Chris Matts on the Timing of Commitments, Interview Part 2
Last week, InfoQ talked to Chris Matts about seeing the Agile community as a learning machine, this week InfoQ continues with Chris Matts by focusing on real options and how this strategy for making decisions.
Chris, you talk quite a bit about "do not commit early unless you know why".....can you explain this phrase?
Real Options are really about focusing on the timing of commitments. To achieve a goal, there are many ways to achieve the goal, each of these different ways is an option, a "real option". And each option has an 'expiration date and time', a time after which it is no longer available as an option. In the real world, each option also has an 'expiry condition'.
How about a very quick real-world example of this?
Imagine you are a building contractor with four builders. You are holding a party and are you want to build a swimming pool. It takes one person four weeks to build it, two people take two weeks and four people will take a week. The option to build it with one person expires four weeks before the party. The option to build it with two people expires two weeks before the party and four can build it in one week.
By deferring the commitment until a week before, you can decide not to build the pool if the party is cancelled up until a week before. However there is now the risk that one of the builders will be ill and you are not ready in time for the party. You have three options for building the pool, each expires at a different point with a different risk profile.
"Do not commit early unless you know why" means you do not chose in advance which of the three options to chose..... unless you have a reason which could be anything. For example, your four builders have nothing better to do this week. By starting now, they complete a task that you know you want doing and it means they will be free in the future when they may be needed.
This means the default behavior should not be to start now or as soon as possible. Instead, know when you should start. Too often we see people make emotional commitments and stick to their decisions even though we know there are now better options.
If you have the information you need to remove the uncertainty from the decision process, then you can commit early. For example, you decide to build a pool regardless of the party and you know your builders are free for the next two weeks and are going to be busy after that, then build the pool early.
You talk alot about 'real options' ... to exercise an option, it is imperative that you know you have a choice. Is your message about real options ("choice") resonating in the Agile community? If so, why so? If not, why not?
There are a growing number of individuals who use real options thinking. Its still a very small part of the Agile Community though. I can always tell when someone "gets it" when they apply it outside of IT. One friend used real options to organise their wedding. Another used real options to chose a school for their son. I love it to hear these kind of stories.
I think there are a number of reasons why real options have not really caught on.
- People are put off by the name and think it is about clever financial mathematics. There is no maths in it at all.
- A lot of people are disappointed when they come to a real options session. I think they expect fancy maths and stuff and when its not there. As a result, they dismiss.
- Real Options is quite subtle. There are a few practices but mainly its a set of principles that you apply according to your situation. I tend to avoid speaking about the practices as I want people to think about the general principles rather than focus on the practices.
- Real options are as much about psychology as anything. The way we make decisions is probably one of the core aspects of our personalities. It has a big impact on our behaviour. Asking people to change how they make decisions after an hour long presentation or after reading an article is a big ask.
- People like certainty. They really dislike uncertainty. Real options asks them to embrace something they dislike.
- Most importantly I am appallingly bad at explaining real options. I was working with them for a couple of years before I really discussed them much with anyone else. As such, the way I think about them is hard to explain. In other words, it is my fault they are not better understood.
What is the link between real options and group-level decision making? Isn't choice in software development decision-making simply a capital allocation problem, where the capital is 'developer attention and effort' ??
Real Options are an information hungry decision process. A group of people have more information than an individual. As such, groups should be better at making decisions than an individuals as long as their goal is to make the best decision for the group rather than making the best decision for their personal benefit.
I use real options for the business investment decision process. However there are lots of decisions to be made on an it project. Architecture, choice of development tools, choice of developers. The trick is to get everyone on the team to bring the right information into the group in a timely manner. When we start talking about the timing of releasing information, or of withholding information, we are in the realms of game theory.
If everyone human being on the planet above age 18 has perfect understanding of real options as a tool for decision making, what does that world look like? Is it better or worse-- and why?
Diversity is another way of saying options. Prejudice goes against the principle of making commitments early. We value options, so we would value people who are different. In effect, we would see an end to prejudice and war. I think the world would end up looking a bit like the world of "Diamond Age" by Neal Stephenson.
I think society is moving in that direction already.
What do you get out of your prolific InfoQ commenting habit? You are committing to an option early (you can comment on anything on InfoQ any time), so you must know why. Tell us why.
I get two things out of commenting on articles. One is a discussion about the subject which may reveal further information. The other is to signal to the readers that sometimes the article is discussing a subject that is not as cut and dried as people think. Occasionally you read an article and the tone indicates that the material is "best practice" or generally agreed. In reality the article is an opinion piece and a comment is a good way of indicating their are other opinions.
I think that most people read Infoq articles in the day or two after they are published. Although the articles stay around forever, most of the attention occurs in the first few days. Therefore if you want the best chance for a good discussion of the article you need to get in early otherwise people will not see your comment.
Lots of people know you and you seem to introduce people constantly at these conferences. Sometimes you challenge people in print and in conversation. One thing that is interesting is how you are interested in influencing the leaders by challenging their thinking respectfully from time to time with well-formed arguments. Why do you exercise this option?
The leaders of the Agile community are a constraint to learning. By removing them as constraints, learning will flow more easily. People will find the real leaders of Agile, the people doing it every day for a living who are struggling to solve their problems. Also, the close followers of leaders are voracious learners with the right risk attitude to try new ideas.
I'm not really interested in the queen, but rather in stealing her army of ants.
What's the difference?
Swimming pool analogy is flawed
Re: What's the difference?
Similar to LRM, but not the same. I explain why here -> decision-coach.com/lean-and-real-options/
Keith Adams Dec 06, 2013