Every leadership development programme promises better decisions.
Workshops. Frameworks. Case studies. Cognitive bias checklists.
The promise is wrong.
Not because the tools are useless.
Because a better decision on this problem is not the goal.
The problem will change. The conditions will change. The stakes will change.
What needs to develop is judgment.
The output is the wrong unit
published Noise in 2021. Their finding: the same person, given the same type of decision on different days, produces systematically different outcomes. Not because the facts changed. Because the decision-maker is not a fixed function. Studies across criminal sentencing, medical diagnosis, and financial forecasting found this variability to be as large as bias, and far less studied.
If the decision-maker is variable, optimising individual decisions treats the symptom. The problem is upstream.
Expertise does not predict judgment
ran a forecasting study over twenty years, involving nearly three hundred experts and twenty-five thousand predictions. Domain experts performed no better than well-informed non-experts. Seniority made no difference. What predicted accuracy was a cognitive style: actively open-minded, calibrated about uncertainty, willing to update on new information, comfortable sitting with an unresolved question.
Those characteristics are not acquired by knowing more about the domain. They are acquired by practicing a specific kind of thinking, repeatedly, under conditions with honest feedback.
The environment the development happens in
distinguished kind learning environments from wicked ones. Kind: clear feedback, timely, reliable. Wicked: feedback is delayed, noisy, sometimes absent. Strategic leadership is a wicked environment. The consequence of a decision made in January may arrive in October, filtered through market conditions, personnel changes, and twenty other decisions made in between.
Development designed for kind environments does not transfer to wicked ones. This is the structural problem with most leadership programmes: they produce learning that is valid in the room it happened in and nowhere else.
What judgment actually is
Judgment is not the accumulation of good decisions. It is the underlying capacity that makes reliable decisions more likely across different conditions, with different stakes, under different kinds of pressure.
resolved this in a 2009 paper, a collaboration between two thinkers who had spent decades disagreeing. Klein believed expert intuition was reliable. Kahneman was skeptical. They found both were right. Intuition can be trusted when the learning environment is kind: when it provides clear, timely feedback that allows accurate mental models to develop. In wicked environments, confident intuition is indistinguishable from confident noise.
Developing judgment in a wicked domain means solving the environment problem first.
What the bias curriculum misses
Knowing about a cognitive bias does not remove it. That finding exists inside the same research tradition that produced the biases. The , the observation that people believe they are less biased than others, is one of the most robust findings in social psychology.
Awareness does not produce change. Change requires a different mechanism.
Knowing about a bias and being free of it are not the same thing. The bias literature contains that finding.
How the immersion activates judgment
FusekiB works on the environment problem directly. Go is a wicked domain made kinder: the feedback loop is compressed from months to hours, the consequences of each decision become visible within the same session, and the debrief forces articulation of reasoning before the outcome is fully known.
In a single immersion, a participant exercises judgment dozens of times. Each stone placed is a commitment to a reading of the position. Each sequence that follows either confirms or challenges that reading. The board does not lie and it does not wait.
The debrief then does the second thing: the reasoning is made explicit. Not what you decided, but why, and whether the why was actually connected to what followed. This is the mechanism Hogarth identified as necessary for judgment to develop in wicked environments: the artificial creation of a kind feedback loop that the natural environment cannot provide.
Across immersions, the cognitive style Tetlock found in accurate forecasters, the discipline to update, the capacity to sit with an unresolved question, the habit of separating what is known from what is assumed, becomes practiced rather than aspirational.
The goal is not a better decision on this problem. It is someone who brings different judgment to every problem that follows.
A framework promises a deliverable. A decision-making tool has a measurable output. Developing judgment is a longer, harder commitment, and the results show up in decisions that have not been faced yet, in conditions that could not have been predicted when the development began.
That is precisely why it is the only approach that works. The problem that will require judgment next year is not the problem you have today. The tool that solved today's problem will not be there. The person who developed their judgment will.