One of the pillars of the Skillgym experience is that the situation feels like a conversation with a real person. But how do you make sure that a virtual character behaves and reacts like a real one? The answer lies in the models we use to build them, and in a design that gives each character a personality, needs and a standpoint of their own.
Why authenticity matters
Authenticity means that the simulated experience is as close as possible to real life: the same elements, resources and constraints the trainee would meet at work. It also means paying attention to how the character reacts, the words they use to describe the situation, and what they think about the meeting once it is over. Without this, a role play is an exercise in style. With it, the learning value matches that of a real conversation.
The three dimensions of a character
Every Skillgym character is built along three dimensions:
- their social background: age, role, history, context;
- their needs, explicit and implicit;
- their personality, the most critical of the three.
A good set of characters covers both the most common counterparts a trainee will meet and the most difficult ones.
Needs: what the character wants
To describe needs we rely on Abraham Maslow's hierarchy, the five-level model that runs from basic needs to growth needs. Each character carries two layers: explicit needs, which surface naturally in the conversation, and implicit needs, subtler impulses the trainee has to discover and address for the conversation to become truly productive.
This is what makes a character a complex individual rather than an exercise in technique. Practicing with them means listening for hidden triggers, motivations and fears. Developing awareness of the other, in turn, develops awareness of oneself.
Personality: how the character thinks and reacts
For personality we refer to widely accepted models: DISC, which describes how people behave in a situation and how others perceive them, along the four dimensions of Dominance, Influence, Steadiness and Conscientiousness; HBDI, or Whole Brain Thinking, which describes thinking preferences and how people process information; and the MBTI, built on Jung's psychological types, which combines preferences on energy, information, decisions and structure into sixteen types.
The models differ in purpose and angle, which is exactly why their integration is useful. A character is not the mechanical output of one model. It is crafted around the fundamental elements that drive human beings, the same elements these models describe, so that it behaves, decides and processes information in a consistent and recognizable way.
Background and the preparation room
Personality drives behavior, but so does history. That is why, before the conversation, the trainee reads about the character: their background, their personality and approach, their needs; where and why the conversation takes place; what role the trainee plays and what the conversation is for. This lets the trainee decide on a strategy the way they would in real life.
Each situation, one style
Role play works because it lets people practice a difficult conversation in a protected environment and then discuss the outcome. The situation is part of the learning: it is within the situation that words, gestures and attitudes gain meaning. Here we draw on situational leadership theory: there is no single best style, and effective leaders adapt their style to the readiness of the person in front of them. Practicing the most common and the most critical situations builds confidence and creates a sense of familiarity when the real one arrives.
From scripted options to real conversations
In the early generations of digital role play, the trainee chose from a few suggested sentences and the character reacted according to a pre-written map. Today the trainee speaks freely, by voice or by text, and the character reacts to the strategy expressed in what was said.
What makes this possible without losing authenticity is governance. The character's personality, needs and standpoint are defined before the conversation and held for its whole duration: a skeptical character does not become agreeable because the trainee is insistent, and opens up only when the trainee says something that, in that character's logic, deserves it. The system always knows what the character is feeling, because the scenario and the character's design determine it from the strategy recognized in the trainee's words. It never analyzes the trainee: the only input is what they say.
Reviewing the character's reactions
After the conversation, the trainee can go back through it and see, at each moment, what the character was feeling and why: a note over the character's video explains the reaction as it happened. The character then gives their own account of the meeting in first person. This is the information nobody gets in real life, and it is the fastest way to understand the connection between what you said and what it produced.
Why we insist on this
From the start of our work on behavior and leadership we were convinced that characters had to rest on solid foundations, and the evidence has confirmed it. Every person we meet carries a complex structure of traits that shapes how we perceive them, even when we do not notice it. Designing a tool as delicate as a digital role play without those nuances would spoil the exercise and cut off the value that human interaction brings to self-growth. With them, a conversation with a character becomes practice for a conversation with a person.