>I still think it's luck based
Actually I digress with this full statement (it's a bit of luck) and here's why.
If you make a cute game (keep in mind I'm saying cute because it's a popular theme, it doesn't need to be cute) and someone enjoys it, he will keep playing it. You will then make a screen pop asking if he liked it, then asking him to rate and review. If he likes the game, there is a good chance he is going to do so.
Then you add rewards for him to make friends in the game. And rewards to invite his friends to the game. If he enjoys the game, he will invite people he knows to play the game, in part for rewards in another for playing together. Not every player will add someone, but those more avid usually add like five or ten people each. Even better if it has connection with Facebook.
It does have a bit with luck, but if you make something solid that people enjoy people will share it. It's natural. And those who usually enjoy games of a theme (such as cute) usually constant look for other games, which gives even greater chance of appearance.
As someone who probably download a lot (and I mean a lot) of mobile trash, including those with little views, those games were usually 'just a single screen' or 'paint something'.
Here, one on my mobile phone: Fancy Pet. It's cute, true, but completely trash. Just a screen that you customize. And this abomination is made on Unity. Now checking, it's so bad they removed it from the store.
Okay, here is another example. Poddle Play. Trash go through a screen thing. It has less than 125 ratings. All is well in the world.
Now Monster Hotel. I got tired a bit, but it's a cute game with a bit more effort. It has almost 5k views.
Pockieland, a game with better graphs (but poor minigames) has 12k hits.
Marketing has a place, but you can build popularity by being good. I never downloaded an obscure app which was extremely fun, and look that I have epic Poodle Play on my phone.