I remember reading about a study a long time ago about how things get popular. What they found out is that if you ask 20 or so randomly selected people to individually rate a set of songs without talking to each other, they will give remarkably consistent results. Individual people will have different tastes, but one randomly selected set of 20 people will have very similar tastes to another randomly selected set of 20 people because individual tastes average out. Each set will have about as many rock fans and pop fans and rap fans and classical fans and so on any other set. That doesn’t mean that the top picks were objectively any better than the other songs, but that they reflect the collective tastes of society.
However, if you select 20 people and let them talk to each other about the songs they’re rating, you’re going to get very inconsistent results, not just between different groups but within the same group on different days. The lowest rated songs were still unpopular, but songs that were mid-rated for the randomly selected people who can’t talk to each other could easily shoot to the top and vice versa. People influence each other’s tastes, even when they’re not trying, and this has a spiraling effect on popularity.
What this means is that winning a game contest is always at least as much a matter of luck as it is of creating the best entry. So don’t enter a game contest with the intention of winning, and don’t take it personally when you don’t win. Treat it less like a race and more like raffle.