Recent review article on Science:
Recent discoveries on the acquisition of the highest levels of human performance
A Güllich, M Barth, DZ Hambrick, BN Macnamara | Science, 2025 https://www.science.org/doi/10.1126/science.adt7790
Editor’s summary …Güllich et al. looked at published research in science, music, chess, and sports and found two patterns: Exceptional young performers reached their peak quickly but narrowly mastered only one interest (e.g., one sport). By contrast, exceptional adults reached peak performance gradually with broader, multidisciplinary practice. However, elite programs are designed to nurture younger talent. —Ekeoma Uzogara
The authors assemble 19 datasets covering 34,839 adult world-class performers across sports, science, (opera) music, chess and academia—the paper extracts numerical results reported in prior papers to convert and summarize values in consistent numbers. In other words, key predictors (e.g. early performance level, age milestones, (inter)-discipline practice) are aligned conceptually across studies and compared between world-class and near-world-class performers using standardized summary statistics reported in the original literature, presented in Table 1 of the paper.
From the author’s sample:
- Early stars and adult superstars are mostly not the same people.
- High early performance does not predict the highest adult performance.
- Early specialization helps early success but hurts peak success.
- Early breadth (multidisciplinary practice) predicts world-class outcomes. World-class performers progress more slowly early in their careers.
- Standard talent-selection and training models are likely misaligned with long-term excellence.
While the paper’s analysis method is potentially prone to selection bias, I find it raised really interesting questions that human (well, parents) are long concerned but never answered: since prodigies are not 100% to become world class masters and existing expertise theories are insufficient to explain these patterns, what is the mechanism that leads to greatness?
(IF, you can treat greatness as a binary institutional label) Also, note how the analysis is silent on the base rate problem—how vanishingly rare world-class outcomes are among ordinary participants. It’s ~10% overlap between prodigies and masters. If thaty’s too weak, what about all population on earth?