The first serious attempt to create a rarity index (IR) applied to the slot car that I know, was given by Roger Gillham in his book of Scalextric (I could not tell you which editions). Roger Gillham just want to give an indication of rarity, it does so from a qualitative and arbitrary character, so a greater or lesser extent such rarity index is influenced by his desire.
---------------------------------------------
At the opposite pole is the rarity index applied to the slot car Rüdiger used by Marx in his book on Stabo car. He recounts the population in whole numbers from which to create statistics. An index of rarity very cold, but very accurate to reality because it is absolutely quantitative.
------------------------------------------------
In this post people are giving very good advice and here's mine advice :
I do not advise making a qualitative index of rarity completely or fully quantitative, but I advise taking the middle path.
Mick, I advise you to do an index of rarity only of what you like, not the whole entire global slot and that this applied statistics to personally use what you like.
On your settings to collect and I do not want to influence tastes, so I'm not going to write my personal rating index, but I mean as Phillippe, when others have other tastes, that´s fantastic !
However I would like to explain my methodology:
I managed to combine the linear scale of the statistical Roger dee Rüdiger through a simple mathematical model based on logarithms, where:
RI (rarity index)
n = number of units that I personally know (it can not have seen the other person).
log = logarithm
RI = log n
When using such a logarithm (base 10),
10,000 units if I know then RI = 4
when n = 1,000 then RI = 3,
when n = 100 then RI = 2
when n = 10 = 1 then RI
when n = 1 then RI = 0
when n = 0 the logarithm does not exist ... and indeed the car (to our knowledge) does not exist! .
This is just one example, you can change the decimal of a base of 5 or more so it fits what you want to collect.