First, I am not a mathematician. So, have patience with me please. I want to be able to assign a number (1-10) to n-amount of people, based on a given action. If that person does less of that action, they receive 1. If they do more (or the most), they receive a 10. And everything in-between for less extremes.
An example of this would be scoring people who watch tv. Someone who watches 40 hrs of tv per week should receive a 10 compared to everyone else who is only watching 20 hrs or 10 hrs.
What is a formula for this? I am trying to score people based on their actions COMPARED to other people doing that same action. Thanks for your time.
It sounds like you want feature scaling.
Min-max normalization (described in the article above) gives you scores between 0 and 1, which you can multiply by 10 if you want. One thing that might be undesirable about this approach is that if one person watches a million hours of TV, everyone else's scores will be low. An alternative approach that guarantees that the scores are evenly distributed is to arrange the people in increasing order of your metric and assign a score based on where they fall in the sorted list. This amounts to computing percentiles.