Sports Data Analytics Is Pulling Us Toward Artificial Intelligence Everywhere
Because we are reaching the end of a decade, all sports and other areas of popular culture are featuring articles on the most important events and people of those sports or fields of interest. This is a link to the Major League Baseball article looking back on the period from 2010-2019.
https://www.mlb.com/news/major-league-baseball-best-of-the-decade?t=best-of-the-decade-2010-19
However, the most important trend in all sports and entertainment has been the transition from decision-making based on traditional human intelligence and experience to decision-making driven heavily by data. This article (I cannot send a link in its entirety because The Wall Street Journal prevents non-subscribers from receiving articles) correctly notes that data analytics have profoundly reshaped all sports.
The use of artificial intelligence and data in sports is the metaphorical “canary in the coal mine” foreshadowing its extensive use everywhere. Great leaders will embrace artificial intelligence and data analytics and use both better than their less innovative peers.
Baseball started earliest in focusing on data, a trend highlighted in Michael Lewis’ 2003 book Moneyball, released as a feature film in 2011. However, all sports now rely heavily on data, and every team in every sport has adapted in varying degrees to what data analytics have yielded.
In professional basketball, as a result of data analytics, the 3-point shot is used more frequently, and players’ skills are evaluated completely differently than they were a decade ago. In professional football, passing and the use of the shotgun offensive formation have significantly increased. Ice hockey has changed in multiple ways as well.
Areas seemingly immune from analytics, such as the creation of hit songs, have incorporated data analytics.
https://www.theverge.com/2018/8/31/17777008/artificial-intelligence-taryn-southern-amper-music
As a film producer, I learned that decisions on theaters and screens available to newly released films were made on the basis of real-time data collected by services like Rentrak. That meant that my film, From the Rough, would not get final theater slots until the week before the release, which made targeted grassroots marketing impossible. It meant that high-cost broad-based TV, print, billboard and online marketing mattered more than ever.
Businesses are slowly, but surely, incorporating artificial intelligence into decision making. They would be moving faster, but large organizations have purchasing departments and mid-management people who are frightened of change and resisting the use of AI in places in which it could significantly improve decision processes.
The company I lead, MoveFlux Corporation, has a platform that helps supercharge sales and new business development. We do not succeed 100& of the time because many organizations have sales and marketing leaders wedded to old, inefficient ways of targeting and prioritizing prospects. However, as a new generation of people more comfortable with AI tools move into leadership positions that should change.
What’s next? I predict that sports will lead the way in a new generation of innovation in using data. Baseball will be at the front of the pack in using data to help find undervalued and overlooked opportunities, particularly in talent evaluation. The 2019 book The MVP Machine could have the same long-term impact as Moneyball did.
The “moneyball” philosophy was to use data to identify undervalued talent, but it assumed that the skill sets of players were inherently limited. In other words, whatever a team saw in a player after doing data analytics was a description of what that player was capable of offering.
The MVP Machine presents a new use of data analytics: identifying ways of making players far better through the use of technology and data analytics on conditioning and skill development. It described the cases of Trevor Bauer, now pitching for the Cincinnati Reds, Justin Turner and Max Muncy, now playing for the Dodgers, and the various pitchers the Houston Astros re-developed, among them, Lance McCullers, Charlie Morton, and Collin McHugh, as examples of how players could be made far better in mid-career or even in late career.
Although the book was written at the beginning of this decade, Rick Kash’s and David Calhoun’s book How Companies Win actually foreshadows this use of advanced analytics. The book makes what seems to be an obvious point, but one that few companies follow almost a decade later: competitive advantage arises from learning about customer needs no one else understands or knows about, and that customers may never have articulated. Kash and Calhoun focus heavily on what they describe as “latent” or “emerging” demand.
Relative to traditional marketing research, they make an obvious point: if a firm learns about customer needs by asking customers, they are using a technique that their competitors can use as well. If customers share insights with them, they can share the same insights with every other competitor. We see this when big businesses issue Requests for Proposals or even earlier-stage Requests for Information. The insights are obvious and highly public. Competitive advantage arises from doing a better job in interpreting weaker, less obvious signals or even needs that the customer has never explicitly articulated.
The question this raises is why customers do not tell us what they need. There are at least two reasons:
- Many customers have given up trying to solve a problem, because no solution has been available over a long period of time. They may have looked at multiple points in time at possible solutions but found what was available either inadequate or too costly.
- Many solutions require customers to reimagine how a new solution might present itself. Recently, I viewed a documentary on Singapore, which is rapidly moving toward self-sufficiency in food sourcing by using “vertical farming.” A firm called Sustenir Agriculture is leading the way.
We have always linked agricultural potential to the availability of land and rural labor. Sustenir is presenting a new paradigm that disconnects food from rural, outdoor, land-based agriculture. It will not solve food problems by itself, but it upends the traditional view that solving the food crisis requires land, labor, and heavy capital equipment.
At Pitney Bowes, one of the best new businesses formed during my tenure arose from spotting a relatively weak market signal and realizing that there was potentially a much bigger opportunity.
We noticed that many customers asked for cash advances on their postage downloads to enable them to complete an end-of-the-month mailing. We charged them a flat $25 fee and ended up with about $6 million a year in revenues, most of which was highly profitable. Our President of Financial Services thought that this signaled a large opportunity for providing revolving credit to these customers.
Our chief operating officer was highly skeptical. After all, every one of these organizations could get money at far lower interest rates from their own corporate treasury departments. At first glance, this should not have been a big opportunity for us, but it turned out to be one. The reason was that these mailroom directors believed that their treasury departments were often too bureaucratic or slow in providing funds for their operations. We gave the mailroom directors far more control over their budgets. Control had an exceptionally high value, even at very high-interest charges. We found a big latent demand and created a $200 million business from it.
Many otherwise intelligent people believe that, in the future, artificial intelligence software will require large amounts of data, highly automated processes that displace humans, and very precise conclusions from a perfected technology system. In fact, the future will belong to those who design artificial intelligence algorithms, using software applications like Tensorflow from the Google Cloud library, to spot emerging patterns and market signals from small data sets.
Success will come from rapid learning and improvement from continuous data feeds, often helped by human intelligence, especially early on in a process of using artificial intelligence. Smart AI practitioners do not spend lots of money on initial data collection. They take whatever data is available and spend a lot of time creating algorithms that learn and improve quickly.
The other counterintuitive insight about artificial intelligence is that those who succeed will be best at finding highly non-obvious causes or connections within the data, as opposed to those who get to obvious connections faster. Human intelligence is satisfactory in spotting the obvious. AI is ultimately better at finding connections that would seem less important or completely irrelevant to humans.
At Pitney Bowes, we stumbled on the fact that one of the most powerful predictors of whether customer who received a postage meter that had been shipped to it would cancel after our 90-day free trial was not the objectively big potential need the equipment might fill, but whether the customer opened the box, plugged in the meter and fed mail into it within a few hours after the box arrived.
Early use was more important than almost any other factor. AI would easily uncover this; it was purely accidental that we discovered this connection years after launching our 90-day free trial program.
The future will belong to those who use the best combination of human and artificial intelligence, not those with the most data or those with the largest IT staffs or those who recruit the most software programmers. They will find and increase value in all processes that no one else has found, whether it is in making an athlete perform far better than he or she had previously performed, increasing the power of a hit song, finding an unarticulated customer need, or bringing two people together.
On the last point, I have often wondered whether my wife and I would have come together if AI date matching tools had been in place. We were and are very different people, but we both adapted rapidly to one another and knew how to maximize value from the other person’s strengths.
We need to recognize the opportunities and the limitations of artificial intelligence and how it works most synergistically with human intelligence to realize its potential. Sports will help show the way because they are the most meritocratic of human endeavors. As fans with 24x7 ability to register our opinions and to vote with our pocketbooks, we force sports franchises that we care about to be focused on continuous improvement.
That is why these seemingly trivial, but interesting, retrospectives on the past decade relative to the transition to more data usage in all sports are so important. They show us where we can be in many other fields of human endeavor in the decades ahead of us.