Indeed someone who by anyhow know the price of bitcoins would play in millions and billions of dollars ! But is that really possible ? Expert analysts predict a rough picture of future bitcoin price that it is likely to hit this level by going through past successes and some triggering events and by some data like current market cap , daily trade volume etc.
But the question is are these reliable enough? The daily trade volume is totally a probabilistic event as it is like total number of customers that visit your shop at a time. Though some interpolate and come with a idea that if at a given point the daily trade volume is 1 million the trades following would be likely to be around this only. If this is so , can we not come with an improved mathematical interpolation theory, the success of that itself we are not sure but we can give a try.
In a general interpolation we assume a given pair of points continuously trace a complex function not this complex i.e z= x+ iy but a difficult path in z=0 plane that can not be represented by some ordinary or any standard function. Rather we have to take into account every single history of points locus and come with a fitting path that all the points have obeyed. With such we are able to draw upon the future likely to be points and thus have control on prediction.
But in the case of stocks and more particularly bitcoin the interpolation considered should not be general but a weighted interpolation. The weight here of course is the mathematical probability of the points. With such tool i believe it would sort out
the future locus to some extent. As the probability itself takes into account several different pictures affecting price i.e it being a multi-variable probabilistic event , it should help in realizing the sought purpose.