How Experts Solve Problems

Watch an expert and a beginner face the same problem, and the difference is not mainly speed or memory. It is what they see. The expert looks at a situation and immediately recognises a familiar shape, while the beginner sees a jumble of disconnected details and starts hunting for something to try.

Understanding this changes how you learn any craft. Expertise is not general cleverness sprinkled over a field; it is a deep store of patterns built inside one domain. Knowing that tells you where to put your effort: not in memorising tricks, but in seeing enough real examples that the patterns start to jump out on their own.

Perceiving patterns, not pieces

The most studied difference between experts and novices is perception. Experts carve a problem into a few meaningful units rather than many small ones. A physicist reads "an inclined plane with friction" where a student reads a scattered list of numbers and angles. Because the expert's eye groups the situation into known types, the relevant principle often arrives before any calculation begins.

Working forward from principles

Novices tend to reason backward: they grab the unknown they want, find an equation containing it, and work back toward the givens, often thrashing through dead ends. Experts more often work forward — they classify the problem by its deep structure, recall the principle that governs that type, and move steadily from what is given toward the answer. Studies of physics problem-solving in the 1980s, notably by Michelene Chi and colleagues, found that experts sort problems by underlying principle while novices sort by surface features like "pulley problems" or "spring problems."

Understanding before acting

Counter-intuitively, experts are often slower to start. They spend a larger share of their time building a rich representation of the problem — restating it, drawing it, checking what type it is — and less time in frantic activity. That front-loaded understanding is what lets the later steps run smoothly, because they are solving the right problem.

Give a chess master and a club player five seconds to view a position from a real game, then ask them to reconstruct it. The master places the pieces almost perfectly; the club player manages only a handful. The master is not photographing the board. They are recognising familiar configurations — a castled king, a known pawn structure — and storing each as a single chunk. A few chunks reconstruct twenty-odd pieces, because each chunk unpacks into several.

Now scatter the same pieces at random across the board, in an arrangement that could never arise in play. The master's advantage nearly vanishes; they recall only a few more pieces than the beginner. If expertise were simply superior memory, the master should still win easily. It does not survive, which shows the skill is pattern recognition tied to meaningful positions — not raw recall.

William Chase and Herbert Simon, at Carnegie Mellon, published "Perception in Chess" in Cognitive Psychology in 1973, building directly on the Dutch psychologist Adriaan de Groot's earlier work. They briefly showed players board positions and asked them to reproduce them. On positions from genuine games, stronger players reconstructed far more pieces than weaker ones. On randomly arranged boards, the gap collapsed.

Chase and Simon explained this with chunking: masters had learned to see the board as a small number of familiar groups rather than many separate pieces, letting each chunk carry several pieces at once. They estimated that reaching mastery involves acquiring a very large vocabulary of such patterns — later estimates run to tens of thousands. The study is a cornerstone of expertise research and has been widely replicated across other domains.

• Experts just have better memories. The random-board result shows the advantage is pattern-based, not general memory.
• Experts are faster at everything. They are often slower to start, investing more time in understanding first.
• Expertise transfers across fields. It is largely domain-specific; a chess master has no edge at physics.
• Talent explains the gap. The dominant factor is a large, hard-won store of domain patterns.

Take three problems you recently solved in a field you know well. For each, write down the type of problem it was and the principle you used — not the surface details. Notice how quickly the type comes to mind here versus in a field you barely know. That difference is your pattern store at work.

Think Like a Maester: Before reaching for a method, ask what familiar shape this problem already resembles.

1. In Chase and Simon's study, why did the master's advantage disappear on random boards?
2. What does "chunking" mean, and how does it help recall?
3. How does forward reasoning differ from the backward reasoning novices often use?
4. Why might an expert spend more time than a novice before starting to act?
5. What does it mean to say expertise is "domain-specific"?

Experts solve problems differently not because they are faster or have better raw memory, but because they perceive meaningful patterns, reason forward from deep principles, and invest time in understanding before acting. Chase and Simon's 1973 chess studies made this concrete: masters recalled real positions far better than novices, but lost that edge on random boards, revealing pattern recognition rather than memory. The practical lesson is that expertise is built by seeing many real examples until the structure of problems becomes visible at a glance.

• Design Thinking
• Solving Real-World Problems