There's an article by Nassim Nicholas Taleb recently published in Edge that is probably going to be required reading in any advanced course on analysis I teach. The last ten or twenty years has seen major developments in statistical tools (mostly as a result of more powerful computers). In some cases they can really improve our understanding of systems. In many other cases they can make us think we understand systems when we don't. This article explains where that transition is most likely to occur. It's a good time to read it, now that various financial institutions have destroyed more wealth in the past year than they have ever created. Did you know that? You might want to read this article.