It seems like a similar challenge as looking at the steam hardware survey stats, it's probably the best public source of information but it's too much of an ove

It seems like a similar challenge as looking at the steam hardware survey stats, it's probably the best public source of information but it's too much of an overview to make anything but the most general conclusions from it. Within that data there's going to be millions of different stories if you could divide it up a bit - what are common combinations of hardware, what are common combinations in different age ranges, what is used for playing games of different ages, what's common in different locations in the world. Valve could possibly tie hardware to software purchasing activity as well what is played.