In Stephen King's On Writing he discusses the idea of an "Ideal Reader." It is a tool to help avoid the trap of overthinking concerns about the reception of your work.
You can’t let the whole world into your story, but you can let in the ones that matter the most. And you should. Call that one person you write for Ideal Reader. He or she is going to be in your writing room all the time: in the flesh once you open the door and let the world back in to shine on the bubble of your dream, in spirit during the sometimes troubling and often exhilarating days of the first draft, when the door is closed. And you know what? You’ll find yourself bending the story even before Ideal Reader glimpses so much as the first sentence. I.R. will help you get outside yourself a little, to actually read your work in progress as an audience would while you’re still working. This is perhaps the best way of all to make sure you stick to story, a way of playing to the audience even while there’s no audience there and you’re totally in charge.
I think this concept has been more and more helpful to me as I've been making games. As games are becoming more varied, the previously etched-in-stone concept of a "good game" is vanishing. There are going to be more and more games that just aren't interesting to you.
I know that the games I make are not for a lot of people. This feeling became overwhelming for some time and I really felt like I wasn't doing the right kind of work. I spent over a month working on a game in the vein of survive-as-long-as-you-can twitch iPhone games. Not that there's anything wrong with those games, I really enjoy some of them, but they're just not fun for me to make. After that failure, I quickly resigned myself to making the games that I enjoyed making again. One approach I used to try and overcome the feeling that I wasn't doing the right work was working on games just for myself. However, with this approach I quickly forgot that people would ever even be playing my games. This quickly became a problem as issues of usability were forgotten, I lost concern for controls, I had no drive to set a release date, and concepts stopped being as clearly conveyed (since everything totally makes sense in my head).
However, once I started thinking about an ideal player it became a lot easier to make design decisions, feel good about my work, and keep in mind that a game is something for others to play. The fact that my games aren't for everyone became irrelevant: the ideal player cared. The people I use as my ideal player don't know they play this role, and I've only ever gotten very limited feedback from them. But every time I put something out there I know that if my ideal players enjoy the game I've done something right.