Researchers may publish their best work at any point in their careers, a new study reports. This is not the same as success being the result of random forces or just plain “dumb luck.”