What Makes Someone Good at Math: The Science of Mathematical Talent

Why does one person win gold at an international mathematics olympiad at age 13, while another struggles to master the multiplication table? Some people solve millennium problems; others spend years redoing the same textbook exercises. The gap seems vast — but what actually explains it?

Is it biology, psychology, motivation, environment, or genetics? With AI systems, the answer is relatively simple: processor power, parameter count, network architecture, training data — and you can explain why one model outperforms another. But what about people?

The key point is this: mathematical talent is real, but it has no single cause. It emerges from several things simultaneously — a distinctive pattern of brain activation, strong working memory and logical reasoning, heredity and personality, and, crucially, practice.

Here is what current research tells us.

How Mathematical Talent Works in the Brain

Math operates through “number sense,” not language.

In 2016, a study placed 15 professional mathematicians and 15 humanities scholars of equivalent academic standing in an MRI scanner and asked them to evaluate complex statements in algebra, analysis, topology, and geometry. In the mathematicians, the activated brain region was entirely distinct from the one associated with language processing.

Mathematical thinking is fundamentally non-linguistic.

It relies on the same neural network involved in simple counting — and even in merely looking at numerals. The circuit a mathematician uses when working through topology is the same one that activates in any person when they see numbers or do mental arithmetic. Advanced mathematics is a “superstructure” built on top of an ancient sense of number common to all humans — and even to infants and animals.

Mathematically gifted people do not have a “different” brain. They simply use this shared circuit more deeply and efficiently.

This may also explain, incidentally, why large language models still cannot produce genuinely original mathematics.

The Cognitive Foundations of Mathematical Talent

Three factors predict mathematical ability most reliably: logical reasoning, memory, and spatial thinking.

The strongest predictor of future mathematical success is fluid intelligence — the ability to detect patterns and reason through novel situations without relying on memorized knowledge. This single factor accounts for roughly 90% of individual differences in mathematical performance.

Two additional factors complete the picture: working memory (how much information you can hold in mind simultaneously while solving a problem) and the ability to mentally rotate and visualize spatial objects.

Genetics and Mathematical Talent: What We Actually Inherit

About half of mathematical ability is heritable — but there is no single “math gene.”

Thousands of genes contribute to mathematical ability, each adding a tiny increment. No single gene is responsible.

What is notable is that the role of genetics grows with age. In adulthood, innate potential becomes more visible because people naturally gravitate toward activities that suit them and gradually approach their own ceiling.

How Practice Shapes Mathematical Ability

The brain literally restructures itself through mathematical work.

Mathematicians show denser gray matter in the parietal and frontal regions — but the longer someone has studied mathematics, the stronger this effect becomes.

This means the structural differences are not purely innate: the brain changes as a direct result of practice itself.

Early Development and the Limits of Drilling

Early mathematical ability is the single strongest predictor of future academic performance — not just in math, but across all subjects.

That said, “simply practicing more” accounts for only a small fraction of the variation between individuals. Practice is necessary, but it does not by itself produce exceptional mathematical talent. (Though it should be noted that this research predates the smartphone era, let alone AI.)

What This Means for Math Education and Talent Development

The conclusion is this: mathematical ability is far from purely innate. The brain region responsible for mathematical thinking is distinct from language processing. The brain grows and restructures through practice. Good teaching and a rich learning environment make an enormous difference.

The only remaining question is what, exactly, to teach — and how, and when. But that is a different story.