Share with your friends










Submit

Analytics Magazine

Sports analytics, Part 2

November/December 2011

The role of predictive analytics, organizational structures and information systems in professional sports.

Sports analyticsNote: In Part 1 of this series, we defined sports analytics as “the management of structured historical data, the application of predictive analytic models that utilize that data, and the use of information systems to inform decision makers and enable them to help their organizations in gaining a competitive advantage on the field of play.” We also looked at the increasingly diverse and sophisticated sources of data that in turn are driving explosive growth in the field of sports analytics. In Part 2, we examine the ways in which predictive models and information delivery systems are leveraging these growing mountains of data to create the types of competitive advantage that every team is after.

Benjamin Alamar Vijay Mehrotra

By Benjamin Alamar (LEFT) and Vijay Mehrotra (RIGHT)

Predictive models are a key component of every effective sports analytics program because these models translate raw data into useful information. For example, there is very little value in the motion capture data described in Part 1 without skilled analysis that transforms that data from millions of raw records into actionable information a decision-maker can understand and trust. Data sources alone are clearly unusable, as no decision-maker would be able to draw conclusions from these mountains of raw data. Once such data is analyzed, however, the analysis results have the potential to become a valuable and unique tool to aid decision-makers in making better decisions.

Teams often begin their use of predictive analytics because they are looking for a tool to reduce the seemingly high error rate in decision-making around their sport’s amateur draft. Given the financial investment and opportunity cost associated with high-round draft choices, teams that spend a draft pick on a player who does not make a significant contribution find they have made an expensive mistake. Conversely, drafting well can make a huge impact on a team’s fortunes, sometimes immediately and often for several seasons to come.

Given these risks and opportunities, the draft is a place where teams have historically spent a lot of time and energy, helping open the door for predictive analytics. The draft also provides a fairly easy starting point for teams because much of the data used is publicly available, building the models does not require the time and attention from decision-makers, and the use of the results does not require buy-in anyone other than the top decision-maker.

In 2005 NBA draft, the Portland Trailblazers commissioned a company called Protrade Sports (since renamed Citizen Sports and acquired by Yahoo!) to create a predictive model for the draft using college data [1]. The model, a logistic regression, was developed with historical NCAA box score data, historical draft information and performance of former college players in the NBA. The output of the model was an estimated probability that the player would be a contributing player in the NBA. This analysis summarized 10 years of NBA drafts into a meaningful and useable measure of each player’s prospects.

More data, more technology

As the complexity of the data available grows, so do the techniques and skill required to create useful information from that data. Starting in 2006, SportVision began using motion capture technology to track the trajectory and speed of every pitch in Major League Baseball. This created a new and complex data stream for teams to analyze and question. Some teams had analysts on staff already that were capable of handling this type of data, but the Tampa Bay Rays realized that they did not and recruited a physics and math professor named Josh Kalk to, among other questions, analyze how pitchers’ release points change for different pitches [2].

Recent years have seen an explosion in the use of information systems designed to support analysts and decision-makers in professional sports. These applications, including desktop and mobile software, as well as customized Web pages, can be thought of as interactive reporting systems. Such systems sometimes serve as a platform for delivering the results of predictive models and supporting “what if”-type analysis. However, to date their primary purpose has been to quickly deliver easily customized data summaries, often called “descriptive analytics.” Thus, just as in the business world, these types of systems (typically desktop packages, interactive Web pages and mobile applications) present the world of professional sports with both opportunities and risks.

As the saying goes, “In the land of the blind, the one eyed man is king.” With a powerful ability to summarize vast amounts of information and deliver the results of these summaries, information systems enable data to be proliferated in an “on-demand” manner quickly and cheaply. Over the last few years, teams such as the Houston Rockets, Philadelphia Eagles and Cleveland Indians have all advertised for database programmers to work on the sport side of the organization. These teams, among many others, have been building the human resources needed to implement an information system capable of providing the organization with a competitive advantage. The first goal of many of these systems is to simply put all of the organizations information in one place, so that decision-makers have more efficient access to the information that they need.

The New Orleans Saints took information systems to the next level, in part by utilizing the ICE System currently available from Stats Inc. on draft day. This system replaced the magnet boards that display the depth charts of every team in the league and eat up considerable human resources to construct and maintain. During the draft, the Saints had access to “virtual” magnet boards that allowed them to easily see every team and maintain a list of draftable players that was updated in real time. They also enjoyed instant access to a large set of information on every player in the league to aid them in efficiently evaluating potential trades. This system ensured that decision-makers had the information that they needed when they needed it, and that this information was accurate.

In addition, for organizations with proprietary data and the knowledge to leverage it, their own customized systems can be seen as the key to capturing competitive advantage. For example, Daryl Morey, the general manager of the Houston Rockets, recently wrote that private data is what ultimately drives competitive advantage [3].

However, Morey is still something of an outlier in the world of sports: an MIT MBA with an undergraduate degree in computer science who has also invested in an army of analysts on staff to help him make sense of the data. It is easy to imagine less technically savvy sports executives who might see the data pumped out by information systems as a low-cost analytics “solution.” But from our experience, rarely will data alone – no matter how quickly it is summarized and graphed – provide actionable new insights.

In fact, as we look at the world of sports today, there appears to be a profound dichotomy: despite a rapidly growing interest in applying analytics, an explosion in data, a plethora of companies peddling information delivery systems and making promises, the actual impact of analytics on the world of professional sports is still somewhat limited.

Big Money, Small Analytical Impact

What makes this truly surprising is the sheer size of the professional sports industry. The National Football League is a $9 billion per year industry, and Major League Baseball reported revenues of $7 billion in 2010. Meanwhile, a financial dispute between the league and the players union has put the NBA’s $3.8 billion dollar business at risk for the upcoming 2011-2012 season. Given the amount of money on the line in the world of professional sports, it is natural to wonder why analytics do not (or at least “not yet”) play a more prominent role in the way decisions are made.

We believe that the most significant structural barrier to the growth of sports analytics is not only the absence of a clear doorway for teams to systematically get involved with this nascent field but also the lack a clear process for developing the skills needed to open that doorway and have an impact on the other side. That is, for most sports executives and for many would-be sports analytics professionals, it is simply not clear how to get started.

Sports executives face a daunting challenge with regards to analytics. Today’s professional sports world is filled with win-at-all-costs pressures, and given the increased availability of data and visibility of analytics, a lot of pressure is on executives to find a way to harness this potential source of competitive advantage. However, most decision-makers have little to no experience or training in the methods and tools of analytics, and as such are not well-equipped to evaluate the landscape of options.

The result is that some organizations start small, at best thinking very incrementally about analytics and at worst simply adding a small amount of staff and/or software as window dressing. Meanwhile, other organizations have absolutely no idea of how to begin and thus simply do nothing.

Another related problem is the lack of a true talent pool for sports analytics professionals. Today, there are a handful of people who whose academic and/or business training has provided them with strong analytic skills, and their passion for sports has led them to attempt to bring their skills to professional sports. There are, however, a larger number of individuals who are passionate about sports and “know enough to be dangerous.” That is, plenty of people have had enough statistics classes or have worked with Excel tables enough to be able to create analytic models. The models may not be well-formed or properly estimated as the purveyor of the model does not have the training needed or exposure to a larger set of tools that are needed, particularly as the data becomes more complex. Most decision-makers in sports do not have the training to identify the strengths and weaknesses of various modeling approaches. This information gap may cause them to opt for the cheapest solution without understanding the value that a more robust approach could deliver.

However, the good news is that there are signs of change, both for executives and analysts. Events such as the Sloan Sports Analytics Conference at MIT provide venues for executives to discuss what is happening in the field and to get a sense of what kinds of analyses, and results, are possible. In addition, for practitioners/researchers interested in the field, such events provide a chance to not only demonstrate their capabilities and findings, but also to learn about what types of real problems professional sports executives are wrestling with.

Meanwhile, a few universities, including Georgia Tech and the University of San Francisco, have begun to offer courses in sports analytics. These courses give students exposure to the kinds of real-world challenges that teams are facing and some sense of the concepts and to the types of tools that they must master to make a real contribution.

But there is still a long way to go.

In Part 3 of this series, we will look to the future, discussing the structures and institutions needed to provide potential analysts with the training that they need to build and grow a sports analytics program and to provide decision-makers with the knowledge and tools they need to implement and lead a strong sports analytics program.

Benjamin Alamar (quantsports@gmail.com) is the founding editor of the Journal of Quantitative Analysis in Sport, a professor of sports management at Menlo College and the director of Basketball Analytics and Research for the Oklahoma City Thunder of the NBA. He is co-author of the annual “Football Outsiders Almanac” and a regular contributor to the Wall Street Journal.

Vijay Mehrotra (vmehrotra@usfca.edu) is an associate professor, Department of Finance and Quantitative Analytics, School of Business and Professional Studies, University of San Francisco. He is also an experienced analytics consultant and entrepreneur, an angel investor in several successful analytics companies and a San Francisco Giants season-ticket holder.

REFERENCES

  1. Ma, Jeff, “The House Advantage: Playing the Odds to Win Big In Business.”
  2. Keri, Jonah, “The Extra 2%.”
  3. http://blogs.hbr.org/cs/2011/08/success_comes_from_better_data.html

business analytics news and articles

Analytics Blog

Electoral College put to the math test


With the campaign two months behind us and the inauguration of Donald Trump two days away, isn’t it time to put the 2016 U.S. presidential election to bed and focus on issues that have yet to be decided? Of course not.


Headlines

What you like in a CEO candidate may not deliver results

Attributes that often get someone hired as CEO may not be the ones that drive success once they are at the helm of the company. That’s one of many provocative insights in a study by consulting firm ghSMART. The study, recently featured in Harvard Business Review, identifies characteristics that differentiate the most effective leaders. Read more →

Paying for online reviews backfires among socially influential

Online user reviews have become an essential tool for consumers who increasingly rely on them to evaluate products and services before purchase. The business models of online review platforms such as Yelp and TripAdvisor and e-commerce sites such as Amazon and Expedia critically depend on them. Should such sites pay users to encourage them to write reviews? Read more →

UPCOMING ANALYTICS EVENTS

INFORMS-SPONSORED EVENTS

CONFERENCES

2017 INFORMS Healthcare Conference
July 26-28, 2017, Rotterdam, the Netherlands

CAP® EXAM SCHEDULE

CAP® Exam computer-based testing sites are available in 700 locations worldwide. Take the exam close to home and on your schedule:


 
For more information, go to 
https://www.certifiedanalytics.org.