Why?
Perhaps today, you can’t recall your old statistics courses
And glaze your eyes when prompting Claude and GPT for sources
Perhaps you are professional, immersed in Zoom-ish meetings
And eager for diversions from the anodyne proceedings
Perhaps you are a student now with corporate aspirations
Pursuing engineering and its sundry permutations
But all of us are longing, for to learn, to teach, to grow
To grapple with ideas worth the time and pain to know
So this, alas, is why I ask for minutes of your time
To follow these digressions into quantitative rhyme
And why, my silly students, should you follow up with me?
’Cause I have a framed diploma with the letters P, H, D?
Credentials? Overrated! My diplomas sit in boxes!
I’m just a guy who ponders math in jeans and wooly socks-es
We forget what statistics professors convey,
But remember the top-hat of little Cat A!
So even if decades from now, as you’re aging
You’ll still find such humor a wee bit engaging
When commonly-offered numeric delusions
Emerge to deliver misleading conclusions
The mind will return to some couplets and wit
Which might redirect from the bull and the shit
And so, in the psyches of subsequent leaders
Might just live the whimsy of poetry readers!
So what is this book about? (Apart from the silliness of Seuss-esque rhyme?)
The overwhelming majority of quantitative analysis is wasteful. The broad heading of “data science” or “analytics” or “business intelligence” is anything but scientific or intelligent. The immaculately-coiffed, well-paid, platinum-status consultant has more in common with an intestinal parasite fed by the largess of its host than we would care to admit.
We’re inundated with data, both the obvious presentations in dashboards and spreadsheets and the subtle ones that emerge in meetings couched in persuasion and commentary. We’re consumers of a product for which we don’t shop and with which we have less operational skill than our household appliances. But we’re still seeing metaphorical boxes arrive on our doorstep as if we know what to do with their contents. Worse, most of the time, when we use those “data-driven” products inaptly, we are entirely unaware. At least when I use my remote control incorrectly and inadvertently cue up some streaming service’s advertised programming I can tell that the screen is not airing the football game for which I had hoped!
In other words, while I’m thrilled so many folks have deigned to pay me for my quantitative insights, thinking clearly about how conclusions emerge from data and how decisions are generated from those conclusions means never needing to hear from my snarky, pretentious self ever again. (How’s that for a pitch?)
Sure, every business is different and every decision is a special, unique snowflake. But from little leagues to the major leagues, the fundamentals do not change. Quantitative reasoning has its own set of principles. Cold outreach, cross-selling, and click-through rates are all flavors of conversion funnels, and as a result, we make similar errors of reasoning in each case.
Corporate urgency, ambitions, incentives, and plain-old hubris challenge our best attempts at sound experimental design. Often, we measure the right metric at the wrong time or apply the right math to the wrong problem. We fixate on errors that are easily measured, but overlook matters that elude such easy assessment, but generate astronomical, invisible costs.
We fixate on the latest tools and techniques for modeling without pausing to determine what it is our model ought to describe or predict, or, heaven forbid, whether the decisions it will inspire are desirable in the first place.
For the math-lovers, there will be some digressions into the land of statistical significance and distributions with fat-tails. But these are means of illustrating ideas that should never require more quantitative training than human curiosity. Most of us endured statistics courses grudgingly, with lamentations of impending exams. I promise no such hardships and ideally, considerably more entertainment value.
Those who immerse themselves in data seek truth. In this, they share an intellectual heritage that traces back to Greek philosophers and enlightenment thinkers who demanded evidence for their assertions. Unlike modern professionals, the guys with the togas knew their job to be truth-seeking and saw their philosophizing as a craft to be refined. Now, we’re probably more concerned with saving face than seeking truth - and for good reason. The latter is hard. We will discuss the modern sophistry of corporate fallacies. These are equal parts humor and horror. And all wit and whimsy aside, the alchemy by which data become decisions is no laughing matter, it is at the heart of the human experience.
If nothing else, having read this book, even if you aren’t ready to crack open a development environment and train machine learning models, you will be far better equipped to be clear-eyed consumers, producers, and interpreters of quantitative findings. Your next email with two charts attached and a link to a dashboard you’ve never seen will be an opportunity to consider incentive, assess the assumptions implied, and heaven for forbid, formulate your own opinions and insights. I promise to deliver more Seuss than snark.