@benny275 says he'd like to hear about design inspiration and how to turn a prototype into something publishers would want to produce.
Interesting question, and I think the answer is different from designer to designer, and even from game to game. Most people talk about designing "theme first" vs "mechanics first," that's a similar question, but yours is more broad I think.
I suspect that for me, the answer is that inspiration comes from any of 3 places... Real life processes that feel like a game to me, Existing games where I wonder "what if this part were different?", or a specific mechanism that I either haven't seen before or haven't seen used in a particular way. I suppose it's worth looking at some of my game designs and seeing if I can remember what inspired them:
Eminent Domain: Combination of "a card game version of Twilight Imperium" and a mechanism I'd already had in mind for tracking a civilization's (or character's) stats that was originally based on Queen's cube tower, later changed to deck building with some aspects inspired by Dominion.
Terra Prime: Grew out of a conversation about "what Roads and Boats in space" might be like, except the people conversing had never played Roads and Boats.
Seth Jaffee's Brain Freeze (free iPad download - check it out): What would happen if you had a simple/trivial decisions (such as Rock/Paper/Scissors) plus time pressure.
And some of my unpublished projects:
Alter Ego: Theme inspired by a game I saw that was a storytelling game trying to be a strategy game. Mechanics invented to fit the theme.
Exhibit: Trying to build a game around a core mechanism that is already a game. Like Zooloretto was built around the mechanism that is Coloretto for example. Exhibit is a game built around the main mechanism of Liar's Dice. I'll note that Eminent Domain is a game built around a main mechanism that ISN'T (but is inspired by) Dominion.
Dice Works: An attempt at a basic dice drafting game when I asked the question "what if players didn't have to take turns?"
How to turn a prototype into something a publisher would want to produce is a tough question. I think the answer is "play it, then refine it. Do that over and over again." When you get something solid (and it's good to begin with), then a publisher might want to produce it. I think it's worth keeping in mind the final components you expect to be in the game, and make the prototype as true-to-life as possible. I wouldn't spend any money on art for a prototype, but I do recommend putting some time into google image search, and think about the layout of your prototype cards and boards to make the game easy to play, so that fiddling with the prototype does not interfere with playing and evaluating the game itself.