The Dreyfus Model: Five Stages from Novice to Expert
From rules to intuition — the five stages and how to move through them
What are the 5 stages of the Dreyfus model?
The Dreyfus model names five stages of skill acquisition — novice, advanced beginner, competent, proficient, and expert — describing a shift from following context-free rules toward intuitive situational understanding. Novice: applies rules literally, with no sense of which ones matter. Advanced beginner: has enough real cases to notice situational features no rule mentions. Competent: overwhelmed by relevant detail, adopts a deliberate plan to decide what to ignore — and, because the plan is theirs, begins to feel genuinely responsible for outcomes. Proficient: no longer chooses a frame consciously, as the situation presents itself as already meaningful, though deciding what to do still takes deliberation. Expert: perception and action fuse — sees what the situation demands and responds without detaching to reason about it. Later Dreyfus writing adds a sixth stage, mastery, in which the performer seeks to expand their intuitive repertoire and transform the domain itself. The model is widely used in professional education and nursing pedagogy, but it is observational and philosophically derived rather than experimentally validated — and researchers including Gobet and Chassy have argued that no evidence establishes the stages as genuinely discrete, and that experts demonstrably do use slow analytical reasoning, such as look-ahead search in chess, where the model predicts pure intuition.
The Dreyfus brothers developed their model partly as a critique of artificial intelligence: they observed that human experts do not operate by applying rules — they respond intuitively to situations as wholes. The novice, by contrast, has nothing but rules. Moving from one to the other is not just a matter of accumulating more rules; it involves a qualitative shift in how situations are perceived and responded to. The five stages of skill acquisition they named — novice, advanced beginner, competent, proficient, expert — track that shift along two axes at once: what the learner attends to (context-free rules, then situational features, then whole situations), and how detached they are from the outcome (following instructions, then owning a plan, then absorbed in the work). The model first appeared in a 1980 Dreyfus & Dreyfus report and was popularised for nursing education by Patricia Benner’s "From Novice to Expert" (1984), which is why most people meet it in a clinical-training context. It is worth being clear about what it is: a philosophically derived, observationally grounded account of how expertise feels from the inside, not a set of stages validated by controlled experiment. It is useful as a map of where you are and what changes next — not as a measurement instrument. Below are the core practices, each with the mechanism that makes it work and an honest read on the evidence.
Practices
- Follow explicit rules and procedures without improvisation at the novice stage
When new to a domain, follow the given rules precisely — understanding comes after compliance, not before.
- Accumulate varied situational episodes to move from novice to competent
Build a library of real cases — not just more rules — to develop situational recognition.
- Allow yourself to be emotionally invested in outcomes at the competent stage
Genuine commitment to outcomes — feeling the loss of failure — is what drives the transition from competent to proficient.
- Practice reading the whole situation, not just tracking individual elements
Develop the ability to perceive a situation as a unified whole rather than as a checklist of components.
- Protect expert intuitive response from rule-reversion under pressure
Experts perform best when they trust their trained intuition; retreating to explicit rules under pressure is often a regression.
- At the expert stage, use deliberate reflection to continue growing
Expert performance is intuitive, but expert development requires deliberately stepping back to examine the intuition.
How these practices connect
Each stage of skill development runs on a different cognitive mechanism, and the practice that works at one stage is often the mistake at another. Rule-following builds the contextual library a novice lacks; accumulating varied episodes turns that library into situational judgment; emotional stake deepens the encoding of what the episodes teach; and once enough episodes are encoded, the system shifts from checking rules against situations to perceiving situations as wholes. The single mechanism underneath all six practices is that expertise is built by matching the type of learning input (explicit rule vs. lived episode vs. emotionally-charged outcome vs. deliberate reflection) to what the learner's current stage can actually use.
Should I just follow the steps exactly, or is it okay to start improvising?
If you're new to the domain, follow the steps precisely rather than improvising. A novice has no contextual experience yet to know when a rule should bend, so improvising substitutes uninformed judgment for the compressed wisdom already built into the rule. The rules aren't the destination — they're scaffolding that keeps you engaged with the domain long enough to start noticing the situations where a rule works well and where it doesn't, which is exactly the raw material the next stage runs on.
I've been doing the same drill over and over and I'm not improving anymore — what's going on?
Repeating the same type of situation until it's automatic produces high competence in that one context but doesn't generalize, because the mechanism that drives you from novice to competent is a library of varied episodes, not repetition of one episode. Situational judgment comes from recognizing patterns across multiple different cases — the maxims that tell you when to bend a rule can't be taught directly, they have to be extracted from variety. If your practice has stopped varying, you've likely plateaued a rule-following stage rather than moved past it.
Does it matter if I actually care about the outcome, or is just showing up and doing the reps enough?
It matters. The shift from competent to proficient specifically requires genuine emotional investment — feeling the loss when something fails, not just logging that it failed. That emotional charge is what causes the encoding of successes and failures to go deeper, driving the kind of self-motivated analysis that plain repetition doesn't produce. Staying in a consequence-free practice environment past basic competence feels safer, but it blocks exactly the encoding mechanism that moves you forward.
Why do experienced people seem to just 'see' the right answer instead of working through a checklist?
Proficient and expert performers respond to the whole situation as a gestalt rather than decomposing it into a checklist of tracked elements, and this holistic read is faster and more accurate for complex, moving situations than sequential rule-checking. It develops only after extended exposure to varied situations with feedback has trained the perceptual system to recognize higher-order patterns across features — which is why it shows up after the episode-accumulation stage, not before it. Reaching for a framework first is appropriate for a novice, but for someone with sufficient experience it actually prevents this holistic recognition from forming.
If an expert chokes under pressure, should they go back to basics and follow the rules again?
Usually no, and this is where the model's stages actively conflict with common coaching instinct: expert performance runs on fast, intuitive response built from a long-accumulated pattern library — what gets called 'absorbed coping' — and under pressure there's a pull to retreat to slower, explicit, analytical rule-following. For a genuine expert, that retreat typically degrades performance rather than stabilizing it, because it de-automatizes a fluency that took years to integrate. The exception is a truly novel situation the expert hasn't encountered before, where deliberate analysis still has the advantage even for experts.
Once you're actually an expert, is there anything left to work on?
Yes, but the work changes shape. Expert performance is intuitive, and that same intuitive fluency resists introspection, which means it also resists getting better on its own — the expert has to deliberately step outside absorbed coping and examine their own patterns as a learner again, not just perform them. This is a qualitatively different activity from the rule-following that built the novice's foundation, but it answers the same underlying need: expanding the library of situations the intuition can draw on. Most people who reach functional mastery stop doing this, which the model treats as a choice rather than a ceiling.
How do you choose between these?
Pick the practice based on what the immediate situation actually calls for, not on which stage you generally consider yourself at. In a genuinely unfamiliar situation, explicit rules or deliberate analysis are the right tool even for someone who is expert elsewhere, because intuition has nothing to draw on yet. In a familiar, high-pressure moment where you have real accumulated experience, trust the intuitive read and resist the urge to fall back to a checklist. When you notice practice has gone flat despite repetition, the fix is more varied episodes, not more reps of the same one. And when performance has plateaued at a comfortable level, the question to ask is whether you've stopped being emotionally invested in the outcome, or stopped deliberately examining your own intuitive patterns — both are stage-specific stalls with different fixes.
What these approaches do not settle
The Dreyfus model is a theoretical account built from cross-domain observation, not a set of trialed interventions, and several of these mechanisms — situational maxims, holistic perception, absorbed coping — are inferred from behavior rather than directly measured, so the model has been criticized for overstating how discrete the stages really are. The practices can also cut against each other if misapplied: emotional investment helps encoding but can tip into performance-impairing anxiety, and expert intuition is domain-specific enough that the same trained response which looks masterful in a familiar case can produce a confident, wrong answer in an unfamiliar one. Treat the stage boundaries as a useful approximation of a mechanism, not a precise diagnostic of where any one person stands.
Practice this with IX Coach
Related concepts
- Tacit Knowledge: The Skill You Can’t Fully Explain
What expert knowledge is made of, how it is acquired, and why it resists documentation
- Mental Representations: The Expert’s Internal Map
The cognitive structures that separate expert performance from prolonged repetition
- Deliberate Practice, Not Just Practice
Focused, feedback-driven practice at the edge of your ability