There are definitely the right number of "N's" in the formula whether you
start a 0 and go to k-1, or you start at 1 and go to k. However, the variable
T (the total claim count) is defined as N(1)+...+N(k). No mention is ever
made in the paper of N(0).
I realize the notation is arbitrary and can start with a subscript of 0 or 1,
but it seems like the paper establishes indices from 1 to k for N and R, then
the formula for S(1,i) comes along and shows indices from 0 to k-1.
/external
/to yin.lawn@cna.com
/end