In preparation for my big data class I do various data analysis exercises myself. I like to write Flash games, not because they earn money (less than iPhone actually) but because they reach more people faster, and I like to see when my games are played. However, releasing a game is somewhat tricky, you have to choose a day when to open your game to the public. I know that for iTunes this moment is critical, so how about releasing a game on a major gaming site? I scraped all game titles of a major gaming site together with release day and total number of hits. The first thing I was curious about was the holiday effect. iTunes is swamped with new games around christmas, how is that with flash games?
What you see is the average number of game releases per month averaged for the last seven years. I scraped the data without the December for 2013, so the last datapoint has more errors. However, December is not a preferred month whatsoever. Okay, so how about game releases per day of the week?
Interestingly we see a high around Thursday and a dip at the weekend. I always assumed that players have more time on the weekend, but I also know that people with office jobs play quite a bit during office hours … What is puzzling now is the question about the best day? We see that most releases happen Thursday, which means the competition is highest on Thursday, which suggests that one should submit Friday to increase visibility. On the other hand, maybe there is a reason for a Thursday release? The next plot shows you the average number of hits a game achieves per day:
This plot tells us that even though competition is highest on Thursdays, games released on Thursdays do much better on average. This still doesn’t resolve the “which day to release” question, but it makes a good suggestion.
Since we are at it, I wondered about the progression of this site, so I plotted average number of hits per year per game.
This is interesting, since it shows that the number of hits declines. How about the total number of games released per month:
The plot shows that fewer games are released per month. While the hits per game per month data could be misleading since newer games did not have a chance to collect the hits they ultimately could get, which makes that data questionable, we see that the number of game releases per month actually declines. Is flash gaming dying?