Objective evidence of rising population ability: a detailed examination of longitudinal chess data

Rising population intelligence should be impacting in the real world and performance in intellectual games such as chess may be good indices of such impact. Previous research suggests that chess players recently have been reaching high performance levels at younger and younger ages, consistent with rising population ability, but this conclusion is based on very limited data. Sample size was small and data mostly were only from 1970. The present study was able to use all international chess rating list data and some new data back to 1880 to investigate this issue thoroughly. Longitudinal data show that the age effect really is something quite new, not being apparent before 1970. The age effect mostly is occurring in the very top players, who are getting on the list at progressively younger ages, but also is occurring in female players, though not as strongly. The age effect mainly seems due to progressively more very talented young players entering the rating list very young and quickly rising. The data also allow some other interpretations of the age effect to be partly excluded. Top players do not seem to be getting any more practice and the peak rating age still seems to be around 35 years old, as it has been for many decades. The effect thus is unlikely to be due to recent changes in the chess environment. The best interpretation is that the natural talent pool has greatly improved in recent years, consistent with rising population ability.