- The Uncertainty Project
- Posts
- š® The role of conviction and narrative in decision making
š® The role of conviction and narrative in decision making
Conviction as a superpower, why we tend to be overly optimistic, and how outcome trees help us map potential outcomes
Howdy!
Weāve had a ton of new people join the community since our post last week, so welcome!! š ..in our weekly newsletter we dive in to a topic around decision making, cover one bias a week, and highlight a tool from the Uncertainty Project.
You can find all of our past posts here.
Last week we talked about goal setting and OKRs from the perspective of decision making.
This week:
š® Topic: The role of conviction and narrative in decision making
š§ Bias 10/50: Optimism Bias
š ļø Tool: Outcome Trees
The role of conviction and narrative in decision making
If youāve been following these posts, you may have noticed the word conviction comes up quite a bit. In the particular case of navigating radical uncertainty, itās the battleground between biases and heuristics (how we make irrational choices and act on flawed judgements) and naturalistic decision making (how these āflawsā enable us to navigate complex, real-world situations)
Before we get into strategic decision making and the role of conviction, lets revisit the āconfidence and intuitionā debate.
As we briefly covered in a previous thread, we see in this 2010 interview between Daniel Kahneman and Gary Klein that the two are at odds on the topic of confidence - specifically, overconfidence.
They both agree that overconfidence is detrimental, but Gary Klein, echoed by Gerd Gigerenzer and supporters of āEcological Rationalityā, makes exceptions for high-validity environments (e.g. medicine, coaching, aviation) where leaders have deep domain expertise. The most common example was highlighted in Malcolm Gladwellās book āBlinkā about a firefighter who intuitively made a snap decision to evacuate a building by ājust knowingā the floor might collapse.
Regardless of this disconnect on over-confidence and intuition, these topics have been covered by countless books authored by various researchers across economics and behavioral science - Richard Thaler, Cass Sunstein, Herbert Simon, Tali Sharot, and Richard Zeckhauser to name a few.
Through this research, we know a great deal about how individuals make judgements and choices - even in some limited real-world scenarios where information and time are limited (e.g. the firefighter example), but this still doesnāt effectively represent āradicallyā uncertain environments.
Other researchers, practitioners, and academics in the overlapping worlds of strategy and decision making warn against directly translating learnings from decision making observed in a lab to real world strategic decision making.
Much mischief can be wrought by transplanting this hypothesis-testing logic, which flourishes in controlled lab settings, into the hurly-burly of real-world settings.
What factors differentiate strategic decision making?
In his book Left Brain, Right Stuff, Phil Rosenzweig argues there are a few primary factors unique to strategic decision making - Influence, Competition, Feedback Loops, and Conviction:
Influence: How we act and intervene can influence the outcome (He argues we often have more control than we may realize)
Competition: In this environment, success is relative, not absolute
Feedback loops: Time to learning is often long and lagging
Conviction: Perception and credibility are important
He argues these are the primary factors that cannot be duplicated in a lab and are even difficult to observe in real-world situations - especially since the context for each decision if often unique.
Many of these factors feel intuitive - but what is conviction? How is it different from the overconfidence behavioral scientists warn of, and why is it important in this environment?
First, we can try to make a distinction between the environments:
Normative Environments: where there are clear actions and subsequent outcomes that are designated as good, desirable, or permissible, and others as bad, undesirable, or impermissible. We have a clear idea of what āshould beā vs what āshouldnāt beā.
Radically Uncertain Environments: where outcomes and utility are unknown, the relationship between actions and outcomes are not stable, and many potential options/scenarios canāt be effectively modeled or predicted.
We might say that these normative environments are high validity, but uncertainty can vary. For example, poker and warfare are high validity environments (where one right or wrong move influences the next and feedback loops are tight), but also highly uncertain environments - luck plays a significant role.
Most research observes situations on some spectrum of these ānormative environmentsā, but these donāt effectively model radical uncertainty - a low validity environment thatās also highly uncertain (business, politics, or something like space exploration)
We use the term radical uncertainty to refer to equivocal situations in which uncertainty about the outcomes of actions is so profound that it is both difficult to set up the problem structure to choose between alternatives and impossible to represent the future in terms of a knowable and exhaustive list of outcomes to which to attach probabilities
So how do we think about strategic decision making in these environments of radical uncertainty? Strategy, management, and leadership practices have been widely explored and debated, but as neuroeconomics and the study of decision making is still a nascent field, relatively little work has been done on the subject of strategic decision making in complex, uncertain environments.
āRadical uncertainty cannot be described in the probabilistic terms applicable to a game of chance. It is not just that we do not know what will happen. We often do not even know the kinds of things that might happen.ā
In the work that has been done, itās evident that the conviction plays a major role in our ability to operate in these environments, and inspire others to do the same.
Weāll dig in to a few of these studies that explore this phenomenon, but first, we just covered this foundational argument around confidence (and overconfidence), so how do we differentiate conviction?
Conviction vs Confidence
Because they are easily conflated, itās helpful to clarify confidence and conviction:
Confidence: The state of feeling certain about the truth of something.
Conviction: a firmly held belief of opinion
The primary difference is that confidence is relative to certainty while conviction is based on beliefs - regardless of evidence or probabilities. If weāve learned anything so far, itās two things:
The pursuit or illusion of certainty is paralyzing and dangerous
Strongly calcified beliefs and assumptions are the enemy of progress
āWhen I find new information I change my mind; What do you do?ā
In tech especially, we glorify overconfidence - āStrong opinions, loosely heldā (a popular quote by Jeff Bezos) exists in some fashion in many ācompany principlesā - but this way of thinking can have detrimental effects if itās giving a hall pass to overconfidence.
In short, because it assumes weāre good at challenging our beliefs, objectively processing information, and viewing reality for what it is.
But weāre not. By default, we anchor to our initial assumptions, intake and share information through a strongly biased filter, and construct our subjective reality in real time.
So we need to find a sweet spot and put guardrails in place that combat overconfidence, challenge assumptions/beliefs, and actively avoid the comfort of the status quo (or regression towards the mean)
As an interesting aside; Itās well studied that our memories are not accurate representations of the past, and some studies suggest that our memory recall may be more optimized for simulating the future than remembering the past. The functions that drive our ability to do this āmental time travelingā are rather similar.
The capacity to hold conviction in the face of uncertainty
Peter Drucker famously said, āWhenever you see a successful business, someone once made a courageous decision.ā
Thereās something very human about the ability to succeed against all odds. There is, of course, danger in these intoxicating and heroic stories - they often dismiss base rates and fall victim to survivorship bias, but nonetheless, conviction plays an instrumental part in the emergence of innovative ideas and āboldā decisions.
Sir, the possibility of successfully navigating an asteroid field is approximately 3,720 to 1
C-3PO
Even the perception of conviction is meaningful. We gravitate towards confidence not only because itās inspiring, but itās a herd heuristic that allows us to make fewer, harder decisions on our own. Itās a lighter cognitive load to trust than it is to construct our own convictions.
There are a few interesting studies that suggest a model for thinking about conviction and strategic decision making:
Image Theory: When we think strategically, particularly to achieve some goal, we create mental images: The self-image, the trajectory image, the action image, and the projected image.
Conviction Narrative Theory: Enables actors to draw on the information, beliefs, causal models, and rules of thumb situated in their social context to identify opportunities worth acting on, to simulate the future outcome of the actions by means of which they plan to achieve those opportunities, and to feel sufficiently convinced about the anticipated outcomes to act.
Image theory has been influential in theorizing how mental images might play a major role in how we model strategic decisions - essentially creating a picture of our beliefs and values, what success looks like, our action plan, and potential ways the plan unfolds.
This is counter to the idea that we analytically reason our way to conclusions and instead, compare new information to these mental images. Below we can see how Lee Roy Beach represents two types of decision candidates (progress or adoption) challenging the mental images and deciding whether or not the model accepts and adopts new updates.
If it would be valuable to go deeper on image theory, let us know! (comment/review).
Now layering in Conviction Narrative Theory, we start to see how these mental models build narratives that we then act on (and continue to act on) among a constellation of changing variables and new information.
Narrative: A process that allows us to construct the everyday meaning of events and happenings along with their causal implications.
Conviction: The subjective confidence to act (and often to carry collaborators with you)
In Conviction Narrative Theory, a preferred narrative for action emerges through conscious deliberation influenced by an overall appraisal of the approach and avoidance feelings evoked by every chunk of the narrative predictions about outcome that form the narrative. So, the credibility of the information contained in the narrative for action, the implicitly and explicitly held beliefs that govern expectations, the decision rules guiding inference, search, and checking that are adopted, the explanatory causal models used in simulating the outcomes of action, the various action tools selected, etc., all evoke emotions.
A preferred and conscious narrative of the outcome of action emerges from the evaluative process, in part explicit, but much of it implicit and automatic, to evoke a state of action readiness.
Itās interesting that both of these theories build on research suggesting itās incredibly difficult to update the underlying beliefs and assumptions that subconsciously drive our decision making.
We, as individuals, vary on our threshold for willingness to hold convictions in the face of imperfect or contradicting information, as well as our tendencies to seek certainty (or fabricate the feeling of certainty). Itās common to point out individuals that are more āconfidentā even if weāre conflating the term.
But instead of dismissing conviction, that by definition is often contrarian or counter to existing information, we may think of it as a unique and powerful factor when making decisions in radical uncertainty.
And also deeply humanā¦
The ability of human actors to draw on feelings of conviction provides an advantage unavailable to a computer generating only scenariosā¦
āā¦While a computationally competent outside observer may be unable to identify secure grounds to support a particular narrative of the future in radical uncertainty and so to commit to a particular decision, a human decision-maker can feel sufficient conviction to act.ā ā David Tuckett, Milena Nikolic: The role of conviction and narrative in decision-making under radical uncertainty
Beware of escalation of commitment
Weāve started to explore the downsides to holding onto convictions too tightly for too long and crossing in to overconfidence and hubris - Annie Duke, in her book āQuitā extensively covers the risks of escalation of commitment and holding on to sunk costs too long.
Thereās plenty of nuance in this topic and a lot more to dig into here that we can cover in the future!
Let us know what you think! You can join the conversation on the thread in The Uncertainty Project. Want to build on or challenge these arguments? Submit a thread!
Bias 10/50: Optimism Bias
Optimism bias is the tendency for individuals and groups to overestimate the likelihood of positive outcomes and underestimate the likelihood of negative ones. This bias can manifest in many ways, from underestimating the time and resources required to complete a project to overestimating the potential success of a new initiative.
In group decision making, optimism bias can lead to a phenomenon known as "groupthink," where members of a group become so convinced of the correctness of their initial assumptions that they fail to consider alternative perspectives or options.
This can, of course, be particularly dangerous when the group is making high-stakes decisions, as it can lead to poor outcomes or even catastrophic failure.
One reason for the prevalence of optimism bias in group decision making is that people tend to want to be seen as positive and forward-thinking, and they may be reluctant to voice dissenting opinions. Additionally, group dynamics can create pressure to conform to the opinions of the majority, further exacerbating the bias.
To combat optimism bias in group decision making, it is important to actively encourage dissenting opinions and to create a culture where it is safe to speak up.
Additionally, tools such as a pre-mortem, nominal group collaboration, and scenario planning can be used to help groups consider multiple perspectives and potential outcomes.
Read more:
Tool: Outcome Tree
An outcome tree is an exercise for modeling potential outcomes and probabilities of those potential outcomes. Outcome trees help do three things:
Align on value by being clear on the desired outcome of the decision
Defining the potential outcomes helps measure risk vs expected value
Build rigor in documenting your expected outcomes to compare with the actual outcome in order to make better decisions in the future
Whenever you make a key decision or take a key action, write down what you expect will happen. Nine or 12 months later, compare the actual results with your expectationsā¦ Practiced consistently, this simple method will show you within a fairly short period of time, maybe two or three years, where your strengths lieāand this is the most important thing to know.
How to build an Outcome Tree
Building an outcome tree is quite simple. The goal is to understand the possible outcomes of the decision in order to have a conversation about confidence and risk.
If you want to take it a step further, the outcome tree can help you understand how to measure success, and even more importantly, put measures in place that could help provide leading indicators of a potential negative outcome.
1) Identify the potential outcomes
The best place to start is by simply listing out potential outcomes. This can be done as a stream of consciousness or prompted by questions. For example, asking "What area all the potential positive outcomes?" and conversely, "What are the potential negative outcomes?".
What often happens is that the outcomes you want or expect tend to come to mind first. As this list starts to move into negative outcomes, interesting conversations start to emerge. If we decide to build a new feature, it's a very real possibility (no matter how much we think it won't be the case) that no one will use it. If that happens, then what? How might we defend against that outcome?
Building out the visual tree is helpful to see the different tracks potential outcomes can take. Maybe it's simple, but maybe there are domino effects as well.
Adapted from Annie Dukeās book, How to Decide
Keep in mind the goal here is to drive conversations that lead to better risk mitigation and eventually better decision making - this is not meant to be an exhaustive list.
2) Understand the probability of the potential outcomes
After we build out the list, the next step is to group our potential outcomes into ones we believe are positive, negative or neutral. We could take this a step further and rate them on a five star scale to add some depth. For example, we may have two negative outcomes, but one may be a far worse outcome than the other.
For each potential outcome, give it a 'probability' for the likelihood this outcome will occur. Again, this activity is to drive conversation, so roughly right is okay. We can simply say 'high', 'low' or even 'very low'. If we're talking about a critical decision, maybe we do take the time to give percentages to these probabilities based on some research or insight.
Nevertheless, with each of these, have a conversation around why we believe the probability is what it is. Do we believe the positive outcomes are high due to bias? Do we have information that supports our assumptions? If not, think about whether or not all the risks are really known. This may influence your confidence score.
Adapted from Annie Dukeās book, How to Decide
3) Measure, learn and adapt
Over time, decisions need to be revisited to log the actual outcome and retrospect on the process and expectations against reality. If a negative outcome occurred, did we predict it as a possibility? Were our assumptions of the risk or probability wrong?
Much of the value of outcome trees come from the activity of logging the process before the actual outcome happens. Once the actual outcome happens, we're subject to hindsight bias. Looking back, we (and others) believe the outcome was obvious all along. Documenting that process helps to adapt the process.
āWhat makes a decision great is not that it has a great outcome. A great decision is the result of a good process..."
In the inverse, a bad outcome doesn't necessarily mean a bad decision was made; plenty of factors are involved, including luck. What determines a good or bad decision is based on the process built around that decision to accept or mitigate risk. In the tree above we can see that not a single outcome is certain, we're looking to accept risk relative to the expected value.
Logging the process, even if simple at first, and learning from the outcomes is the single most powerful way to make consistently better decisions as a team or organization.ā
Resources:
We hope this post was helpful and interesting! Have feedback? Just reply to this email! It would be great to get in touch!
How was this week's post?We'd love to know what you think! (click one) |
Reply