# Calculating score

I have a number of values, grouped by some category. For every category, I've calculated average and standard deviation. Now I want to calculate a score (say 1 to 1000) for every category.

I don't know what the maximum average nor maximum standard deviation is (data set will keep growing), but if necessary for this calculation, I can set a maximum for both (again, I'm assuming that might be necessary). But, I would rather use some sort of logarithmic growth formula, so I wouldn't need to worry as much about maximums.

What are the different options/formulas I can use to do this?

-
What does the individual values indicate (lenght, height, grades, etc.) and what would your score mean? (or how will it be used?) - Such measures are sometimes referred to as Indexes. –  Emmad Kareem Feb 23 '12 at 22:24
individual values are media views aggregated per channel. The score is popularity taking consistency into the account (with standard deviation). –  Dmitry Grinberg Feb 23 '12 at 22:28
I guess a reasonable score would be the percentage of viewers of each channel compared to total number of viewers. –  Emmad Kareem Feb 23 '12 at 22:37
You may also want to move the question to: stats.stackexchange.com/?as=1 –  Emmad Kareem Feb 23 '12 at 22:41
New channels will be added and they can't skew other scores. only media view changes within a channel can influence the score. –  Dmitry Grinberg Feb 23 '12 at 22:48