Analytics Lessons From The Ice Rink
The Need For Solid Foundations And Real-Time Assessment Strategies
Spend a bit of time at a public ice skating rink and you begin to notice — a lot of people on the ice are novices. They stand… errr, glide … ummm, stumble along in stark contrast to the handful of skaters who have been doing this for years. If you let it, say if you often write analytic analogies, this poses an interesting analytic thought experiment.
It doesn’t take very long to differentiate those who are avid skaters from those trying their luck for the first time. However, it does take time. How difficult would it be to try to make that determination from a snapshot?
Most analytic exercises actually take this form. For as powerful as time series is, most businesses have neither the volume of data, nor luxury of time to wait for time series. New customers need to be assessed on arrival. One time users may never create meaningful time series. And more and more solutions seek to act in real-time.
So what if you needed to assess experienced skaters from novices with a simple snapshot? Could you?
You could, easily enough, look for the polar bears and penguins. But trying to move beyond that, gets a little difficult.
One interesting option would involve assessing the skaters foundation. Specifically, I am referring to the geometry of the skaters blades. Novice skaters spend a tremendous amount of time on unstable foundations, blades at varied and curious angles. But significantly, not all their time. So any given snapshot may show an otherwise unstable skater in a very stable position. Even the best machine learning algorithm will occasionally be fooled by this.
Let’s Talk About Foundations A Bit More
Solid and stable foundations are the key distinguishing between experts and novices in many skills. Along with ice skating, we can easily include analytics. While balance, posture, and control of skate blades are critical to expert skating, analytics has its own critical components to its foundation — data. Analysts must have proper structure and control to their data. One look at the data structures employed by any analyst is a key give away to their analytic prowess.
And What About Angles?
Great skaters employ great angles to achieve speed and force. Great analysts employ great angles with the same aim. Any skater can look balanced and solid with a strong foundation. But to be exceptional, to be nimble, to be fast — an analyst must understand the right angles to apply.
Examples both abound and demand more detail than this simple article can afford. One simple idea includes the technique of reversing the hypothesis. It is sometimes easier to disprove something than to prove it. Sometimes an analyst needs to change the level they attack a problem. Analyzing profit may not be necessary, if only a difference in overall sales would likely change the overall profit level… or maybe one should analyze the costs instead.
Finally, analytics has a complexity that can make it more akin to a sport like ice hockey or figure skating. This also allows for some to attempt to fake broad prowess with highly specialized skills. Don’t be fooled.
These aspects of ice skating often requires speed, balance, agility, precision and power. It is possible to achieve great success in one or even two areas with such poor foundations that overall success is impossible. That doesn’t stop people from trying!
Analytics is no different. It is likely even more common. Competition has a habit of weeding out those pretenders on the ice. Competition is more indirect in analytics — allowing many to fake it, for a while. Eventually the reality that bad analytics leads to bad business results — exposes the weak skaters here as well. Build solid foundations, learn great angles, embrace the full set of skills, and avoid skating on thin ice.
