A propos the link this morning about lying about infant mortality (but also a propos everything else) we urge all readers and all Americans to read Darrell Huff's classic How to Lie with Statistics.
You do not need to be a math whiz to understand it, but if you don't know this basic stuff, you will be easily duped. Duped expecially by the MSM, which is not only biased in data presentation but is also widely ignorant about the most basic statistics.
In the link we mentioned above, the report was making an "apples and oranges error", known as a "Category Error." It's the same error involved in the report that showed the world's best cancer center, Sloan-Kettering Memorial in NYC, as having the worst mortality rates of American hospitals. Data can be technically accurate, yet meaningless. Similarly, the most talented docs often have the lowest success rates because they take on the toughest cases.
Money manager acquaintances have told me how they make their numbers look good: they select their best time frame to present, and they decide whether to include or exclude new money added during that time frame. That is a fallacy known as selection bias. A good example at the bottom here: how different would the impression be if you just charted June to November?
Speaking of money managers, I have also enjoyed Jones' How to Lie with Charts and Graphs. Fun to peruse.