Well, the price being 252$, it means that everybody who who owns bitcoin thinks that they are worth more than 245$; but, on the other hand, everybody in the world who has some money to spare thinks that 1 BTC is not worth 255$. The latter think that the chances of it being worth more than 1300$ anytime in the next few years are much less than 20%.
I don't know where you get the stat but I feel sorry for them.
It is not a survey, just math. If anyone with money to invest thought that the chances of BTC hitting 1300$ in the next 2 years were higher than 20%, they should buy it even for 300$. But they aren't, so...
Why not 19% or 21%?
Just rounded the number. To get a more accurate value one should consider also the chances of it reaching 1200 but not 1300, 1100 but not 1200, 1000 but not 1100, etc. Then one would get much less than 20% for "will get above 1300".
The 20% number is the upper bound, assuming that those buyers consider only two possibilities: "will get higher than 1300$ sometime in the next 2 years" or "crashes to 0$ right after I buy it, and stays there forever". Then p*1300 + (1-p)*0 < 255, which gives p < 255/1300. If they are considering other cases besides those two (which of course they are), then p must be even less than that.
Anyway, note that it is what "the market" must be thinking, not what I think.
Terrible backwards logic.