To make things even more confusing, the games I truely enjoy are often games no one else likes, and I mean the type that get reviews of 4-5/10. So it's kind of a bad idea for me to use them.
That's even better because that means there is room for improvement. As an exercise you could find out what makes that game suck, then clone the good parts and improve the others, replacing or adding new elements. For example, I like a crappy game called Prision Tycoon. The idea behind this game is interesting and there are plenty of opportunities to make it entertaining, but it seems the developers made several bad decision during execution and also release the game in a broken/unoptimized state. So it can be greatly improved and many people play this broken game because there isn't anything similar, so there is also a market.
+1 to all of the above, escept, there's Prison Architect now, so the market for THAT specific gameplay isn't as easy to grab. But you're likely to find other niches. As a matter of fact, as an indie, your one way to success (and not be compared to AAA publishers) is to find a niche you like.
Pick one of the games you like, which has a very bad review.
Look up the reviews (the ones that bother to be 'grey' and tell you what sucks/rocks).
Focus on the potential you see (why is it fun for you?).
Check the 'downsides' you see, and the 'downsides' the reviews see. Can you assess them in one way or another?
Build a game with the above ingredient, and if you're lucky, you could be the next Mojang (of course, the odds of that occurring are very low, but indie sleeper hits are not unheard of).
