Where do the magic numbers (p, b1-b5) come from? Are these empirically found using curve fitting?

Also, wouldn't it be better to write things like

   (+ (- (* (exp (* (- r) t)) strike) s) call)))

in a more schemer-friendly way as

  (+ call (- s) (* strike (exp (- (* r t)))))

to avoid parenthesis syndrome?

(I was probably wrong to expect self-explanatory code, e.g. with docstrings, but IMO the readability could be improved)