Over the years, I've tried more or less everything I can think of.
Sometimes it's just to see how something works, sometimes it's to satisfy my own curiosity.
We've done Pay What You Want products, given things away, charged up front, a few crowd-funds etc.
A potentially controversial stance is that I've increasingly opted to charge for beta versions of games: From a dollar or two to 5 bucks or so.
"What is this outrage? Why should I pay to test your broken-ass game?" I hear you say.
Sure. I hear you back.
I have mostly settled on this for a couple of reasons:
First, to me a beta version is something substantial enough that it could be a game.
If it's a full game with the sort of content you'd get from the competition but lacking testing, I don't feel bad charging a couple of bucks for it.
You know what you are getting (it says beta right there!) and you can decide based on previous products if it's worth the risk.
If the draft is so rough I am not sure it's even functional, it never sees the light of day outside my personal circle of critical eyes.
Second, I find that the level and quality of feedback is simply better if I charge.
Pay what you want or freebies are great but it seems people also don't feel invested in it.
If people had to put in their paypal account, there's more of a sense of investment (even if its just the price of a cup of coffee).
When I've charged more than "coffee money" for an early version, I try to do a discount code or something later on, so you can view the early version as a buy-in to the full one.
That also has the advantage that if something falls through, you still have the initial version.
I'm sure we'll continue different options and ideas in the future, but ultimately that's why I prefer charging for the beta version of a game.