How historical training norms and learning science make the case for smaller groups

A new argument has been circulating in law enforcement simulation: that more trainees in a single Virtual Reality (VR) scenario means better value. The logic, on its surface, seems appealing. When you have more officers trained per session, the math suggests you have greater throughput. But does that align to the actual science of how officers learn?

For over 30 years, the industry standard for simulation-based training has been one to two trainees per scenario, with groups of four reserved for specific collaborative objectives. That norm emerged for several reasons, and the research backs it up.

 

Training mirrors the real world, and the real world sends one or two officers first

Before examining what the research says about group size, it’s worth considering what officers are actually being trained to do. The vast majority of calls for service are handled initially by one or two officers on scene. Crisis contacts, traffic stops, mental health encounters, domestic disturbances, and verbal confrontations are typically handled by one or two officers interacting directly with a subject.

Training designed to build communication skills, de-escalation judgment, and decision-making must reflect that operational reality. When MILO VR is used for de-escalation or crisis intervention, the goal is to develop the individual officer’s capacity to read a situation, select a tactic, and execute it under pressure. That development happens in focused, direct contact with a scenario — not while navigating the noise of seven other trainees vying for the same decision space.

 

The science of simulation group size

The research on simulation-based learning group size is fairly consistent in its direction: smaller is better for complex interpersonal tasks, and gains begin to erode as group size grows beyond four.

A 2025 systematic review published in the Journal of Medical Education and Curricular Development examined optimal group sizes across high-stakes simulation training and found that larger group sizes hindered an instructor’s ability to identify and correct errors, underscoring the importance of selecting optimal group sizes to ensure effective feedback and performance improvement—particularly for procedural training. The same review found that learners consistently preferred smaller groups for debriefing and complex scenarios, precisely because those formats demanded a higher level of cognitive engagement.

A randomized study published in Medical Education found that time spent on non-learning and passive learning activities increased with group size, whereas constructive and interactive learning activities remained constant. Put differently: as you add more participants to a simulation, each trainee spends more time watching and less time actually learning. While there is benefit in pairing observation with reflection, critical thinking, and group discussion, for skills that require active engagement—verbal intervention, reading behavioral cues, choosing a force option—passive observation is not a substitute.

The implications for de-escalation training are particularly clear. Research on the deliberate practice model, which underpins much of MILO’s cognitive approach, consistently shows that goal-oriented training in the presence of a coach who can provide personalized, immediate feedback has successfully improved performance across multiple domains, including sports, music, and combat. That feedback loop is the mechanism through which skills transfer. It is difficult to maintain in a group of two. It is nearly impossible to maintain in a group of eight.

 

What gets lost in a crowd

The National Policing Institute’s framework for de-escalation training describes a demanding set of competencies: verbal and non-verbal communication skills to effectively engage members of the public through active listening, with time viewed as a resource and officers given multiple response options based on agency policy, criminal activity, and goals. These are individual officer skills. They are developed through practice, feedback, and repetition—not through group observation.

When a trainee watches a scenario instead of inhabiting it, they lose something important: the stress response. The physiological activation that comes with standing in front of a hostile subject, deciding what to say and how to say it, is the very thing simulation is designed to replicate. As more participants are added to a single scenario, each trainee spends less time making decisions under pressure and receives less individualized feedback.

Just as important, effective training must address each trainee’s specific needs. One officer may need to improve active listening, another may rely on repetitive commands, and another may need to moderate an increasingly aggressive tone. In an instructor-led environment with one or two trainees, the instructor can adjust the scenario in real time to target individual performance gaps, provide immediate coaching, and allow the trainee to apply the feedback.

Research on collaborative decision-making also raises a caution that applies directly here. Studies on group problem-solving have found that increasing group size can actually decrease group performance for some problems—particularly those where the correct response is not easily demonstrable to other group members. De-escalation judgment is precisely that kind of problem. There is no clearly signaled “right answer” that a group of eight will converge on together. The decision is internal, individual, and shaped by training that has been delivered personally and corrected in real time.

 

The instructor’s role doesn’t scale the same way

One of MILO’s core commitments is to the instructor. The simulation is the tool; the instructor is the training. An instructor managing eight simultaneous trainees in a branching VR scenario is an instructor whose attention is spread too thin to do their job well.

The question worth asking is: who benefits from eight or more trainees in a single session? It isn’t the officers, whose hands-on time and instructor access are reduced by every person added to the group. It isn’t the instructor, whose ability to observe, intervene, and debrief is diluted at scale. The benefit (if there is one) is administrative. It moves bodies through a training block faster. Bodies are a throughput metric, not a learning metric.

 

Smaller groups, stronger outcomes

The historical norm of one to two trainees, or occasionally up to four for team-based objectives, reflects the accumulated experience of an industry leader who has been running simulation training for three decades. It reflects how officers actually work and the cognitive science of how skills are acquired under stress. Instructors know from standing in training rooms that the moment they lose the ability to watch every trainee closely and respond to what they see, the training starts to degrade.

MILO VR is built around human performance, including decision-making, communication, and the kind of judgment that develops through deliberate, coached practice. That’s a proven philosophy that succeeds at one, two, and sometimes four. More participants per scenario isn’t a training advantage.

With resource and budget constraints, it’s especially important for leaders to use caution when considering new training technologies and features. For the skills MILO VR is designed to build, we don’t do trade-offs that cost officers the very thing training is supposed to provide. Because just like you, we follow the evidence.