Calling Bullshit: The Art of Skepticism in a Data-Driven World

Finally got around to reading "Calling Bullshit," and it's basically a survival guide for the 21st century. Bergstrom and West do a brilliant job of deconstructing how data can be used to mislead us without technically lying. I loved the deep dive into p-hacking and the "black box" of algorithms; they make complex statistical concepts feel surprisingly intuitive for a layperson. The writing is sharp and frequently funny, keeping the tone light even when the subject matter feels overwhelming. It’s not just about debunking fake news, but about understanding the structural reasons why misinformation spreads so fast. In my experience, the section on data visualization alone is worth the price of admission. It truly feels like essential reading for anyone who wants to navigate the internet without losing their mind.

Ever wonder why it takes five minutes to make up a lie but five hours to debunk it? This book introduces Brandolini’s Law, and truth is, that concept alone changed how I view my social media feed. The authors move past old-school rhetoric to tackle the "new-school" bullshit that hides behind fancy charts and scientific jargon. It’s written with a fantastic sense of humor that makes even the technical sections on selection bias and correlation versus causation fly by. I particularly appreciated the distinction they make between spotting bullshit for yourself and the social performance of "calling" it out. They warn us not to become "well-actually" guys, which is a trap many of us fall into. This is a powerful tool for informed citizens.

Picked this up hoping for a way to win arguments online, but I walked away with something much more valuable: a mirror. The authors brilliantly point out that the person most likely to bull-shit you is actually yourself, especially when a factoid confirms your existing prejudices. Their advice to pause when something feels "too good or too bad to be true" is simple yet profoundly difficult to implement in the heat of the moment. The writing style is engaging, mixing academic rigor with casual anecdotes that make the material feel incredibly accessible. I loved the section on "Fermi estimation" as a way to quickly check the logic of a claim. It’s a fantastic guide for anyone who wants to be a more discerning consumer of information in a data-drenched world.

Wow. This is a must-read for anyone who feels exhausted by the constant "tsunami of crap" we face every day. Bergstrom and West are total masters at taking technical, intimidating material and making it feel like a conversation at a bar. I especially liked the examples involving the Covid epidemic and how science being done in real-time creates the perfect breeding ground for misinformation. The book is accurate, well-researched, and manages to be funny without being flippant about the very real dangers of bad data. It teaches you how to look for the "mendacious element" in a claim without needing a PhD in mathematics. This is truly the art of skepticism refined for a world where everyone has a megaphone.

As someone who works in data, I found the breakdown of "mathiness" to be the most satisfying part of this entire book. The authors explain how people use complex-looking formulas to intimidate others into silence, which is a tactic we see everywhere in business today. Frankly, the way they explain the "Trust Equation" as a nonsensical formula was both hilarious and eye-opening. The book is filled with great examples that use plain English to explain why certain graphs are designed to deceive our eyes. My only real gripe is that it feels very US-centric, often ignoring international contexts or cases that might be more familiar to global readers. Still, it provides an amazing guide to critical thinking that is desperately needed in our current media landscape.

The chapter on AI and machine learning was the absolute highlight of this book for me. Most people treat algorithms like magic oracles, but the authors strip away that mystique to show how "garbage in" always leads to "garbage out." It’s a sobering look at how automated systems can perpetuate bias while maintaining an air of objective authority. I also found the discussion of p-hacking in scientific journals to be incredibly illuminating, if a bit disheartening for my faith in academia. While the book can get a little repetitive with its warnings about confirmation bias, the examples are current and relevant. Not gonna lie, I’ll never look at a "wine is good for your heart" headline the same way again. It’s a very readable and necessary reality check.

Personally, I think every high school student should be required to read the section on misleading graphics. We are constantly bombarded with "ducks" and "glass slippers" that warp our perception of reality, and this book gives you the goggles to see through them. To be fair, some of the statistical deep dives went a little over my head, but the authors always brought it back to a practical takeaway. I appreciated that they didn't just focus on "fake news" but also on well-intended scientists who accidentally misinterpret their own data. It makes the whole problem of bullshit feel more like a collective challenge than a battle against villains. My only complaint is that the section on actually calling out the bullshit was a bit short. I wanted more scripts for those awkward family dinners!

Frankly, the authors deserve a lot of credit for making a book about statistics actually feel fun to read. They tackle everything from the Wakefield vaccine debacle to the Prosecutor's Fallacy with clarity and a healthy dose of wit. I did find the book a bit long-winded in the middle sections, but the "Top Tips" summary at the end is pure gold for quick reference. It’s a wonderful resource for learning how to spot selection bias and the dangers of unrepresentative samples in the media. While it doesn't solve the problem of people who want to believe lies, it empowers the rest of us to stay grounded in reality. It’s a solid, well-written guide that turns complex data into something you can actually use.

Was this book actually bull-shitting me? I found it incredibly dense, with a relentless barrage of examples that made the reading experience feel heavier than it needed to be. While the information on p-values and statistical methods is certainly accurate, it felt a bit like a textbook disguised as a popular science read. To be fair, the authors are clearly experts, but the "negative lilt" of the prose wore me down after a few chapters. I would have preferred more interactive exercises where I could practice calling bullshit on headlines before they gave away the answer. It's a solid refresher for anyone who has studied stats, but I didn't feel like I was truly being trained in a new skill. It’s useful, just a bit of a slog to get through.

Look, the title is "badass" and the marketing is great, but there just wasn't enough new information here to justify three hundred pages. Much of what is covered—correlation isn't causation, checking sources, etc.—is ground that has been trodden a thousand times before in better books. The truth is, if you’ve watched a couple of TED Talks on data literacy, you’ve already got the gist of what Bergstrom and West are selling. I also found the tone to be a bit condescending at times, as if the authors assume the reader is constantly being duped by the simplest tricks. It’s fine for a total beginner, but for anyone with a modicum of skepticism, it’s a lot of fluff around a few simple ideas. I’d recommend just reading the last chapter and moving on.