Not strictly a technical question. I'm looking at different statistical methods for calculating the popularity and value of a game. For anyone doing analysis or app development using Steam web API/Big Picture API, what are your thoughts on these different approaches?
Apps often have multiple genres and categories and it impacts other stats. For example, say I want to calculate the fraction of time I spend playing different genres, like Action games vs Adventure games. What if there is a game which has both Action and Adventure in its genres? Here are the two simplest approaches I am considering:
Say I want to figure out which game I should purchase. If I look at the games my friends own I can predict the dollars/hours based on the current prices and the playtime of my friends. However, maybe I have some friends who play Steam games 100 hours a week, and some who only play ten. If I calculate the value by raw playtime, the players who use Steam 100 hours a week can skew results. If I calculate the value by fractional playtime (aka "indexed"), the players who use Steam 10 hours a week can skew results.
Thanks in advance for any thoughts or feedback, and my apologies for formatting.
You are currently replying to another tree, would you like to make a top level comment? Reply
In order to reply to this thread,
you must login via Steam.
💗 Support SteamDB by donating
This site is not affiliated with Valve, Steam, or any of their partners. All copyrights reserved to their respective owners.