Many of us put a lot of effort into reducing uncertainty, but sometimes figuring out the unknown is not going to change the decision we make.
In this second post in my series on reducing uncertainty, I focus on how your decisions are impacted by what you don’t know.
The question to ask yourself is “how much does the unknown really matter in this case?”
When you’re driving across town on a well-known route, do you usually use GPS? On the one hand, live traffic forecasts can be helpful if it’s an unlucky day with a crash in your way. On the other, some of us are creatures of habit and like to take our usual route even if there is a delay.
To see how uncertainty and your decisions are connected, ask yourself how much of a delay it would take to make using GPS worth it. You might even realize that no matter how bad traffic is, like on some city streets in the suburbs, it’s basically impossible to run into that much of a delay.
There is a useful graph called a decision tree that can help you figure out whether the effort to eliminate the uncertainty is worth it. Like my recent post on learning curves, I’ve modified the standard decision tree a little bit to represent the information I care about.
Figure: A decision tree. At the top is the initial choice to use GPS or not. Then after you’ve made your decision, there’s an unknown delay on your drive. The next layer asks “would you change your path if you knew there was a delay?” If not, or there’s no delay, you don’t need GPS.
Typically, the information you put towards the bottom are the outcomes – what we get after we have made a sequence of decisions and uncertainties unfolded. Following that standard I would have put some numbers like a “5 minute delay” in certain boxes.
But in this case we’re not all that interested in the specific delays since our first question is “is it even possible the delay would change our decisions?” To capture this question in the graph I added a second decision “Would you change your path if you knew there was a delay?”
To use this decision tree to decide on using GPS or not, you just need to ask yourself about the blue and pink highlighted boxes. For the drive you’re on, and the delays you might face, would you do anything differently if you knew there was a delay?
Now, I’ve been acting up to this point like all you need to know is whether a delay would change your mind. Some people really like to have their GPS guiding them, and others find it mildly annoying. Do you really need to hear five directions for how to drive out of the neighborhood you’ve lived in your whole life?
That moment of annoyance is part of our outcomes. Aha, a tradeoff!
To make the choice of whether to use GPS or not, you have to decide the marginal cost of using it every time versus the marginal benefit of the information you gain.
If GPS annoys you, and you will rarely or never change your route, it would make no sense to use it. Once you’ve figured out your breakeven point, you now have a policy for making the decision to use GPS or not.
For me, that policy is “am I driving on I-25 in Denver? (i.e., a highway)” If I am, GPS is worth it because there are frequent, and substantial, delays. But more importantly, I usually choose to change my path to avoid them.
