Plenty of people play a lottery, even knowing they can expect a loss (my mum is a maths teacher, yet still enjoys playing the lottery on occassion). Expectation is not always the best metric to use when thinking about games of chance (for example, see the St. Petersburg Paradox: http://en.wikipedia.org/wiki/St._Petersburg_paradox). I'm perhaps playing devil's advocate a bit here, but consider a person who earns a low wage and has no possibility of ever increasing this wage except by winning a low probability game of chance that costs a very small nominal amount to play and entails an expected loss with each play (ie a lottery). If the cost of play is a small proportion of weekly earnings, it seems reasonable (to me at least) that from a utility perspective, the person would choose to play the lottery. The key point is that if the payout is large enough, then the person will stop playing if they win (hopefully), so the concept of expectation (ie, number of plays going to infinity) is not necessarily the right metric.

Actually, "other countries" could easily solve the patent problem in the US by banding together and refusing to adhere to the Agreement on Trade-Related Aspects of Intellectual Property Rights (see http://en.wikipedia.org/wiki/Agreement_on_Trade-Related_Aspects_of_Intellectual_Property_Rights). The US would be forced to overhaul the system, or else watch half their economy relocate overseas.

From the article, it looks like the "Other" category accounts for almost half the downloads. I'm a bit dubious about any conclusions drawn from a dataset with this many unidentified points.