Randall Bolten Blog

At a dinner party last night, one of the guests posed a question: Imagine a roomful of people chosen at random. How many people need to be in the room for there to be at least a 50% probability that at least two of the people have the same birthday? The answer to this puzzle tells us a lot about how we process risk & uncertainty, and how we should think when we build financial models.

The Powerball payout has reached the jaw-dropping, mind-numbing (insert your favorite hyper-adjective here) level of $1.4 billion. I must speak: state lotteries are one of the most sinister and unfair ideas our elected officials have ever had. And simple, straightforward numerical thinking would have clarified everything. Let us count the reasons why I feel this way:

In an interview today National Public Radio reported on home construction data that the U.S. Commerce Department had been reporting erroneously for ten years. Not a sexy topic, but read on, since it addresses the whole subject of why we bother to collect data, and whether we need to spend more of our tax dollars in that effort.

Dishonesty can take many forms. Many are easy to spot, or at least will get caught out by a little research – or by the increasingly large number of journalists and organizations focused on fact-checking. Other forms are subtler. My “favorite” with respect to numerical information is cherry-picking – using a few carefully selected facts to support or refute an argument. The facts cited are usually correct, and the presenter might be well-intentioned, but the approach is intellectually dishonest. Let's consider some examples...

It’s ironic that two of the most popular forms of data visualization are named after delicious but fattening and unhealthy foods: the pie chart and its developmentally challenged sibling, the doughnut chart. To illustrate my point, let’s look at how we might present the distribution of scoring on the NBA’s current top team, the Golden State Warriors. We first see a pretty but non-nutritious doughnut chart, then consider some much better ways to look at that data.

Last time we used a GRAPH to appreciate the long-term average performance of the U.S. stock market. It’s an important message for U.S. workers trying to save up for retirement. In this post, we look at equally important issue – the risk and uncertainty around that average, and this time it’s a TABLE that works best.

A really good example of data visualization used properly is looking at the performance of the U.S. stock market over a very long period of time. Moreover, it’s an important subject that’s critical to almost everyone’s financial security. In this post, we look at an example of data visualization that can make an intimidating subject less terrifying.

Like the Deadly Sins of Quantation listed in my book, “Painting with Numbers,” the art/science of designing incentive compensation plans has its own Deadly Sins. These Deadly Sins of Incentive Compensation render these very expensive plans useless or even destructive to the enterprise. I have a few of my own, but I’d love to hear your “favorites.”

(I posted this blog exactly one year ago, and I post it again as a public service.) It’s that time of year again, when we have to figure out our best health insurance alternative in the face of wildly gyrating prices and complex competing plans. If you’re a finance professional, you’ll be doing your coworkers and your friends a huge favor if you can help them through this important and expensive decision. And data visualization plays a big part in this task.

In an “Investment Writing” post, Susan Weiner observes that the frontrunner presidential candidates in both parties are the ones that speak the simplest English. We can assess the complexity of writing or speaking with metrics like the Flesch-Kincaid Grade Level score. Should we have similar metrics for our quantation?