0 votes
by (1.3k points)

Let's say I wanted to assign an attraction score to a character.

If I had a list of notable features:

hair	= 75
eyes	= 70
lips	= 85
face	= 80
legs	= 75
fitness	= 65
wealth	= 30

If I took those values (which range from 1 to 100), and averaged them across the fields, would that produce an OK score? I think not. I think certain values should be scaled to reflect their importance. Or possibly different attributes might be more important dependent on the viewer.

Does anyone have views on this? Anyone worked on similar systems? I haven't. I haven't really brushed up on statistics either, for that matter. Or distribution curves. But... has someone here does such research? You wanna talk about it? Potential formulas. Concepts?

1 Answer

0 votes
by (159k points)
selected by
 
Best answer

I believe what you want is a weight function (generally called weighting) which is commonly used when analysing the results of a survey. I am unsure exactly how you would apply it to your particular use-case. 

by (1.3k points)
Yeah, I went with the weighting approach. It worked out pretty well, I think. So, I'll mark this one as answered too!
...