Principle of mathematical induction states that if a subset $S$ of a successor set $\omega$ is also a successor set, then $S=\omega$. In primitive terms, it is formulated as:
if $S \subset \omega$, if $0 \in S$, and if $n^+ \in S$ whenever $n \in S$, then $S=\omega$.
Now, I wonder what if a statement that I want to prove holds for numbers starting from $b \in \omega$ for example. Would it be correct to introduce
a set $L$ that contains all natural numbers that satisfy the statement
a function $s: \omega \to \omega, s(n) = n^+$
a function $r: \omega \to \omega, \begin{cases} r(0) = b\\ r(n^+) = s(r(n))\\ \end{cases}$
a set $S$ that contains all natural numbers for which $r(n)\in L$
and show that $0$, $n$ and $n^+$ are in $S$? Because I can't just start proving inductively from $b$ and say that it holds for all numbers starting from $b$, as the principle of induction clearly states that $0$ has to be in $S$ too. So, I introduce a trick, i.e. mapping to perform induction.
Is it correct to do so? If not, then how to formally show that I can start from any number?