Appears that height of Roland Garros junior champions is a good statistical predictor. If just look at the numbers, it would appear that if a player is 190+ AND wins junior RG he´s on track to top 20, even top 10. If he´s under 188cm the statistical chances are not good
(10 out of 10 players under 190 didn´t make a quick ascend, 3 our of 3 players over 190cm height did. Rublev is close). Interesting.
Hmm, sample size seems far too small to me to be sure that height is the main explanatory factor. And even if it is, probably just indicates extent to which won in juniors because physical development relatively well advanced at a young age compared to peers.
To be really useful, you would have to include the others who played each year and look at who succeeded and failed and whether or not winning/height are well correlated.
In fact if you look at recent results you might conclude it is better to be runner up than win at junior RG. The two two junior RG finalists in recent years who have made top ten (Zverev and Thiem) both lost at RG; in the last six years, 4/6 runners-up are doing better than the person who defeated them, and I suspect if you looked at the field more broadly the effect would be even stronger:
2016: Blancaneaux (now 482) df Auger-Aliassime (now 231)
2015: Tommy Paul (now 355) df Taylor Fritz (now 128)
2014: Rublev (95)
df Munar (266)
2013: Garin (220) df Zverev (12)
(178) df Peliwo (315)
2011: Fratangelo (132) df Thiem (8)