“I think that what I did is just a slightly more algorithmic, large-scale, and machine-learning-based version of what everyone does on the site. ”
Over at Maize N Brew, Space Coyote breaks down Michigan’s fateful 2 point conversion attempt against Ohio.
It is excellent analysis. More, please!
I think this is correct. US consumers are still repairing their household balance sheets which have been distorted due to super easy credit.
From the comments:
“The United States economy is like a poker game where the chips have become concentrated in fewer and fewer hands, and where the other fellows can stay in the game only by borrowing. When their credit runs out the game will stop.”
-Marriner Stoddard Eccles, Beckoning Frontiers (1951)
The little country that could.
When presented with a bad deal by the international community and the banks, Iceland said, “No thanks” and demonstrated the validity of the Austrian view that a short sharp shock is preferable to a long drawn out soft landing.
My uncle, a former seminarian who has two masters’ degrees (one in English Literature and one in Business Administration) and has been running his own company for nearly fifty years, once told me that it isn’t enough to be good at just one thing. It isn’t enough just to be great at sales. You need to be great at sales and operations. One dimensional resources abound and the way to stand above the crowd is to be great at many things.
Analytics and business intelligence suffer from this same uni-dimensionality. Kaiser Fung, author of Numbers Rule Your World: The Hidden Influence of Probability and Statistics on Everything You Do and VP of Strategic Analytics at SiriusXM, seems to agree.
Fung says the point of Numbers Rule Your World is that data analysts really should speak to people in English, and in the language of business as opposed to in the language of math and in the language of statistics. Of course the language of math and statistics is important for analyzing the data and reaching conclusions, but once analysts know what the insights are, they should look at communicating those insights in a way that’s different from doing the math.
That’s the challenge Fung is talking about—the challenge of communicating the findings.
“I think that’s where the statistics community has not done enough yet,” he says. “We are very obsessed with coming up with methodologies and coming [up with] new techniques that may be marginally better than the existing techniques. That stuff is great for academia and research, [but] you’re talking about practice and business interpretations.”
Unfortunately, this kind of multi-discipline, multi-dimensional thinking is rare. Finance people congregate around finance people because they share a common language and often a similar set of experiences. IT people spend time trying to wow each other with their mastery of code or a nifty new configuration. When the two groups meet to discuss how to get something done, they often end up talking past each other. It takes a rare breed of person to act as the Rosetta Stone between the two groups. I can’t speak for disciplines like math or engineering, but Fung seems to think part of the problem is the way statistics & analytics are presented to students.
Fung points out that business thinking – how to teach data analysts to talk to the business in language the business understands – is often the hardest to develop because colleges and grad schools (outside of business schools) just don’t teach it
I’ll buy that. Math is a language just like Portuguese or SQL and unfortunately at the undergraduate level it is very hard to see applications of things like the Taylor or Mclaurin series (I pick on those because I never saw the application of either in my undergraduate calculus classes beyond clever math-nerd party tricks). The problem is that there are very few professors who are, pardon the pun, polymaths. My uncle is one of them. My high school physics teacher, who would recite Shakespeare as easily as he could rattle off the number of electrons in each orbital for Carbon or Neon, was another. I’m lucky to have had them as role models.
As business analytics professionals, we bear the responsibility of being examples for others of the value of speaking the language of business and analytics.