An Experiment
Steam Charts is an experiment with Steam's player numbers. Player count data has been collected from Steam since July 2012, and a frontend was first put together in December 2012. The idea behind this is to see the long-term trends of game popularity on Steam.
You can already see some notable trends with only 6 months of data. For example, Dota 2 has the strongest player base ever seen on Steam, and is still growing strongly as of this writing. Counter-Strike continues to be strong, but CS:S seems to be losing some of its player base to CS:GO.
To me this sort of data seems valuable, especially as time goes on. It would be great to see more of this sort of data available in the rest of the games industry. With this data, there are some interesting things to consider in the future, such as adding news event markers, or comparing more than one game on the same line graph...say CS:S vs CS:GO.
This being a bit of an experiment with data, I'd love to hear your feedback. Also, as you interpret the data around the site, if anything is unclear at all, or if there are any technical issues, please let me know. You can do this by commenting on the Trello board or via Twitter.
A few things to note: This data includes all games available on Steam, not just the top 100 from Steam's stats page. Sometime I'll have to write a more technical post about how the data is collected. The main limitation is that the number of players for each game is collected once an hour, which means during any given hour there could have been many more players than when the data was collected, but over time (and in aggregate) I would say the margin of error here is insignificant.