At a glance: Decision quality is not about being smarter. It is about having predictive models that work, the cognitive conditions to apply them, and the discipline to separate what you can control from what you cannot.
The usual advice about making better decisions is to know your biases and try to compensate for them.
That advice is fine as far as it goes, and it misses what experienced professionals have actually built.
Years of practice with real feedback have given you working models that generate good calls before you can explain them.
The work now is to maintain those models, protect the conditions under which you use them, and learn to tell a sound decision from a lucky result.
Self-direction is what makes all three possible, because you control when and how your most important thinking happens.
Decision quality is not about eliminating uncertainty. It is about applying models that work, under conditions that support clear thinking, with the discipline to separate your reasoning from your results.
The advice that is true but incomplete
The conventional story about decision-making is about avoiding bias.:
- Think more clearly
- Notice your cognitive traps
- Resist wishful thinking
None are wrong, and for experienced professionals it leaves out the most important thing you own.
Years of practice have built something more useful than a list of biases to watch for. You will have built working models of how things actually work: what a client problem signals before the client has finished describing it, which projects slip, which apparent emergencies resolve on their own.
Keith Stanovich, the cognitive scientist whose work on rational thinking sits behind much of this, calls the rules and strategies you can pull from memory to make decisions your mindware.
The useful kind, refined by real feedback on real consequences, he calls productive mindware. The question for a wisepreneur is not how to think more clearly in the abstract. It is how to look after the decision-making you have already built, and how to recognise the conditions under which it works best.
I want to suggest that the people best placed to do this are independent professionals who direct their own work. Not because they are smarter, but because they control the things that decide whether good judgment gets a chance to operate.
Why experience and expertise are not the same thing
Michael Mauboussin, drawing on the psychologist Greg Northcraft, makes the distinction plain.
- An expert is someone who has a predictive model that works.
- An experienced person is someone who has been around a long time.
On the stable, predictable side of a problem these two tend to line up. On the unstable, uncertain side, where most interesting professional work lives, they come apart.
Plenty of people accumulate years without building models that reliably predict anything.
They develop what Stanovich calls contaminated mindware: assumptions, habits, and industry orthodoxies that feel like expertise but were never tested against outcomes. They have confidence without calibration. Their decisions feel informed while running on a flattering sample of remembered wins.
Cedric Chin, who writes the business-expertise publication Commoncog and was a guest on the Wisepreneurs Podcast, went looking for what separates the two. He spent years studying how people get good at business quickly, and found his answer in a field the United States military funded in the late 1980s called Naturalistic Decision Making.
Its researchers developed a way to extract the tacit models that sit in the heads of genuine experts.
Tacit means it cannot easily be put into words. As Cedric describes it, you ask an expert how they knew what to do, and they say it just felt right, and that is the most they can give you. That feeling is the output of models built through hundreds of encounters with real consequences.
This is the experience I mean when I say your brain generates an assessment before you can articulate why. Over many years it has been storing patterns from every problem you have worked on, and a new situation calls them up. The assessment is not a guess. It is prediction from models that have been tested.
Cedric also found the clearest evidence that this is real and not a flattering story we tell ourselves. He came across the work of Lia DiBello, a researcher who built an assessment of business expertise that companies brought her in to run on their executive teams.
What made it a good measure, in his words, was that it tested your predictive ability, your ability to predict what was going to happen in the business. That is Mauboussin's point arrived at from a completely different direction. The mark of expertise is a model that predicts, not a long record of attendance.
The distinction matters commercially. An AI system can process more data, faster, than you can. It has no models grounded in years of embodied professional experience. A junior colleague has energy and current technical skill, and lacks the accumulated encounters that train real predictive capability. The models that actually work are what a client pays for when they hire your discernment rather than your procedures.
Why many models beat one
Scott Page, who studies how groups and individuals predict and decide, argues that the best decision-makers do not lean on a single framework. They use ensembles of models, several lenses on the same problem, each catching something the others miss. He calls the people who do this many-model thinkers.
This is what extended practice quietly builds. A consultant of twenty years does not meet a client problem with one model. They bring a way of reading the commercial structure, a feel for the human dynamics, knowledge specific to the industry, and the stored sense of what worked in situations that rhymed with this one. Each model is partial on its own. Together they generate a richer read than any single one could.
Cedric's account of DiBello's research sharpens this. She found that expert business operators carry a shared model with three parts, what she calls the business triad: the market, operations, and capital.
Strong operators are decent across all three and know to bring in others where they are weak.
Cedric describes this as distributed expertise, the recognition that a capable team is made of people who are better and worse at different parts of the same model and shore up each other's gaps. The same is true inside one experienced head. Your judgment is an internal ensemble, several partial models weighted against each other.
Page's idea has a direct bearing on how you should think about AI. It can run any single model faster than you can. The decision about which models apply to this situation, how to weigh them when they disagree, and when to override one because the context holds something the data does not, comes from the many-model arrangement you have built through practice. The integration across models is the part that does not delegate.
AI runs models. Your accumulated judgment decides which models matter for this situation, and how to weigh them when they conflict. That integration is what years of practice build, and it is the part that does not automate.
The discipline of deciding against something
Having good models is not the whole of deciding well. You also need the discipline to use them on the right thing.
This is where a conversation I had with Peter Compo, author of The Emergent Approach to Strategy, changed how I think.
Peter argues that most strategies fail because they are really just long lists. You take a goal, break it into a hundred sub-goals, and call the list a strategy. His alternative starts with a different question: what is actually in the way? He calls it the bottleneck, the one constraint that, until you deal with it, makes everything else you do beside the point.
A strategy, in his framing, is the single central rule you follow to clear that bottleneck. His line has stayed useful to me as a test: when everything matters, nothing matters.
That is a decision-quality principle as much as a strategy one. Contaminated mindware treats every factor as equally urgent, because untested assumptions come with no sense of weight. Productive mindware does the opposite. It tells you which two things in front of you carry the decision and which eight are noise.
Peter's framework gives that instinct a structure you can actually run: name the aspiration, find the bottleneck, decide against a rule rather than against a wish. I had done many things across twenty-five years of running a marketing practice without ever being that disciplined about it, and reading his book named something I had been doing only loosely.
The connection to your own models is direct. The reason an experienced professional can pick the two things that matter is that their models have already filtered the situation. The discipline of deciding against a clear constraint, rather than reacting to whatever is loudest, is how you keep those models in charge of the decision.
Telling a good decision from a good result
Annie Duke spent two decades as a professional poker player before she became a decision strategist, and she names one of the most corrosive habits in professional life: resulting. Resulting is judging the quality of a decision by the quality of its outcome. Good result, must have been a good decision. Bad result, must have been a bad one.
It is the wrong test, and the error compounds across a career. Duke puts it simply. The quality of our lives is the sum of decision quality and luck. In poker, as in most professional work, hidden information is everywhere and luck plays a real part.
A sound decision under uncertainty can produce a poor result because of things you could not have known. A weak decision can produce a good result through fortune. If you grade your past decisions only by how they turned out, you will sometimes abandon strategies that were genuinely sound and keep ones that were actually weak. Your models get corrupted by noise rather than refined by signal.
Tonianne DeMaria, who co-wrote Personal Kanban and works on the psychology of how people manage their work, named the same trap from a different angle when she came on the podcast. We have a tendency, she said, to judge other people by the outcome of their actions, while we judge ourselves by our intentions. And that is not fair. It is resulting pointed outward. We extend ourselves the grace of understanding our reasoning, and we deny it to everyone else by reading only their results. The discipline Duke teaches is to extend that same grace, and that same scrutiny, to your own past decisions.
The practice is to evaluate the decision separately from the outcome. The question is not did it work out. The question is, given what I knew at the time, did I apply my best available models and reason it through. When the answer is yes and the result was poor, the correct response is not to change the strategy but to accept that uncertainty is irreducible.
When the answer is no and the result was good, the correct response is not a victory lap but an honest look at where the thinking was thin. This is Stanovich's reflective mind doing its job, overriding the easy reflex that equates results with reasoning. It is a skill that deepens with deliberate practice and rusts without it.
Grade the decision by your reasoning, not by how it turned out. A sound decision can land in the bad-luck quadrant, and a weak one can land in the lucky quadrant.
The cognitive conditions that judgment depends on
Gloria Mark studies the attention of knowledge workers, and her research points to something most professionals feel without managing. Making decisions uses up mental resources. Evaluate information, weigh alternatives, do it again and again through the day, and the executive function that powers careful reasoning starts to deplete. We get more impulsive. Our decisions get less consistent. The quality of judgment drops, not because expertise has gone anywhere, but because the biology that carries it is tired. Mark's image is exact: a hard decision late in a depleted day is like climbing uphill when you are already worn out.
For an employee, most of this is out of their hands. Meeting schedules, email cadence, and the demands of the organisation decide when and how often they have to choose. Someone else sets the conditions under which their judgment operates.
Here is where self-direction stops being a lifestyle preference and becomes a structural advantage. When you control your own calendar, you control the conditions under which your most important decisions happen. You can put the work that needs your deepest judgment where your resources are freshest. You can protect the break that restores executive function before a critical call. You can batch the small, routine decisions so they do not bleed the capacity you need for the few that matter.
Tonianne DeMaria has built her working life around exactly this, and she described it in practical terms. She does not want people to manage their time, she said. She wants them to manage their energy. She knows she is freshest in the morning, so that is when she does the most cognitively demanding work. And she is precise about why prioritising itself is so costly: deciding what matters most is one of the most cognitively demanding tasks there is, because it asks you to picture a future state and weigh risk you cannot yet see.
The highest-value decision is also the most tiring one to make, which is exactly why it should not be made on an empty tank. She is blunt about the biology too. The prefrontal cortex is where executive function lives, and once a person is overwhelmed or afraid, that is the part that gets compromised first.
The same expertise produces different qualities depending on the conditions you decide on. Self-direction lets you make the decisions that matter where the conditions are best.
None of this is productivity advice in the usual sense. It is the recognition that professional judgment has biological preconditions, and that a self-directed professional can build those preconditions on purpose. The most valuable thinking you do deserves to happen when you are most able to do it.
The wider case for staying the author of your own thinking
Laetitia Vitaud writes about the future of work, and on her third visit to the podcast, episode 71, she made an argument that runs underneath everything here. Older professionals, she thinks, will hold the upper hand with AI, precisely because they were trained to think critically and to analyse before these tools arrived.
They can read the output of an AI system with a sceptical mind, because they have the models to judge it against.
As I put it to her in that conversation, you sort of know when something is not right, your gut tells you before you can say why. That feeling is contaminated mindware getting caught by productive mindware. It is the audit happening automatically, and it depends on having done the hard thinking yourself for years first.
Laetitia is honest about the risk on the other side. She described what she calls the WALL-E effect, after the film in which people who let machines do everything slowly lose the ability to walk or think for themselves. She uses AI as a capable assistant and notices her own writing muscle weakening when she leans on it too much, and she worries her thinking could follow. The worry is the right one, and it is the strongest possible argument for the maintenance habit this whole piece is about. Models that are not used drift from reality. Judgment that is handed off does not stay sharp.
She also offered a reframe I keep returning to in my own practice. The modern working week, eight hours a day, five days a year-round, came from the factory, from the idea that output should be constant whatever the season or the state of the worker. Most of us no longer work that way, and creative and relational work does not obey constant productivity.
There are seasons in the year and seasons in a life, times to push and times to restore. A self-directed professional can work with that rhythm rather than against it. This is the same point as Gloria Mark's, widened out from the day to the year. The conditions that good judgment needs are not only about this morning. They are about not grinding the instrument down.
Auditing optimism without killing it
Rolf Dobelli, in The Art of Thinking Clearly, is right that we are all prone to wishful thinking. We overestimate what we can achieve, underrate the complexity ahead, and keep our eyes on our own plan while the outside world prepares to disrupt it. The framework here adds one thing he does not.
Wishful thinking is a specific failure where contaminated mindware overrides productive mindware. When your optimism about a project comes from genuine pattern recognition across similar past work, it is not wishful thinking. It is a calibrated prediction from a model that has been tested. When it comes from wanting the outcome rather than from evidence, the untested assumptions are running the show.
The discipline is to audit optimism rather than suppress it. Before you commit to a timeline, a strategy, or a price, ask where the prediction is coming from. Is this my tested models talking, refined by real feedback? Or is it an assumption I have never checked, an industry habit I absorbed without testing, or plain desire wearing the costume of analysis?
Annie Duke's test for this is refreshingly practical. Get comfortable with saying I am not sure. When you can say I think there is a seventy per cent chance this works, and here is what I will do if it does not, you are thinking in bets.
When you say this will definitely work because I want it to, you are thinking in wishes.
The first is your real models in action. The second is untested confidence refusing to admit what it does not know. Certainty in the face of genuine uncertainty is not strength. It is the part of your thinking that has stopped checking.
Looking after your decision-making as a professional practice
The practical upshot is that decision quality is a discipline with maintenance requirements, not a gift you either have or lack. A few things hold it in good repair.
Keep your models current by staying close to the substance of your field. Models drift when your engagement with real conditions lapses. Every genuine encounter with the work refines the predictions your judgment runs on.
Protect the conditions under which you make important decisions. Use your control over your own calendar to put the work that needs your deepest judgment where your executive function is freshest, and build enough slack that the decisions that matter most are not made on a depleted tank.
Audit your reasoning separately from your results. When something works, ask whether the process was sound or whether you were lucky. When something fails, ask whether the process was flawed or whether uncertainty simply landed against you. Done consistently, this refines your models instead of corrupting them with noise.
Treat comfort with uncertainty as a mark of expertise rather than a gap in it. The professional who says I am seventy per cent confident in this direction and here is how we manage the other thirty shows deeper judgment than the one who claims certainty.
I should say plainly that almost none of the independent professionals I have interviewed did this thinking entirely alone. Most had a coach, a mentor, or a strategic partner, someone who could ask the questions they were too close to their own expertise to ask themselves. Auditing your own optimism is genuinely hard, because the same models that serve you also hide their own blind spots. A thinking partner is often what makes the audit honest.
Certainty in the face of genuine uncertainty is not confidence. It is the part of your thinking that has stopped checking.
Frequently asked questions
Can AI make better decisions than an experienced professional?
AI can run an individual model faster and handle more data. The judgment about which models apply to a specific situation, how to weigh them when they conflict, and when to override one because the context holds something the data does not, comes from the many-model thinking that sustained practice builds. AI is a powerful tool to coordinate within your own thinking. The decision about what to do with its output stays with you.
How do I tell whether my optimism is calibrated or wishful?
Ask where the prediction comes from. If it comes from pattern recognition grounded in similar past situations, it is calibrated. If it comes from wanting the outcome and from assumptions you have never tested, it is wishful. Annie Duke's practical test helps: can you put a probability on it and say what you will do if it does not happen? If you find yourself saying it will definitely work, that certainty is the tell.
Does decision fatigue really affect professional judgment?
Gloria Mark's research shows that sustained decision-making depletes the executive function that careful reasoning needs, which makes choices more impulsive and less consistent. For an independent professional this is not abstract. The quality of your most important work depends on the conditions you do it in. Scheduling the decisions that matter when your resources are fresh, and protecting a break before a critical call, directly protects the quality of your output.
References
Michael Mauboussin. Think Twice, and his Knowledge@Wharton interview on the success equation. The source of the experience-versus-expertise distinction and the line that the key to expertise is having a predictive model that works.
Keith Stanovich. What Intelligence Tests Miss (2009). The mindware framework, including the distinction between the productive and contaminated kinds, and the role of the reflective mind in overriding easy but wrong responses.
Scott Page. The Model Thinker (2018). The many-model thinker idea and the argument that ensembles of models outpredict any single framework.
Annie Duke. Thinking in Bets (2018). Resulting, the separation of decision quality from outcome quality, decisions as bets under uncertainty, and getting comfortable with not being sure.
Gloria Mark. Attention Span (2023), and her article on why making decisions is exhausting. Research on how sustained decision-making depletes executive function.
Rolf Dobelli. The Art of Thinking Clearly (2013). The wishful thinking and optimism bias observations that prompted the original version of this article.
Cedric Chin. Founder of Commoncog and a guest on the Wisepreneurs Podcast. His account of Naturalistic Decision Making, tacit models of expertise, and Lia DiBello's research on the business triad and predictive validity grounds the section on experience versus expertise.
Peter Compo. Author of The Emergent Approach to Strategy and a guest on the Wisepreneurs Podcast. His framing of strategy as a single rule for clearing a bottleneck shapes the section on the discipline of deciding against something.
Tonianne DeMaria. Co-author of Personal Kanban and a guest on the Wisepreneurs Podcast. Her observations on judging others by outcomes while judging ourselves by intentions, and on managing energy rather than time, ground the sections on resulting and on cognitive conditions.
Laetitia Vitaud. Writer on the future of work and a guest on episode 71 of the Wisepreneurs Podcast. Her argument about the critical-thinking advantage of experienced professionals, the risk of skills atrophying, and the seasonality of sustainable work grounds the section on staying the author of your own thinking.
Where to take this next
If you want a Monday-morning version of this, pick one decision you are carrying right now and run it through three questions.
- What is the actual bottleneck here, the one thing that makes the rest beside the point?
- Am I judging this by my reasoning or by how the last similar thing turned out?
- And am I about to decide this on fresh resources, or on a depleted afternoon?
Three questions will not remove the uncertainty. They will tell you whether your real models are in charge of the call.
Working out which of your models actually predict, and which assumptions are quietly running the show, is hard to do alone, because you are too close to your own expertise to see its blind spots.
This is what the Positioning and Practice Audit is for: a short, focused engagement to get a clear read on where your judgment is your strongest asset and how to build your practice around it.
The audit fee credits in full against the Marketing Partnership if you continue, read more on the Marketing Partnership page