Select the books you want.
Price: $395 $235 (you save $160)
All these books cover the same fundamental programming and data science topics. If you read one, you should be able to apply the concepts to other areas, whether that's other sports or a day job.
That's why for most people, I'd recommend most just buying whichever sport you like most.
But, because people have asked, here are the main differences/reasons you might pick up more than one:
Each sport has different ways of getting data via APIs. Whether directly from the league APIs (baseball, hockey) or via a third party Python library (basketball).
These are fairly specialized and we walk through each in depth in all the books, so if you wanted to get a good handle on getting data for multiple sports (or just want API practice), it might be worth getting more than one.
More generally, each book includes an appendix on places to get ready made datasets. For some sports (e.g. football and soccer) the datasets available are high quality and ready to download. For other sports (hockey and basketball) getting your own data via the API is probably the best way to go.
The hockey, basketball, and soccer versions include shot charts. Getting these to look right is a tad finniky, and these include instructions, a background image, and code examples for how to get it right.
While all books cover the same core modeling concepts (linear regression, logistic regression, fixed effects, Random Forests, cross validation etc) the examples are a bit different, especially in football, baseball vs the other shooting sports.
The end of chapter problems (and solutions!) are different in each book, so if you want extra practice, more sports might help.