Scan barcode
A review by sarahetc
How Not to Be Wrong: The Power of Mathematical Thinking by Jordan Ellenberg
3.0
I don't remember why I put this on the Big List, or when, but I'm glad it came up. And then I was even more glad when the introduction read a little like Bill Bryson and promised to make math accessible and answer the question, "When will I ever use this in real life?" Ellenberg can turn a phrase, for sure, and assured me I would need to know nothing but basic arithmetic to understand what he was going to do. Sign! Me! Up! And then I did a little dance, despite being tucked nicely into bed in my bookfort, with heating pad.
And that lasted through the first chapter. I hung in for a good long while because of clever illustrations and again, a great turn of phrase and the absolutely injudicious use of footnotes to make dorky jokes. But it turns out I also needed to remember geometry, which is fine. I'm okay at that. And then have an understanding of calculus, which is less fine because my understanding of it is limited to what's been dramatized by Neal Stephenson. So I powered for a few additional chapters and gained a learning or two, but hit a wall when it was time for the chapter on P-value.
The book fronts like a Bryson-esque explanation of math qua math but it's really a book on statistics and probability. Those things are cool, too, but I wish either he or his publishers or both had just been honest about it. I wanted the Whole Big Math Picture Written Especially for People Who Were Freakishly Good at the Verbal Parts of Standardized Tests. I ran into several chapters that all merged to take me back to Math 660, which I only took like 18 months ago, except not as friendly as my professor was. No mention of a quincunx or the Guiness factory. I already knew P-value. I never did know how to find alphas in a z-table and I still don't know. I know upper and lower control limits and special cause variation, but Ellenberg didn't really mention those. Mostly because you don't argue with standard deviation-- which, now that I think about it, he went out of his way not to talk about. Hey!
So, yeah. How Not to Be Wrong is not about how to build/grow a more mathematical mindset or see the flourishing logical beauty in the world around you. It's about how to make sure your efforts in statistics and probability support the case you're trying to make, whatever that might be. And, of course, the converse of that.
And that lasted through the first chapter. I hung in for a good long while because of clever illustrations and again, a great turn of phrase and the absolutely injudicious use of footnotes to make dorky jokes. But it turns out I also needed to remember geometry, which is fine. I'm okay at that. And then have an understanding of calculus, which is less fine because my understanding of it is limited to what's been dramatized by Neal Stephenson. So I powered for a few additional chapters and gained a learning or two, but hit a wall when it was time for the chapter on P-value.
The book fronts like a Bryson-esque explanation of math qua math but it's really a book on statistics and probability. Those things are cool, too, but I wish either he or his publishers or both had just been honest about it. I wanted the Whole Big Math Picture Written Especially for People Who Were Freakishly Good at the Verbal Parts of Standardized Tests. I ran into several chapters that all merged to take me back to Math 660, which I only took like 18 months ago, except not as friendly as my professor was. No mention of a quincunx or the Guiness factory. I already knew P-value. I never did know how to find alphas in a z-table and I still don't know. I know upper and lower control limits and special cause variation, but Ellenberg didn't really mention those. Mostly because you don't argue with standard deviation-- which, now that I think about it, he went out of his way not to talk about. Hey!
So, yeah. How Not to Be Wrong is not about how to build/grow a more mathematical mindset or see the flourishing logical beauty in the world around you. It's about how to make sure your efforts in statistics and probability support the case you're trying to make, whatever that might be. And, of course, the converse of that.