Redefining Excellence: New Science Study Reveals How Top Human Performance Is Acquired

Journal: Science (December 2025) | Scale: 34,839 world-class performers | DOI: 10.1126/science.adt7790

Background

Traditional talent-development theory assumes that top achievement stems from early exceptional performance (e.g., child prodigies), specific innate abilities (high IQ, athletic talent), and years of intensive domain-specific training. This model dominates education selection and elite training systems worldwide, favoring early identification and forced early specialization.

An international interdisciplinary team led by Professor Arne Güllich argues that most prior research focused on junior or sub-elite groups, missing the true early trajectories of world-class performers. This Science review uses large-scale data to reveal the nonlinear path to the top.

Key Findings

• Early stars are not future legends. Child prodigies and adult world champions are largely different people. Only ~13% of international-level athletes remain top in both junior and adult stages. 82% of junior international-level athletes never reach the adult international level, while 72% of adult world champions were not top juniors.
• Gradual development paths. Future world-class performers rise relatively slowly. Among chess grandmasters ranked world top-3, their average Elo rating at age 14 was *lower* than that of players who later ranked 4-10. Early speed does not determine final height.
• Breadth beats early specialization. Future Nobel laureates often explored multiple scientific fields and had non-scientific hobbies; top athletes typically played multiple sports in adolescence rather than single-sport drilling. This breadth appears to lay the foundation for later depth.

Proposed Mechanisms

The study offers three hypotheses explaining why broad early exploration outperforms early deliberate practice:

1. Search-and-match: Exposure to multiple disciplines increases the probability of finding the domain best matching one's talents and preferences, optimizing person-domain fit. 2. Enhanced learning capital: Diverse learning experiences build transferable cognitive skills (pattern recognition, adaptive learning), enabling faster learning and greater innovation during later specialization. 3. Limited risks: Avoiding over-specialization reduces physical injury, psychological burnout, and opportunity costs (e.g., lack of general education), preserving career longevity and motivation.

Conceptually: WorldClassPerformance = (OptimalMatch + HighLearningCapital + LowRisk) * SustainedEffort

Implications for Education and Talent Development

• De-label: Do not label children as "gifted" or "mediocre" too early.
• Encourage breadth: Support deep engagement in two or three different domains (e.g., language and math, music and science) rather than single-track drilling.
• Long-horizon evaluation: Training programs should be assessed on long-term peak performance, not short-term gains.

As Einstein—great physicist and passionate violinist—demonstrated, excellence often grows from diverse soil.

---

Source: Science, DOI: 10.1126/science.adt7790