[george_proost]

Hi dwh,

Thanks for the reply... but no not really that way, and certainly not with that effect. Let me try expalin it in another way.

standard RIP logging - running totals in brackets

period1.
=====
Jack referes Jill and Simon -- Jack(2 RIP's)

period2.
======
Jill refers Pete -- Jill(1 RIP) and Jack(2,5 RIP's)

period3.
======
Pete referrs Paul --- Pete(1 RIP) Jill(1,5 RIP's) Jack(2,75 RIP's)

If the period to be measured is 'period 3' the actual standings would be :

Pete(1 RIP)
Jill(0,5)
Jack(0,75)

a member therefor ALWAYS for any given period WILL have a RIP value if a 'child' refereree at any level were to obtain a registration.

If Jack had refered 1 more member in period 3 in the current system of referals then would be a tie 1 RIP each. In the RIP system at the start of the period, depending on how large the referral tree is of a member and how deep there will ALWAYS be a RIP value ... albeit a very small one eg:

--1 direct referal in period
--0,5 referal once-revomed
--0,25 referal twice-removed
--0,125 ..
--0,0625 ..
--0,03125 and so on

This accurately measures the effect of refferals of a user over time within a given time period.

Hope this is clearer ..