Keith Law

Of course I thought this book was fantastic. If you always wondered about the statistics behind baseball, this is an excellent, easy to follow primer. Law first explains the statistics that have been used since the early days of baseball and basically shows one by one why they suck. Then, he addresses the new statistics we use today and why they are at least better and actually tell us something. Finally, the last third of the book addresses the MOUNTAINS of data that is being provided by StatCast (which apparently tracks every single movement on the field) and how interestingly, in some ways the analytical approach that used to be a competitive advantage has now been adopted by every team, and it's up to the individual teams to dig deep to find better ways to use the masses of data. It used to be you could be a kid with a spreadsheet to be part of a baseball analytics team; now you need a PhD.

Great read for lovers of the game.… (más)

This isn't just a useful book for baseball or sports analytics. Law's approach to making sense of data in the applied field of baseball points out the many flaws and pitfalls of any analytical pursuit. Cases of "managing to the stat" or simply relying on counting stats to the exclusion of useful context are profoundly atavistic, but Law keeps pointing out how these lay approaches persevere and thrive despite more coherent methods.

I generally don't get much out of sports analytics books--they tend to be introductory primers to many of the concepts analytics nerds are already deeply familiar with. Where Smart Baseball excels is that Law does more just than show the advantage of a contextual number over a mere counting stat. Instead, he spends a great deal of a time exploring how the fallacies that went into creating myths like the relevancy of a Save or Pitcher Win were inculcated into fan and even subject matter expert's understanding of the game. It's this constant refrain about the emergence of error in commonsense stats that makes the book an extremely useful polemic against the use of shallow approaches to big data and broader analytics.… (más)

Fascinating, eye-opening, infuriating book on baseball statistics. This book is Moneyball on steroids, and while it does take some of the rough edges off of sabermetrics in the end it really doesn't make the more esoteric of them more palatable. Maybe I'm just too old school (or just too old), and while some of the new stats discussed are useful indeed, some are just BS, as is most of the discussion in the book on the HOF voting and player career evaluation solely on statistics. Still, a useful read.… (más)

A good book. I've been reading Keith Law for a long time, now; he rarely disappoints.

There are two ways to view this book:
* It's a baseball book, discussing the cognitive biases that often shape decisions made by teams.
* It's a book about cognitive biases that uses baseball examples.
The author claims both reads are legitimate in the first few paragraphs.

I can imagine this as a college textbook. Law seems to think that would be an economics or MBA course, but I could see it used in a psych, sociology, or even philosophy classroom. But its sometimes baseball-analytical (aka sabermetric) background would probably confuse some of the students, even though Law's quite good at explaining those things.

From a baseball fan's perspective the author's non-baseball examples could well be considered a distraction, though I found them interesting--they certainly help illustrate his main points. And because Law wrote the chapters with the intention that each stand alone, there's some repetition that could be annoying but is pretty harmless.

==========

One of the non-baseball discussions really caught my eye. Several pages in chapter 4 discuss vaccine misinformation, and the ways in which it spreads. There's also a bit of discussion about the difficulty of combatting conspiracy theories.

This in a book written in 2019 and published about the time Covid hit.… (más)