The study explains that people in general tend to produce more papers and everything else early in their careers, which gives the misleading appearance that their first papers are most likely to be the most successful.
The study is “Quantifying the evolution of individual scientific impact,” by Roberta Sinatra [pictured here], Dashun Wang, Pierre Deville, Chaoming Song, and Albert-László Barabási, published in Science [04 Nov 2016: Vol. 354, Issue 6312].
The scientists also prepared a vivid set of visualizations, which they put online.
And Nature magazine prepared a video that explains it all:
(Thanks to Andrea Rapisarda, who won an Ig Nobel Prize for his own randomness research, for bringing this to our attention.)