Foucault said
This is really only true on the river, when hand values are static. On previous streets, it's also relevant how much equity your "value" and "bluff" hands each have vs potential calling hands. You should really experiment with something like GTORB, Piosolver, or CREV if this is something you're interested in getting right.  

Yeah, good caveat. Definitely true.

For example, if you're giving your opponent 2:1 by betting full pot, you need to be bluffing 33% of the time. You're also correct that if you're betting half the pot and giving villain 3:1, you need to be bluffing 25% of the time.

This can be really difficult to use in-game, but it does help to do it away from the table as often as possible, and it also helps to think about bet sizing in terms of how polarized your range is - the bigger you bet, the more of your bluff hands you can bluff with, because you're allowed to have a higher bluffing frequency, so when your range is polarized it benefits you to go bigger. If your range is more linear or merged, you want your sizing to be smaller, since you have fewer outright bluffs in the first place.

I'm looking at working on building some ranges in situations primarily where I would bluff and am looking for some help.  

First things first... Here's my starting point.

I have read that building these ranges are impacted by our bet sizing.  So as an example:

If there is 100 in the pot.  And we are looking to bet 50, regardless of bluff or value, we should have roughly 25% bluffs to 75% value as he's getting 3/1 on his money.  If we are to bet 100 into this same pot we should have roughly 33% bluffs to 67% value since he's getting 2/1 on his money.  Is this a good starting point?  Does this make sense?  Then from this point am I able to reverse engineer range construction?

I find I am doing some things well up to the point where I am river bluffing a hand that may or may not be way outside what should be a reasonable bluffing range.

Any input is appreciated.  Thanks in advance!