Are probabilities predictive and statistics a posteriori ?

1. Oct 10, 2015

jk22

could we say like the title that if we throw coins n times the result were s=nhead/n this were statistics and we could find a prediction for the next toss p(s)=?

I thought we could reason like : the next tosses shall equilibrate towards 1/2 so for example p(s)=1-s ?

2. Oct 10, 2015

FactChecker

You can say that. But the probabilities don't have to be predictive. Even after a coin toss has been done, if you don't see the result, you use a 50/50 odds for guessing the result. That is not predictive, since the toss has already happened.