How do We Assess Our Beliefs About Apple?

I write about Apple. I have opinions about how Apple is doing as a company. How do I know that my judgments bear any resemblance to reality?

Recently, I was reminded of the required thinking process in a book that I’ve been reading in the evenings. The Big Picture. It’s by Dr. Sean Carroll, a theoretical physicist at the California Institute of Technology. In Chapter 9, author Carroll is preparing to talk about how scientists know what they know, and how reliable their knowledge is about certain things.

Analyzing Apple or any company is, essentially, a real science. Via Shutterstock.

Stay with me here. There won’t be any math.

Author Carroll starts with an introduction to the Rev. Thomas Bayes (1702-1761) who was an English Presbyterian minister. Oh, and quite a good mathematician. We’re introduced to the methodology he worked out for “the best way of moving toward reliability in our understanding.”

On the same page, author Carroll more carefully defines the scientist’s argument that while they may not know everything, they know a lot of things. Here’s the quote that caught my eye as he answers the question about how we know the reliability of our understanding:

To even ask such a question is to admit that our knowledge, at least in part, is not perfectly reliable. This admission is the first step on the road to wisdom. The second step on that road is to understand that, while nothing is perfectly reliable, our beliefs aren’t all equally unreliable either. Some are more solid than others. A nice way of keeping track of our various degrees of belief, and updating them when new information comes our way, was the contribution for which Bayes is remembered today.

Degrees of Belief

Without repeating his Chapter 9, Carroll goes on to explain degrees of belief. Statisticians call theses credences. For example, if I told you that a man on a bicycle just rode past my house, knowing my location (Colorado) and your own world experience, you’d place a high credence on my casual remark.

However, if I told you that a headless man just rode by my house on a horse, you’d place a low credence on that fact. Later, I explain that a movie studio was filming a movie, the Legend of Sleepy Hollow, in my neighborhood. Suddenly, with this new information, your credence, your estimation of the validity of my remark, goes way up.

Thomas Bayes formalized this process in a way that lends itself to statistical assessment of validity. So, when physicists at the Large Hadron Collider at CERN reported in 2012 that they discovered the Higgs Boson, it was a statistical result, based on very high standard of confidence. This is a process that, without prior training or experience, is lost on most casual readers and authors.

Page 2: How does all this apply to Apple?

Back to Apple

In the latter months of 2016, an uncomfortable feeling was emerging from the community of Apple observers. Only one new Mac was announced in 2016, the MacBook Pro with Touch Bar. By the end of 2016, it became clear that was all we’d get from Apple, and the future of the Mac was called into question.

Then, on January 31, Apple had its Q1 2017 Earnings Report. Not only were Mac sales up year-over-year (just slightly) but Mac revenue was stellar. Apple must have sold a lot of (very expensive) MacBook Pros in Q1 to make up for what looked to be sagging sales in the previous three quarters.

Apple stores in my neighborhood were jammed all during the holidays.

Not only were Mac unit sales and revenue comforting, but the rest of the company seemed to be firing on all cylinders. I promised no equations, but I’ll sneak one in.

f{Macs, iPhones, iPads, Apple Watch, Apple TV, Services} => Revenue.

In other words, the executive focus, a function of all areas of the business produced the revenue and earnings that were reported. This is objective reality that has to be folded into the credence of the claim: “Apple is doing well as a company.”

I see many cases where authors seem to either ignore certains kinds of data, even performance data, or put the emphasis on the wrong elements of Apple’s business. Some even start with the conclusion, (Apple is doomed, Tim Cook is awful, etc), then pick and chose particular facts to support the proposition. That may be entertaining, but it isn’t analysis.

Evaluating Apple’s Black Box

I recognized that Apple’s Q1 2017 quarter was a holiday quarter, and customers tend to be expansive around that time of year. And yet. Apple’s performance was a record breaking Q1. Mac sales were up year- over-year. iPhone sales set a record. Services set a record and are on the upswing.

So when I look at Apple as a whole, and I update my thinking about what I know about Apple, I try to fold in data that’s objective and relevant (Apple customer enthusiasm => sales) and lower the credence of some other data (Apple doesn’t care about such-and-such.)

Apple is like a black box that has an input (supply chain), complex innards, and outputs (products). Most of use can’t really see inside the the black box to see the inner workings. Often, we draw conclusions about what must be going on inside to account for the output, but that’s not always reliable.

In summary, Apple is objective reality. Observing Apple is like observing the workings of the universe. We form theories. We try to do some feeble experiments on Apple by quizzing its executives at the Earnings Reports, and we formulate ideas about how Apple is functioning. How well we do that depends on how well we update our basic theories based on new information and the tools we use to assess credences of what we learn.

I’ve always tried  to do that in my own writing. Dr. Carroll’s beautiful explanation of Bayesian logic and credences has inspired me to work even harder on that kind of analysis.

Share this post