About three hours into a two-day simulation with a senior leadership cohort from a large manufacturing company, something changed in the room. The scenario had been running well. The team was engaged. The business problem was complex enough to require genuine collective intelligence, and the group had been working collaboratively and thoughtfully. And then, as the time pressure increased and the resource constraints tightened and the complexity of the interdependencies became impossible to manage comfortably, one of the most senior leaders in the group, a person who had spent six months in a leadership journey consistently articulating a sophisticated understanding of inclusive decision-making, began quietly but systematically taking the decisions himself.
He was not aware he was doing it. Nobody called it out in the moment, because it was happening in increments, each one individually defensible. A suggestion offered with enough authority that it functioned as a direction. A question that closed off a colleague’s line of thinking rather than opening it further. A commitment made to another team in the simulation without consulting the people beside him. A summary of the group’s position that was slightly ahead of where the group’s conversation had actually arrived, so that by the time anyone noticed, the position felt established rather than proposed.
By the end of the session, the group had produced an outcome that was largely driven by one person’s judgment in a simulation specifically designed to require and reward collective intelligence. The individual outcome was not bad. It was, in fact, reasonably good. But it was not what the scenario was designed to produce, and the process by which it had been reached had reproduced, in a controlled setting, exactly the leadership pattern that six months of programme work had been designed to change.
In the debrief, I put the observation on the table. What I saw. What the group experienced. What the data from the simulation’s output logs showed about how the decisions had actually been made. It landed. Not comfortably, but cleanly. And what emerged over the following forty-five minutes was something that six months of classroom learning had not produced: a genuine, specific, personal conversation about the gap between this leader’s stated understanding of inclusive leadership and his actual behaviour under the conditions that most closely approximated the conditions in which that understanding was most needed.
That gap is what simulation is for.
The training transfer problem and why it is structural rather than individual
The persistent gap between what people know and what they do under pressure is one of the most well-documented and least resolved problems in human performance research. The estimates of training transfer, the proportion of formally trained behaviour that reliably appears in sustained workplace performance, vary across studies but the direction is consistent: somewhere between ten and thirty percent of what is learned in formal training contexts translates into observable behavioural change in the workplace. The precise numbers are less important than the direction of the evidence, which is that the default expectation of significant transfer is systematically optimistic.
The mechanisms are well-understood even if the solutions remain elusive. Formal training occurs in conditions that are optimised for learning in specific and important ways: low stakes, ample time for reflection, social norms that support curiosity and experimentation, absence of real consequences for error, the presence of skilled facilitators oriented toward learning rather than performance. These conditions are genuinely conducive to learning. They are also, in almost every important respect, the opposite of the conditions in which the learning needs to apply.
Workplace application occurs under conditions of time pressure, real consequences, social complexity, hierarchical dynamics, and the accumulated cognitive and emotional load of everything else that is happening simultaneously. The cognitive and emotional resources that behaviour change requires, the deliberate override of established habits, the retrieval and application of new frameworks under conditions of stress, the maintenance of reflective awareness while also attending to the immediate demands of the situation, are precisely the resources that are depleted by the conditions that characterise real workplace performance. The gap between knowing and doing is not primarily a motivation problem. It is an encoding problem: the knowledge was encoded in conditions that are too different from the conditions of application for the transfer to occur reliably.
What simulation changes about this equation
Simulation addresses the transfer problem not by eliminating the gap between learning conditions and application conditions but by dramatically reducing it. A well-designed simulation creates an environment that is close enough to real organisational conditions that the same patterns, habits, and responses that operate in real contexts are activated in the simulation. The pressure is real enough, the stakes feel real enough, the social dynamics are real enough, that the leader cannot simply perform the new framework they have been learning. They have to actually deploy it, or they fall back on the older, more deeply encoded patterns that have been operating for years.
This is both the specific value of simulation and the specific reason that simulation is often more uncomfortable for participants than other developmental methods. The simulation is not asking leaders to demonstrate what they have learned. It is creating the conditions in which what they have learned either holds under pressure or reveals itself as not yet deeply enough encoded to hold. Both outcomes are valuable. The leader who discovers that their new framework holds under realistic pressure has confirmation that genuine change has occurred. The leader who discovers that it does not has information of a different and more immediately actionable kind: they know specifically what conditions activate their defaults, and that specificity is the foundation of genuine behavioural change.
What simulations reliably reveal that other methods cannot
Three patterns emerge with sufficient consistency across simulation contexts to warrant specific attention as diagnostic of leadership development needs.
The first is the hierarchy response under pressure. A substantial proportion of leaders who are genuinely collaborative in low-stakes and low-pressure environments revert to directive, unilateral decision-making under time pressure with significant stakes. The reversion is rarely fully conscious. It is activated by specific environmental cues, the experience of time running out, the sense that the collective process is too slow, the anxiety of not being in control, that trigger patterns of behaviour that were established and reinforced over years of professional socialisation in hierarchical organisations. The simulation makes this pattern visible as observed behaviour rather than self-report, which is considerably more powerful as developmental data because the leader cannot manage it, dismiss it, or contextualise it away in the way they can manage their own account of their behaviour.
The second pattern is information management under resource scarcity. When the simulation creates conditions of genuine competition for limited resources, a significant proportion of leaders who would describe themselves as collaborative and transparent begin to manage their information strategically: sharing selectively, withholding when sharing might disadvantage their team, framing information in ways that serve their immediate interests rather than the collective intelligence of the system. This pattern frequently surprises the leaders in whom it appears, because it contradicts their self-understanding in ways that they find genuinely uncomfortable. That discomfort is the developmental material. The leader who discovers this pattern about themselves has access to a quality of self-knowledge that would take months of reflection to arrive at through other developmental methods, if it arrived at all.
The third pattern is the management of being wrong. Some leaders reveal, in simulation contexts where the consequences are learning rather than career, that they are considerably less graceful under conditions of genuine error than they believed themselves to be. The defensive justification, the rapid attribution to circumstances, the reluctance to name the error directly and move on, appear in the simulation with the same character they have in real organisational life, but in a context where they can be named and examined without the political costs that would accompany naming them in the real situation. The debrief conversation about this pattern frequently opens up the most important developmental territory available in the programme.
The debrief as the developmental moment
The simulation is not the development. The simulation creates the behavioural data. The debrief creates the conditions in which that data becomes learning. The quality of the facilitation in the debrief room is what determines whether the simulation produces genuine insight that changes something real, or produces an interesting story that the leader tells about themselves without it changing anything.
A good debrief is specific about behaviour rather than general about character. It names what was observed, in specific terms, without collapsing into evaluation of who the person is. It invites the leader’s own account of their experience before offering an external observation, because the gap between the leader’s account and the observation is often itself the most diagnostic piece of data available. It connects what was observed in the simulation to what the leader recognises from their actual organisational context, so that the learning does not remain contained in the fiction of the scenario but travels back to the real situations where it is consequential. And it holds the developmental tension between naming what was difficult to see and maintaining the relational safety that makes honest engagement possible.
This quality of debrief facilitation is a specific and demanding skill. It is not well served by either harsh confrontation, which produces defensive shutdown, or by gentle affirmation, which produces pleasant engagement without genuine developmental movement. It requires the specific capacity to hold the leader in contact with something uncomfortable about themselves while maintaining enough relational safety and genuine respect that the contact is developmental rather than merely distressing. This capacity is not universal among facilitators, and it is one of the primary factors that distinguishes simulation-based development that produces genuine change from simulation-based development that produces engaging experiences without lasting impact.
The specific design principles that make simulation work
Not all simulations are equally effective, and the factors that determine effectiveness are specific enough to be worth naming. The scenario needs to be complex enough that it genuinely cannot be resolved by individual excellence alone, but not so complex that the complexity itself becomes the obstacle rather than the leadership pattern under examination. The time pressure needs to be real enough to activate authentic patterns but not so extreme that it simply produces stress without the reflective space needed for genuine engagement. The stakes need to feel real enough that leaders cannot simply perform the desired behaviour, but the psychological safety of the container needs to be established clearly enough that genuine risk-taking in the debrief is possible.
The most important design principle is the alignment between the leadership patterns the simulation is designed to surface and the developmental goals of the programme it is part of. A simulation that reveals significant and interesting patterns that are irrelevant to the developmental work the programme is focused on wastes the most powerful developmental opportunity the programme contains. The simulation should be designed backward from the specific leadership patterns that the programme is most concerned to develop, creating the specific conditions that those patterns are most likely to surface in their authentic rather than their performed form.
A simulation does not tell you what a leader knows about handling pressure. It tells you what they actually do. And the gap between those two things is almost always the most important developmental data in the room.