Prioritization needs no ambiguity

When we are truly convinced that the next task on our to-do list will significantly advance our goals, we naturally become more enthusiastic about tackling it. Some parts of you believe this to be the case, and other parts very obviously don't. Thus, there is uncertainty about how to prioritize. I think various systems for prioritization get much of their power from addressing some core ambiguity. People thus self-sort into those systems based on whether that ambiguity was a critical bottleneck for them. This post isn't about outlining another set of antidotes but merely mapping out some (not necessarily all) various kinds of ambiguity.

Short-term goals offer clear and immediate feedback, along with satisfying rewards. On the other hand, long-term goals require sustained effort but have the potential to yield higher effort-to-reward ratios if pursued successfully. While a simple first-in-first-out scheme might suffice if our workload were manageable, numerous opportunities constantly arise, leading to a continuous need for prioritization. More opportunities are presented as we refine our prioritization skills, resulting in an ever-growing workload. It's evident that prioritization is an ongoing necessity.

Ambiguity as a matrix: risk over uncertainty

When I use ambiguity in this post, I'll refer to risk and uncertainty as potential roadblocks. Consider planning a picnic. You want everything to go smoothly, so think about what could go wrong and how to prepare.

Risk

You check the weather forecast and see a 30% chance of rain on the day of your event. This is a risk. You can plan for this by bringing umbrellas or setting up tents. You know there's a certain probability of rain, and you can prepare for it because it's a known factor. Why it's Risk: The rain is a potential problem you can predict based on the weather forecasts. You can calculate the chance (30%) and plan to minimize its impact. You might think of them as known-unknowns.

Uncertainty

Although you planned, suddenly, an unexpected storm hit without warning, even though the weather forecast said it would be sunny. Or maybe a nearby construction site unexpectedly decided to test loud machinery that day, disrupting your event. These are uncertainties. Why it's Uncertainty: You didn't predict or plan for these events because they were outside the information you had or thought to consider. You couldn't prepare for them because they were not part of your original planning model. You might think of them as unknown-unknowns.

One task after the other

We are already prioritizing in several ways, and confusion about which would be best in a given situation can be a roadblock. Let's see some common models:

• In First-Due prioritization, we do whatever has the nearest deadline.
• In Longest-Chain prioritization, we prioritize whatever task will take the most time or have the most significant number of sub-tasks to complete.
• In Shortest-Chain prioritization, we want to reduce the total list size of tasks as much as possible to complete all the shortest tasks quickly.
• In Most-Salient prioritization, we allow tasks' vividness and emotional immediacy to serve as the goal.
• In Most-Likely-Failure prioritization, we look for tasks with a highly uncertain step and see if we can test that step. If it fails, we can discard the whole task and thus increase total throughput.
• In Most-Reusable prioritization, we focus on those tasks whose partial or complete solutions will be most helpful in completing multiple other tasks. This also might be thought of as a sub-type of Highest-Information-Gain.
• In Expected-Value prioritization, we focus on those tasks that will result in potentially the biggest payoffs, presumably creating resources for engaging with other tasks. While this seems like a good approach, we soon realize that we've shifted the problem up a level, as now we have to manage the fact that there are varying types of rewards and our marginal utility for a specific resource, as well as the ability to convert between different types of value, may change over time.
• Due to the well-known effects of loss aversion, it's also worth specifically naming a commonly encountered subtype: Expected-Loss prioritization. Catastrophization is a subtype focusing on the chance of being wiped out (often overemphasized because of the Most-Salient consideration).

Delay

Many people default to a strategy of Delay, and it is worth pointing out that considering this just procrastination makes it hard to understand the real benefits and costs (of this strategy). Delaying helps people deal with complex prioritization problems. When there are many tasks or decisions to make, each with its priorities and consequences, it can be overwhelming to decide what to tackle first. Procrastination simplifies this complexity by allowing immediate concerns to rise to the top.

Instead of analyzing all dependencies and carefully choosing what to prioritize, people often use simple rules of thumb (heuristics) to decide. One such heuristic is: "Who will be angry with me soonest if I don't do X?" This approach combines the idea of what's due first ("First-Due") and what seems most important or pressing at the moment ("Most-Salient"). This is usually a common strategy used by people tackling akrasia.

One task now

Ambiguity about individual tasks serves as an additional roadblock. One easy way of combating this ambiguity is by asking all the context questions: who, what, where, why, and how. To this, I might add a couple of less-known ones that add additional specificity:

Which 

It's used as a drill-down step if the answers to any of our other questions are too general to be of use.

Who does this affect?
The picnic guests
Which?

This suggests that we usually have some tacit intuitive sense of the appropriate scope of a given task or sub-task and that this scope may or may not be well-calibrated.

Source

 (From where did I get this task?) As a backward-facing question, this accounts for ambiguity around a task's origin. Did we make our jobs harder when we stripped the task of that context when recording it?

After some tasks, we get to goals, right?

Techniques such as goal factoring aim to simplify prioritization by prompting an examination of how sub-tasks contribute to overarching values and goals. Some people reading this might jump to OKRs or other tools to outline the same principle. Here are three examples:

• Task-Outcome Ambiguity:
  • Arises when we are unsure about the actual effects or outcomes of completing a specific task. There is uncertainty about whether the task will lead to the desired results.
  • Example: If you're assigned to develop a new marketing strategy, task-outcome ambiguity would be the uncertainty over whether the plan will increase sales or improve brand recognition. You don't fully know the impact the task will have.
• Instrumental-Goal Ambiguity:
  • This ambiguity involves uncertainty about how well our chosen methods (such as goals or proxy measures) align with our future preferences or values. It's about the connection between the means (what we're doing now) and the ends (what we want in the future).
  • Example: If a company sets a goal to increase customer engagement by using social media likes as a proxy measure, there is ambiguity about whether this will truly reflect long-term customer loyalty or satisfaction. This relates to Goodhart's Law, which states that when a measure becomes a target, it can lose its value as a good measure.
• Part-Whole Relation Ambiguity:
  • It deals with uncertainty about how individual actions satisfy longer-term preferences or goals. It's about understanding which steps are necessary and sufficient to achieve the broader goal.
  • Example: If someone's long-term goal is to be healthy, individual actions like dieting, exercising, or meditating contribute to this goal. Part-whole relation ambiguity would involve uncertainty about which actions are essential and how much each contributes to the overall health goal.

Food for later thoughts

There are a few things here, in a random order, which I might tackle in future posts.

• What exactly are we accomplishing when we confront uncertainty in prioritization? One possible explanation is that we are consistently transforming a partially ordered set of tasks into a more ordered set of tasks, up to the point where we have enough order for our 'good enough' heuristics to avoid any major losses. And as with everything related to bringing order, entropy is always present. There are likely other explanations that shed light on different aspects of the issue.
• Ambiguity about the correct level of abstraction to explore/exploit. When trying to do our taxes, instead of getting anything done, we might write an essay about prioritization and ambiguity 🙃.
• Risk aversion is different than uncertainty aversion. It feels like there's potentially a lot to unpack there.
• Motivational systems, like logic, emotions, personal needs, and ethics, help us manage the overwhelming number of choices we face by narrowing down the options to make decision-making easier.
• We are attacking others' ambiguity aversion directly as an emotional intervention. What are we afraid of when we avoid ambiguity, and what is the positive thing that part is trying to get for us? There is likely much more here than just 'cognition is expensive,' and this post itself could be seen as generating the space to forgive oneself for having failed in this way because the problem was much more complex than we might have given it credit for.
• Ambiguity is a liquid that backs up into whatever system we install to manage it. Sure, you could deploy technique X that you learned to prioritize better (GTD, KonMarie, Eisenhower Matrices), but that would favor the tasks you deploy them on over others. There's ambiguity on whether that's a good idea. Related to ambiguity about the correct level to explore exploit on as well as Aether variables, bikeshedding, and wastebasket taxons. i.e., Moving uncertainty around to hide it from ourselves when we don't know how to deal with it.

Take me away

• Make sure to ask the "Which" and "Source" questions. They might give the extra context needed to know what to do with the task.
• Write tasks horizontally along your daily planner. Use the space beneath them to add context and details. One of the features of this format is that it lets you write 2-3 tasks on a page; don't treat it as a bug.
• Don't be afraid to pick different styles of prioritization based on the time and place. These are tools, use them to the best of your abilities.