Michael K. answered 04/26/19
PhD professional for Math, Physics, and CS Tutoring and Martial Arts
Let's use the information we have to see what we can find as an explicit representation for the f(n+1)(x) to see how to prove it by induction...
f0(x) = 1/(2-x) = f(x)
f(n+1)(x) = f0(fn(x))
f1(x) = f0(f0(x)) = f(f(x)) = f(1/(2-x)) = 1/(2 - 1/(2-x)) = 1/[ (3-2x)/2-x ] = 2-x/(3-2x)
f2(x) = f0(f1(x)) = f(f1(x)) = 1/(2 - [ (2-x)/(3-2x) ] ) = 1/[ (4-3x)/(3-2x) ] = (3-2x)/(4-3x)
f3(x) = f0(f2(x)) = f(f2(x)) = 1/(2 - [ (3-2x)/(4-3x) ] ) = 1/[ (5-4x)/(4-3x) ] = (4-3x)/(5-4x)
A pattern can be seen such that...
fn-1(x) = f0(f(n-2)(x)) = f(f(n-2)(x)) = [ (n) - (n-1)x ] / [ (n+1) - nx ]
Does fn(x) follow?
fn(x) = f0(f(n-1)(x)) = f(f(n-1)(x)) = 1/(2 - [ (n) - (n-1)x ] / [ (n+1) - nx ])
After simplification we arrive at the above we get...
fn(x) = [ (n+1) - nx ] / [ (n+2) - (n+1)x ]
This is indeed the expression we would have if we used the representation in bold based on n->n+1