# Index of coincidence for completely random text over k alphabets

I'm confused about /Friedman's method 2 (using a table with column of k size of key, and n/k rows, n is the total size for cipher text)

$$m = \Bigl(\varphi(l)-\varphi(o)\Bigr)/\Bigl(\varphi(T)-\varphi(o)\Bigr)$$

P.S.
$\varphi(l)$ - stands for index of coincidence for language
$\varphi(T)$ - stands for index of coincidence for language
$\varphi(o)$ - stands for index of coincidence for completely random text

Just don't know what is the value for $\varphi(o)$ how we calculate it? and m - stands for how many slice we divide the cipher text to analysis or the key size?

-