Not Everything Deserves a Decision: Homer, the Inbox, and Getting Things Done

Mark C. Otto and MinJae Lee 

Homer Simpson has been making bad decisions for more than 30 seasons. He picks the wrong button at the nuclear plant, backs the horse whose name he cannot spell, and does it all again next week. The show is comedy precisely because Homer’s decisions are terrible. Meanwhile, the creative team behind him is one of the best-organized creative shops in television. Something must be answering the small questions well enough that the big ones ever get answered at all. Statisticians can borrow the team’s methods for the small decisions that fill the rest of our lives.

In a 2016 episode of David Allen’s Getting Things Done podcast, three of the show’s writers—Rob LaZebnik, Dan Greaney, and Danny Chun—describe their creative-by-committee process. What they describe is not decision analysis. It is more granular, faster, and closer to what most of us spend our days on. It is organizing decisions that require little thought.

Up to this point, this series has been on decisions that deserve a structured framework. Frame the problem, list your objectives, generate alternatives, estimate consequences, and work through the tradeoffs discussed last month.

In a 2004 Decision Analysis essay titled “Making Better Decision Makers,” Ralph Keeney offered a canonical account of how our decisions are resolved. Of every 10,000 decisions Keeney figures we face, 7,000 are of small consequence, another 2,000 are no-brainers, and the next 750 can be resolved by clear thinking. That leaves 220 that require parts or all of the SDM process and 30 that require Keeney-level decision analysis.

David Allen’s Getting Things Done, or GTD, is a discipline for the 9,750. We statisticians often carry the mental load of everything unfiled—every promised follow-up, unanswered collaborator question, half-remembered idea, incomplete task, and upcoming deadline. This load is what makes even the today’s handful of decisions feel exhausting. The decisions themselves are not the problem. The other several thousand items, silently consuming cognitive bandwidth, are.

First, Allen observes that the human mind is very good at generating ideas and very bad at storing them. Anything you try to store in your head will bother you at the wrong time. You remember the cat food at 3 a.m., not at the pet store, or that brilliant turn of phrase while you are in traffic.

Second, the level of nagging is not tied to the importance of the task. You need to buy cat food and replace your programmer, and they bug you just as much, at equally inconvenient times. You have to get these out of your head, regardless of their size.

The GTD move is to get every unfiled item out of your head and into what Allen calls an inbox—one physical or digital location, or a small set of them, where every incoming item lands. Not to act on it. To hold it until you can process it deliberately. Your head is for having ideas, not for holding them. The inbox holds them.

GTD has five steps for processing your stuff: capture; clarify; organize; review; and engage. The clarify questions are key: Is it actionable? If not—if it is reference material, equipment, decoration, or supplies—it gets stored. For actionable items, we ask, “Does it take less than two minutes?” If yes, do it now; the tracking overhead outweighs the benefit of doing it later. If no, the item is either delegated (with a note on your Waiting For list) or deferred to your Projects or Next Actions list.

In GTD, a project is anything that requires more than one step. You need two things: (1) what the finished project looks like, and (2) what the very next physical action is.

We statisticians are trained to plan analyses from top to bottom. Be a Markov process; take the first step; discover the rest.

Organize is where the material goes. Reference goes to where you can find it again. Projects go to a projects list. Next actions are sorted by context—the place, tool, or state of mind you need to be in to do them. @computer. @office. @home. @phone. @errands. You do not want to look at a list of 40 next actions when you are standing in the grocery store; you want the four items that are @grocery. Contexts turn the day into a series of batch-processed queues.

Engage is the payoff. When you have a few minutes, you look at the context you’re in and pick from the list what fits your time and energy. No re-deciding. No re-remembering. The thinking is already done. The meeting hasn’t started; make a few calls from @phone.

The review step separates people whose systems work from those whose systems don’t. Once a week, at a time you protect, go through the whole apparatus. Collect any loose papers into the inbox. Empty your head onto the same page: What have you been carrying around that hasn’t been filed? Get email to zero. Look at your Next Actions list and mark off what is done. Review the past two weeks of your calendar for anything that generated a follow-up. Review any future event that needs preparation. Check the Waiting For list and the people you are waiting on. Review Projects to make sure each still has a next action. Review Someday/Maybe and cross off the things you have stopped wanting or get creative with the ones you do.

Statisticians recognize this as calibration to keep your system from drifting into untrusted territory. Otherwise, Homer is back in your head, thinking about cat food. That is the horizontal workflow. The GTD vertical system is the Horizons of Focus, which ties into the rest of this series.

At ground level is what you are doing right now: the next action. At 10,000′, the current projects. At 20,000′, your Areas of Focus—the roles you play: parent, partner, data masseuse, committee chair, homeowner, friend. At 30,000′, your one- to three-year goals. At 40,000′, the vision of what the world looks like if your work succeeds. Then, at 50,000′, the mission—what you are doing with your life. These are the strategic objectives from last month’s article. The elevation analogy helps you go up for perspective and down for more concreteness and action. You don’t need to get lost either in the weeds or the clouds—move up and down deliberately.

Consider what this suggests about a particular kind of institutional failure. Volkswagen’s emissions scandal, Boeing’s 737 MAX, Wells Fargo’s fake accounts, and Twitter’s various collapses were not caused by a single catastrophic decision at the top. They were caused by thousands of small decisions at the ground level that were disconnected from any mission at 50,000 feet. When the ladder is broken, ground-level decisions optimize for ground-level pressures—quarterly numbers, next quarter’s promotion, next week’s demo—and the accumulation eventually shows up as a public disaster. Big problems are made from small decisions.

One more note, aimed at anyone senior enough to have felt the following pain: The higher you rise in an organization, the more of your day is taken up by interruptions.

This is not a distortion of the job. It is the job. Interruptions are what a leader is paid to catch. Everything routine is managed below. When someone appears in your office with the next crisis, you set your current work aside in the inbox, deal with the crisis, and return to the inbox when it is done. There is no lost train of thought. It survives in your inbox, and that next interruption may just as well be an opportunity.

A research study can be viewed as a GTD project: define what “done” looks like—an answered scientific question; a completed analysis; a paper, report, or other client deliverable—and identify, not the next intellectual action, but the next physical one. “Open the data file and run summary statistics.” This is familiar territory for collaborative statisticians, helping investigators turn broad scientific questions into clearly defined, estimable quantities.

And the weekly review is what any well-run analysis pipeline needs: Check the queue, catch the drift, confirm the next step, and move on.

Mark Otto

Mark C. Otto is a retired, but not retiring, US Census Bureau and US Fish and Wildlife Service statistical scientist. He is chair of the American Statistical Association Committee on Applied Statisticians and has served as the president of the Washington Statistical Society and as the Membership Council vice chair. (Consider volunteering.) He is an ASA Fellow. In these interesting times, he works toward long-term cultural change with the Alliance for Decision Education and Braver Angels.

MinJae Lee

MinJae Lee is a biostatistician and professor at McGovern Medical School, UT Health Houston, where she leads the division of clinical and translational sciences. Her research focuses on developing and applying innovative statistical methods to address real-world research challenges. She is a member of the ASA Committee on Applied Statisticians and serves on the Nature Medicine Statistical Advisory Panel.