Leaked data from Valve's API was available for a very short amount of time earlier this week, offering developers a clear look at the number of games Steam has actually sold before the leak was properly closed.
While it's likely the last accurate snapshot of sales numbers devs will get for a while, Ars Technica compiled a list of the 1,000 best-selling games on Steam as of this week using a new data derivation method written by Tyler Glaiel.
The data method revolves around the percentage of players who have completed developer-defined achievements associated with many games on the service, and because this data is derived from Steam's API for each game, it's more accurate than the old Steam Spy estimation method.
However, developer of Steam Spy Sergey Galyonkin was able to integrate the data method written by Glaiel into the machine learning algorithm used for Steam Spy, where the data was momentarily displayed on individual game pages earlier in the week.
The result is a massive CSV file containing sales data for every game on Steam that offers its users achievements. The data may become obsolete very quickly, but it's important for devs who want to better understand how the PC games market works.