The whole thing is in flux, constantly evolving and changing. The bar keeps getting raised, such that a game people raved about and became a massive success just 6 months ago today would barely get any attention. That said, momentum counts. If you started following a game and got into it 12 months ago then despite it now looking a bit dated, the chances are you are still following it, still invested in the ending.
The fact is that GPUs keep getting bigger and faster, making higher quality artwork easier, or enabling better renders than ever. That means that the 'average' or expected quality of the visuals keeps going up. On top of that, because the capabilities keep increasing, so do the techniques.
However, this constant raising of the bar does not at all just apply to the visuals. We've come to expect more in a story and plot too than just yet another rip-off of Big Brother to introduce us to yet another game described as "You play a young man..blah blah... college/school... blah blah... move to a new home... landlady/mother... two daughters..."
In part that is about originality, of course. We've played that story, way too many times. If I want to play it again I can take another path on one I already know I enjoyed and that got completed. If I so much as see those words (young man, college, 2 daughters) in a description of a game, it is 50:50 chance that I walk away immediately without even giving it any further chance.
Certainly your game can 'get by' and be 'good enough' if the writing isn't great, the story isn't particularly unique, or the writing is Engrish, but good enough doesn't cut it. There are thousands of other games that are 'good enough'. If you want a game to actually be successful, be remarkable, then it has to exceed expectations and go straight past the 'good' into 'awesome'. It is not enough to be 'just as good' as the games they already follow and support, because they already follow and support a game that is that good, and they are already invested in it. Your game needs to be better by a big enough margin to compete for their time, make them abandon something else.
That's because nowhere in the world is there anyone sitting around saying "Wow, I have far too much money, I need to find something to spend it on". And the only thing even more limited than money, even more precious, is time.
Each and every time we, any of us, play a new game (or just one new to us), we change an important factor in measuring what is 'best' by changing our experience. Humans like novelty, something new and different. However, we also like more of the same, some sense of stability. And which of those conflicting desires takes precedence on any given point is largely down to our experience and mood. The games you have played influence your tastes in the games you want to play next.
I know. I know. You are sat there saying "Geez, I didn't want a lecture, I just wanted to get some stats on which games people play the most, or like the most"
But here's a secret. Statistics are bullshit.
Take a classic statistic, such as the one about the average family having 2.4 children (that's the UK average). NOBODY actually has 2.4 children. But it's way more deceptive than that. You may think "Well, that's obvious, and it just means that most families either have 2 children or 3 children". And immediately you are massively wrong. Most families have no children. That's because when your statistics include families with 6 and more children, families with quintuplets, etc. You need quite a few families with no children to bring the average down to 2.4 of the nonsensical, misinformative statistics.
What the best, or most popular game is based on prior downloads or views, doesn't tell you how many of those were by people who decided right afterward to drop that game. You can't even look at just downloads in the last week, because if we already downloaded it the week before that, why would we do it again? You can't even somehow mine the data to just look at views and downloads by new users, people who only found the site in the last week, because that then slants the stats towards first-timer friendly games, not the best or most popular, or the one's they'll think were good a month from now.
