Purchasing Baseball Hacks motivated me to refresh my horrifically stale understanding of basic statistics and probability. The last time I seriously engaged with stats is over 20 years ago for a sophomore course, and oddly I've never been forced to reconnect. Casting about for some additional material, I landed on two works authored by Jim Albert, a statistician at Bowling Green State University.
His Teaching Statistics Using Baseball is a thin work which is best used as a supplement to a traditional stats course. For example, there's a couple of places where the book assumes knowledge of MINITAB, a popular entry level stats package. Curve Ball: Baseball, Statistics, and the Role of Chance in the Game, written with Jay Bennett, is a more complete book, with few assumptions of the reader. Curve Ball clearly uses a lot of basic material from Teaching Statistics, so they're somewhat redundant. If I had to only recommend one it would be Curve Ball.
However, both books are good at highlighting the role of chance and working through concrete examples that separate performance, which is something observed, and ability, which is something claimed to be innate. Their best example is the streakiness of batting in players. One can often see negative and positive streaks in a player's hitting performance, but is that streakiness part of the player's inherent ability? In situations involving chance, some observed performance is simply attributable to randomness. Curve Ball and Teaching Statistics both delve into how to model and reason about these issues with concrete examples that are interesting to the average sports fan.
Also, the examples are easily translatable to hacking in Python, especially since Retrosheet makes a lot of baseball data available. Fair warning though. Their game event data files reflect the inherent messiness of scoring baseball, so parsing them isn't trivial, but that's a fun challenge unto itself.