I agree it's kind of crude but I think there are ways to make the data kind of make sense for our purposes.
Imagine a scale. Take your lieral percentage and conservative percentage and put them on each scale. That will tell you how liberal or conservative you are.
So if you are 90% liberal and 30% conservative, you're mostly liberal and vice versa.
If I were modeling this for a game I would take the conservative value and subtract the liberal value. At 0 you're an "independent". Negative values would be liberal and positive values conservative. If your net value is >67 or <-67 you're probably close to a "radical" conservative/liberal.