Ooh, top 100 is way too high in crypto.
I have heard of people doing a similar (theoretical) experiment with the very bottom 10 coins with defined market caps. The results were very positive, mainly from one or two coins which Mooned while the others did nothing or died.
That's pretty much the idea of it, although it's meant to be around 8/10 increase while 2/10 fail (ordinary shares)