You seem to be having a problem with the definition, and the use of the word multiplicity. If you go back to the Wolfram definition of multiplicity you linked you will see that it refers to a power series example.
Well you can regard $f(x)=(x-1)^2$ as a power-series expansion about $x=1$, equivalent to: $$0\cdot (x-1)^0+0\cdot (x-1)^1+ 1\cdot (x-1)^2 + 0\cdot (x-1)^3+0\cdot (x-1)^4 \dots$$
and the example transfers over, with $$f(1)=(1-1)^2=0; f'(1)=2(1-1)=0; f''(1)=2\neq 0$$
It has multiplicity $2$ at $x=1$ in this sense because the first two derivatives are zero, but not the third.
But really the power series example is best regarded as a generalisation of the simple case of polynomials, and the natural definition of multiplicity in the case of polynomials is normally that the multiplicity $p(x)$ at $a$ is the highest power of $x-a$ which is a factor of $p(x)$. This is like saying that $2$ is a double factor of $12=2^2\times 3$.
It is also true that $p(x)$ has multiplicity at least $r$ at $x=a$ if the first $r$ derivatives (starting with $f^{(0)}=f(x)$) evaluate to zero at $x=a$. If $f^{(r)}(a)\neq 0$ the multiplicity is exactly $r$. But this is not so natural a definition in elementary work with polynomials.