s hard enough to decide what you want to measure. But the examples go way further, showing how even math PhDs can get correlations and probabilities horribly wrong. The author doesn’t even seem to be trying to make the point of how hard it is to get the math right. He seems to be trying to merely make the point that it’s important to get the math right. It just turns out that every example is more about getting the math wrong.

The first examples he gives aren’t even from business, starting with the Monty Hall problem. This problem evokes heated debates, with very smart people giving different answers and standing by those different answers insistently. The vast majority of people get it wrong, including people with advanced study in mathematics. I got it wrong at first, and it took several tries to finally accept the right answer. (In my defense, I did figure it out before getting out of college!)

In another example, May arguably gets it wrong himself:

You’re playing a video game that gives you a choice: Fight the alien superwarrior or three human soldiers in a row. The game informs you that your probability of defeating the alien superwarrior is 1 in 7. The probability of defeating a human soldier is 1 in 2. What do you do? Most people would fight the human soldiers. It seems to make intuitive sense. The odds seem to be in your favor. But they’re not. Your probability of winning three battles in a row would be 1/2 X 1/2 X 1/2, or 1/8. You have a better shot at beating the alien superwarrior. It’s a simple problem of probability. [Matthew E. May, The Elegant Solution, page 158]

Here’s the thing: If the chance of beating each human soldier is purely random and independent of beating any other human soldier, then the probability of beating three in a row is indeed 1/8. But if there’s any skill involved in the game, then the chances are not independent. When I’m playing a video game, I typically will grow through levels of skill where certain kinds of opponents become easy to defeat. When I start playing the game, maybe I almost never defeat “human soldiers”, but after some time playing I get to a level of skill where I almost always defeat them. So if I can defeat the first soldier, I will likely defeat all three; if the probability of defeating the first soldier is 1/2, the probably of defeating all three is arguably nearly 1/2 as well, which is clearly better than the 1/7 chance of beating the “alien super-warrior”.

I was reminded of this chapter from The Elegant Solution when I saw a flurry of blog posts on a similar problem recently: See Jeff Atwood’s question, his own answer, Paul Buchheit’s response, and the discussion on Hacker News… The problem is just as simple as the Monty Hall problem, and the response it just as heated. Paul Buchheit points out that the simple English statement of the problem can be parsed two different ways, which result in two completely different answers (both of which I verified myself by Monte Carlo simulation!). In another realm, Semyon Dukach suggests that the current financial crisis is due precisely to the difficulty of numerical intuition.

The examples of success with numbers in The Elegant Solution all come down to identifying very simple metrics underlying the businesses in question. Jim Collins gives similar counsel in Good To Great. But the problems mentioned here show that even given a very simple mathematical statement, you can still get into lots of trouble. In The Goal, Eli Goldratt gives some fascinating advice about how to orient your business toward the true goal of business (I won’t spoil the plot by saying what “the goal” is; it really is worth reading the book). But here again, the whole story line of The Goal shows just how unintuitive those principles are, and how long and painful (though valuable) a process it can be to learn them through real-world experience.

So what do we do? A couple pieces advice I’ve picked up over the years might help:

1. Measure, don’t guess. Specify the problem precisely enough to implement a Monte Carlo simulation, and then run it several times. This is the only way I’ve been able to convince myself of the answers to tricky problems like the Monty Hall problem. The more I think about the problem, the less sure I get about the answer. This principle also applies to optimization of code: Start with the simplest thing that could possibly work, and then measure how it performs under realistic conditions. The thing that needs optimization is often not what I guessed it might be.

2. Measure “up”. Mary and Tom Poppendieck talk about this principle in detail in their Lean Software Development books (the phrase is mentioned on page 40 of Implementing Lean Software Development). It is an antidote to local optimization: The more you focus on localized metrics, the more confused you’ll get. As they write, “The solution is to … raise the measurement one level and decrease the number of measurements. Find a higher-level measurement that will drive the right results for the lower-level metrics and establish a basis for making trade-offs.”

What other advice do you keep in mind when numbers get tricky?

## “Agile” versus “agile”

There seems to be a lot of confusion these days about whether something or other is really Agile, and what that means. Here’s my take on how to sort that out.

Growing up, I lived in a city called Grand Rapids. As you’d expect, there’s a river running through it. Does the river actually have rapids, and if so, are they truly grand? Do other rivers have rapids more rapid, or perhaps more grand? Those questions are interesting, but from the perspective of the name, it doesn’t matter. A long time ago, somebody thought that was a good name for the place, and it stuck.

## Capital-A Agile

A decade ago and more, a bunch of people were working on new software processes. They were very different than what had come before, but they all had something in common. It was hard to put a finger on exactly what that was, but eventually they got together and came up with four value statements and twelve principles. And they came up with a single word: Agile. As in “Agile Manifesto” and “Agile Software Development”.

Was this perfect? No. Was it meant to explain everything about software development for all time? No. Was it a software development process on its own? Definitely not. But it was a declaration of common purpose, a list of things they could agree on.

That’s what capital-A Agile is: a bunch of people seeing that they had something in common, and attempting to say what that common thing was. They gave it a name and a partial definition. And most importantly, they formed a community that is still working out what that means and how best to do it.

## Small-a agile

It’s important to note that they weren’t saying that they had an exclusive lock on agility, or even what made software development agile. As with the naming of Grand Rapids, they were pointing at a particular spot in the landscape of ideas and naming it. The word agile has a variety of meanings, and there are a lot of aspects to software development to which you could apply those meanings. They weren’t trying to lay claim to agility as a whole, any more than Grand Rapids is claiming all the rapids in the world, especially the grand ones.

That also means that there are plenty of ways to develop software that aren’t Agile. After all, software got made long before the Agile Manifesto was written. And there are surely ways of being agile that aren’t included in either the Manifesto or in the current practices of the Agile community. Heck, that’s part of why we get together every year, and talk so extensively on line. Processes based on continuous improvement gives you a real taste for continuously improving that process.

## Saying “that’s not Agile”

So given this, what does it mean when somebody says “that’s not Agile”? To me, it just means that the thing they’re pointing at is a different spot in the landscape of ideas.

Some people get upset when they hear that, because they believe they’re doing well at making software, or because they think they’re being pretty small-a agile. They may or may not be right, but that doesn’t matter. If people in the Agile community say that something isn’t Agile, then it probably isn’t, the same way the city of Grand Rapids gets to decide where the city limits are.

If it bothers you to get told that something isn’t Agile, you have three basic choices:

• Find out more about what we mean about Agile. We’re generally a friendly bunch, glad to show you around. Join a mailing list, come to an event, or even ask in the comment box below.
• Persuade us we’re wrong. If Agile is the city we’ve built on the landscape of ideas, it’s a city that’s grown a lot over the years. It in effect started with a number of different little towns growing together. More recently came Lean, but these days it’s getting hard to even tell where the boundaries used to be. We’re very open to new construction, and your idea might be the next big development.
• Start your own thing. Agile may be the big thing of the moment, but it will eventually be as obsolete as the cavalry charge. Just because we say that something isn’t Agile doesn’t mean that it’s not a good idea. If you’re sure of yourself, do what the Agile founders did: stake out your own territory in the landscape of ideas and give it a name.

Regardless, there’s no need to get upset. Having different approaches or coming from different schools of thought doesn’t mean we don’t share the same goals in the end.

## Why estimate?

Much has been written on the subject of how to estimate the effort involved in developing software products. Some Agile teams use Story Points as their estimation units. Others estimate in Ideal Days. Still others use one of these plus Task Hours to help refine their estimates. Planning Poker is a popular method for quickly arriving at group consensus on estimates.

My own earliest attempts at estimating predate the Agile era by a couple of decades. In the beginning I used a technique that has probably been tried (with varying degrees of success) by many others:

1. Break the development effort into as many small tasks as you can think of
2. Estimate each of the tasks in hours
3. Add all the estimates together
4. Apply sufficient “padding” for comfort
5. Hope the boss buys it

Estimation’s main purpose

There’s no doubt that the more recent Agile estimating techniques are much better than what I started out with, and I’m glad that so much effort has been put into developing these methods. Even so, I believe it’s easy to lose sight of the purpose behind estimating, which can be summarized as follows:

The purpose of estimating is to provide the customer with a predictable development schedule.

Predictable is the operative word here. With a truly predictable schedule, the customer can derive the information that is most important from his or her point of view. Given a predictable schedule, the customer can answer important questions, such as:

• How much will this development cost?
• Will we be able to hit the market window?

In other words, the customer is probably not interested in the accuracy or precision of your estimates in and of themselves. The customer is more likely interested in the predictive power of your estimates. Given that, you want to focus your estimating efforts on things that provide predictability.

Duration should be derived, not estimated

I like to encourage teams to focus on estimating User Stories in Story Points, and to discourage them from estimating tasks at all, largely because coming up with Task Hours doesn’t seem to contribute much to the goal of providing the customer with a predictable schedule. Admittedly, estimating Task Hours may help to inform the team about how good or bad it is at estimating duration, but that’s a metric that’s not directly useful to the customer.

What is directly useful is duration that is derived from a team’s demonstrated velocity. Given a velocity in units of Story Points per iteration, plus knowledge of how many Story Points remain in a development effort, the customer can quickly and easily predict the development schedule.