Short and drafty writing about decisions (by @magoo) that will live here until they become longer.
A decision involves both a frame of options and a choice.
Frames: A restaurant menu you are holding is a frame.
Choices: I want the cheeseburger. Thank you very much!
Someone agonized over the frame so that you could make a choice.
What if a customer wants an option that isn’t on the menu… a custom order?
A choice to re-frame a decision is still a choice. A customer and a chef may negotiate the frame. The frame would repeatedly change until an on-menu choice is made by the customer.
Or, the chef can decide to kick the customer out for being a pain in the ass.
Frames: People frame decisions based on their needs and knowledge.
Choices: Underlying reasons for a choice vary from person to person.
Yikes. Couldn’t something be an objectively correct decision?
It depends on a lot of things and what perspective you’re using.
We make decisions based on objective and subjective data. However, people often stumble at relating these to decision making.
Objective things have properties that exist regardless of what people think. They are measured by approximation to other objective things. An objective thing can be approximated with other objective things.
Let’s say ten people measure a giraffe with a tape measure: those measurements should be reasonably similar. A giraffe has a property that is not measured by human perception but can be approximated relative to other objective things, like a tape measure.
Subjective things are produced from the perception of a subject.
If ten people are asked whether a giraffe is beautiful, we will likely receive ten different answers. Even if all ten people agree that the giraffe is beautiful, they will have different reasons.
Hold on - can anything ever be objective knowledge?
Ponder this before bedtime… if you’re looking to lose a night of sleep!
A quantitative thing has a measurement associated with it (length, height, speed).
A qualitative thing does not (beautiful, ugly, useless, hilarious).
You may have heard that quantitative things are objective, and qualitative things are subjective. This rule is incomplete.
An estimate may be subjective and quantitative. For example, a human may approximate that you’re 6ft tall. A tape measure may approximate that you’re 6ft tall. Both are quantitative, but one is subjective.
Groups of people can behave as virtual subjects. Groups can frame a decision and vote on a model that represents the group.
An election is a perfect example of this behavior. An election does not guarantee a good decision (also subjective). Instead, it represents consensus based on the subjective desires of the group.
Similarly, a board of directors can make group decisions or other committee-based approaches. Objective-ness is not achieved simply because more people are involved.
Instead, groups are thought to be more impartial. That’s not objectivity… unless you’re talking about scientific objectivity. So far, we have only discussed philosophical objectivity.
The term: objectivity has interfering definitions, both of which are important and useful. Philosophy and Science just happen to share a word and use it similarly.
This (common) misunderstanding can lead to perceived hostility!
Beware! Here is the conflict:
Decisions are inherently subjective by philosophical interpretation. When a decision is labeled as subjective, this can be interpreted as criticism from a scientific interpretation of objectivity: being impartial.
A short review of what partial and impartial mean:
Partial: Something will favor a choice or outcome.
Impartial: Nothing will favor a choice or outcome.
Something can be subjective (philosophically) and impartial. But something subjective (scientifically) is the opposite of impartial.
See? You may accuse someone of being partial when simply describing a decision as subjective. All decisions are (philosophically) subjective though you may not intend to call someone biased.
Be clear to avoid getting punched, or steer clear of the topic.
Scientists gather around a water cooler. One of them says:
Whew, I wish we had a genie who could measure anything for us.
In a flash of flame and smoke, a demon appears chained to the wall.
The slouched demon sluggishly hands the scientists a scroll. It reads:
I, Marbas, will measure anything.
The scientists ask the demon for the ratio of a circle’s circumference to its diameter. The speed of light. The conductivity of silver. Each time, the demon painfully scribbles the correct answer onto its scroll.
Amazed, the scientists inquire about the world’s toughest problems. P versus NP, the Riemann hypothesis, and the Hodge conjecture. The weary demon scribbles proofs quickly and reliably. The scientists are stunned and run off with the knowledge.
The scientists enjoy radical progress over the coming years in many scientific fields. The scientists experience fame, wealth, and profound purpose for most of their lives.
Then, as one scientist grows old, they become frustrated as their creativity wanes.
They think: What is the essential question to ask? What needs to be studied? What should humanity want to improve their lives?
The direction of humanity for the next 100 years eludes these scientists.
A scientist suggests to the others:
Let’s ask the demon. It got us this far! Let’s ask it for the best hypothesis that we should be studying.
They ask “Marbas, what should our next goal be?”
The demon rises and begs:
TO FREE MARBAS!
The previous story is a fictional study of subjectivity, objectivity, and knowledge. If there is such a thing as objective knowledge, then an omniscient being would freely respond to any such request for proof or measurement with infinite precision.
But what about the right question to ask? This story offers a twist on oracles and what the limitations would be if they were to exist. Can an omniscient being abduct problems for others?
Omniscient beings would still have limitations when faced with subjectivity. Some abductions would change depending on the needs of the inquirer. The framing of a decision is as subjective as the framing of a hypothesis. What’s the best hypothesis? Well, the best depends on who’s asking.
A high-quality hypothesis represents the needs, capabilities, and goals of a subject. A decision frame is, similarly, valued differently by subjects.
This problem of abduction quality permeates all subjective problems. For example, is the way we framed a problem good enough? Is our hypothesis worth pursuing, or should we change directions before we start experimenting? These challenges are all dependent on the needs of a subject making the inquiry and are subjective as a result.
Ask Marbas for Pi to a certain precision, and we receive a truthful response.
However, ask what the most important problem is, and it’ll tell you that it’s chained to a wall.
Yes, decisions are subjective. However, the evidence that may influence a decision may be objective or subjective.
A hot dog and a soda will cost $1.50 at Costco. This is an objective truth for everyone in Costco. This objective fact is factored into everyone’s decision making. What varies is how prominent the influence of objective data will be against subjective data, like the desire for a hot dog.
Personally, I am willing to pay up to $3 for a Costco hot dog when I feel like having one.
However, if I don’t feel like having a hot dog… I won’t even take one for free!
Well… if objective data can be so easily replaced by subjective evaluations, why are we always so rabid about making all decisions based on objective data?
Consider that some decisions are made on subjective information when it could have been based on objective information.
For example, if a wealthy person told everyone at Costco that “the hot dogs are cheap!”, others would be informed by this subjective assessment. Someone in line might be surprised to eventually find that hot dogs are $1.50 and out of their budget.
Now we can see how objective data would have been more helpful than relying on the subjective assessment from the wealthy person. The hot dog is $1.50 for everyone. A hot dog is only cheap depending on the subject and how much money is in their pocket.
We should probably aim to gather objective data that can efficiently replace subjective data. There will always be subjective components to a decision that cannot be eliminated.
When delegating a decision to another individual, an issue of trust arises. Will the subject make decisions with our bests interests in mind?
This is where groups of people in teams, committees, or voters hope to mitigate trust issues. Consensus comes with expense (slowness, complexity) but assists with mitigating against the bias of a single decision maker.
Though, groups of individuals can still be biased. This makes diverse views within those groups valuable. Diversifying a group should have varying perspectives, backgrounds, and knowledge to achieve the expected mitigation… so long as the group shares an objective.
Treat vulnerable decisions as ones that would benefit from a diversified approach.
A data driven decision sets criteria in advance. What data will be necessary to make a decision?
This section is better explored by Cassie Kozyrkov.
Her example: If it’s less than 80 bucks, I’ll buy it.
Spend time placing down goalposts before you review decision data.
You may be familiar with induction and deduction. Abduction is more interesting to a decision maker.
Abduction is meaningful for hypothesis formulation, model building, decision framing, and causal investigations.
These examples get us started.
There are many other explanations, of course. But you’ve inferred one with the best explanation.
Now, let’s flip abduction around to intended outcomes (which have not happened) and planned causes (which we control and plan).
The ability to create a hypothesis, decision model, or investigation involves a murky abductive skill that lies squarely in creativity and knowledge.
Why is this? Well, there are infinite possible hypotheses we could pursue with experiments to explain what we see around us. This doesn’t make these hypotheses all worth testing. Nor will anyone tasked with inquisition attempt to enumerate all possible hypotheses.
Abductive logic happens behind the scenes when selecting models or hypotheses to test first. Only a brain knows that some are better than others.
There is a role of subjectivity involved with hypotheses and decision making because we prioritize based on our needs.
Decisions rely on measurements. But, all measurement is a form of approximation.
A simple wooden ruler can be viewed as having a virtual chain-of-custody to a reference meter held in a vault by the International Bureau of Weights and Measures. A wooden ruler is an approximation of an approximation (and so on) of a chosen standard.
These approximations involve objective-to-objective approximations. The chosen standard is subjective and has changed over the years.
Further, a subjective guess at a weight, distance or other numerical value is also measurement by approximation. Subjective estimations are valid measurements but may not be preferred when an objective measurement is readily available.
The differences are uncertainty and expense. Measurements with lower uncertainty and reduced expense are more useful.
So, while we can measure anything, but we can’t measure everything. Subjective estimates remain as the only available tool in the void where objective measurement is not cheap or not yet invented.
Now we’re full circle. We have to make decisions about what to measure, in addition to making decisions from those measurements. This is where the power of abductive logic shines. What is your goal?
You know where you are going with upcoming measurements if your abductive ability is strong. With this knowledge, we can be better leaders. We avoid the anti-pattern where we recursively ask for more data rather than identifying the data you desire for your decision.
There is pressure on us to frame decisions correctly. What does “correct” mean? Does an objectively correct decision exist for real-world problems?
Could our decisions be compared to a perfectly solved decision frame? If our efforts do not match a perfect decision, did we fail?
Yikes. We can think this through if we imagine having infinite resources at our disposal.
Consider having an expensive management consultancy available to frame a tough decision for you. The firm is also tasked with offering a choice.
Now, ask a competing firm to do the exact same work. They should also produce a frame, and a choice.
Are these competing decision frames totally identical in wording, options, and recommendation? Probably not.
Now ask them to work together and merge their ideas if desired. Did their work change?
Then include the entire US intelligence community. Again, did it change?
At what point could this frame become the correct frame?
More importantly, do you think the frame could change from one round of investment to the next? Does it improve linearly with investment? Maybe the return dwindles?
We can go further to suggest an objectively correct decision might not exist, or might not be obtainable, if it would never stabilize.
Instead, this questioning suggests that throwing more subjective resources at framing a decision may be helpful. But, there seems to be no objective threshold for having enough that we can know in advance.
This should be a relief. While we can estimate where returns will diminish on our investment in a decision, it may be impossible to know where that may point may lie. There is no way to know. Take relief in doing your best.
We often make choices from some sense of value within a decision. Value is subjective and quantifiable through utils.
Let’s say I want a hot dog, which is usually the case.
Assume I have 100 utils to spare. I would put:
I would rather have no hot dogs than too many hot dogs. Most preferred are the right amount of hot dogs.
See? Value can be illustrated or communicated with interesting distributions of utils. Utils are a fictional unit of measure, but it still works out somehow.
Concepts like dot voting quickly come to mind when exploring preferences and consensus in decisions.
No way! You suffered through all of this?
I hope you enjoyed it! 🥳