It would make sense that there's always a solution, since I'm pretty sure that equations of the form $a^{x}$ diverge faster than $x^{a}$, but does anyone have a proof?
EDIT: Whoops, sorry, I meant $P(x)-{b}^{x}+1$.
EDIT2: I really fell short on this question. I also meant $b>1$