Analtyics!(?) (3/6/12)

By Paul Bessire

Tuesday, March 6 at 8:30 PM ET

Today's blog will recap the MIT Sloan Sports Analytics Conference and the greater context of "analytics." Please note that we have just launched a Podcast (some of the content below was covered in the inaugural podcast; we will also have several new podcasts for the NCAA Tournament next week) and Tuley's Vegas Beat column in our new Features section. Also, as a reminder, the MLB Preview will be coming out this Thursday, March 8th with O/U team win total picks (available to subscribers. While I intend to get another blog in before Selection Sunday, I do anticipate a return of The Madness on-going NCAA Tournament blog next Thursday (though it will likely be broken up into daily tournament blogs).


My favorite thing about the word "analytics" is that it is not a word - at least not in the way that our culture has come to classify a word, meaning that, if you type it, a squiggly red line appears below it. Given that analytics is not a word, it is extremely difficult to define. Yet, it is the buzz word right now in many business communities - sports or otherwise. Web "analytics" (notably from Google - another word, especially when used as a verb, that fits in the "not a word but used all the time" category) dictate the way the internet is run. There is an "Analytics" magazine. Hundreds of companies have appeared recently utilizing the word "analytics." My alma mater changed its Masters of Science in Quantitative Analysis program (of which I was a part) to a Masters of Science in Business "Analytics" (not nearly as nerdy, but probably more appropriate, plus it doubles as an "MSBA," which is a more accurate representation of what it is than simply an MS degree). And the MIT Sloan School of Management hosts an annual "Sports Analytics Conference" each March that was attended by 2,200+ people last week.

I was one of those people. The conference was extremely well-run, well-received, informative and completely packed with interesting panels, presentations, discussions, content and interesting attendees and speakers. Before I get to anything else, I need to commend MIT Sloan (and all others involved at the Hynes Convention Center in Boston) for the great work... But one thing is for certain, it is evident that we have a long way to go before common sense surpasses bullheaded convention. And if the sports industry is to be considered more advanced in this respect than most industries, there is a great deal of catching up to do worldwide.

A definition that I prefer for "analytics" is: the utilization of technology to answer questions correctly that the average human mind cannot without bias or error. One definition of "analytics" that seems to be conveyed internally at ESPN (I'm reading between the lines here, but I heard this expression almost 100 times at panels) is that "analytics" tells a story. In short, "analytics" is the discovery and application of the truth.

What separates those of us who utilize "analytics" from those of us who resist it is that we understand that we are not smart enough to properly evaluate all aspects of a decision on our own, especially without our own, natural biases getting in the way. There is nothing that a computer can do with information that I do not wish I can do in my head. The notion of "gut," or "intuition" as being different than what technology tries to help us decide with analytics is preposterous. No coach or general manager or CEO or truck driver or anyone who has to make a decision at any point throughout the day is going make a decision that he or she thinks is not in the best interest of achieving his or her goals. Technology forces us to define those goals and then aids us in analyzing the factors that go into the decision and likelihoods of results from making all possible decisions. Computers are not "smarter" than we are (they may be in as little as 20 years, but we can discuss that at another time). Our minds are very strong machines, yet - as with many sets of data fed into a computer can also have - can be cluttered with too much noise, bias and external activity.

This is why I have never understood any resistance to the utilization and consumption of technology in decision making. By employing technology, I am not doing anything or trying to solve any problem that someone else is not; I just do it quicker, with minimal bias and with more factors considered. At the Basketball and Coaching Analytics panels last Friday, Jeff Van Gundy's comments (antics) have been characterized as "stealing the show." While JVG did so mostly by talking the loudest and the most often, he also played the role of old school codger who was not down with these newfangled ways by consistently poking fun at any "analytics" that sounded like common sense. Many thought it was funny; several applauded and cheered it on. I get the joke, but I also feel bad for him and, more importantly what he (and others like Jackie MacMullan on the various panels) represents; a much greater cultural issue.

Sure, any analysis that says that Michael Jordan was good at basketball sounds like common sense. It should. The vast majority of findings of any technological analysis should line up with what we believe to be true. We have to give ourselves that much credit. But "analytics" should still represent our greatest possible attempt at truth for a topic. Sometimes, for whatever reason, as it did for Christopher Columbus, Bill James and countless others who have helped to define and shape what we believe as truth at this time, the truth violates conventional/consensus (not necessarily always unanimous - Earl Weaver probably understood what Bill James was doing long before he did it as an example and I am sure there are many people who naturally understand or have understood concepts that have come from convention-defying "analytics"). To applaud someone or something that is not interested in knowing or utilizing facts (whether they are found through technological advances that have exponentially evolved mankind over several centuries or otherwise) is to celebrate inefficiency and seek bias and error. That does not make much "common" sense.

The point is not to turn the tables on Jeff Van Gundy, but rather to question why anyone attempts to delineate between decision makers. Some may use technology. Others may not. We are all trying to make the best decisions we can. In my case, I concede that I am not smart enough (some may be - that's the best part about this conversation, some people are really, really good at making decisions and grasping the real truth; most are not but we can make up that ground with tools that we have built to help us) nor free from all biases, so I employ devices and tactics that we have developed to surpass those obstacles. And those devices and tactics are only going to get better.

Coaches do not like to think that they can be replaced or even told what to do by an inanimate object (no one does). Players do not like to think that they are simply made of numbers (who isn't?). Whether it's words or numbers (to me, they are interchangeable, to most, let's say at least more than 50%, they are not) telling the story, a player is who he is and knowing the facts about him can only help maximize his potential.

All businesses in every industry should leverage information and technology to make the best possible decisions they can make. Any business that does not is inevitably making inefficient decisions. Get that and you will instantly be better at any job that you have (especially if you can free your mind from the day-to-day to take steps back and remove inefficiencies from all that you and your business are doing). At its core, that is the premise of this site - making the most efficient ("optimal" if such a thing exists) decisions about investing in a market.

This is what those who shined at the conference, like Bill Simmons, Bill James and Mark Cuban, get. They are "common sense" guys (a phrase Simmons often uses). They are willing to take a step back, question anything, gather the facts, analyze all relevant inputs and outputs and draw conclusions that are free from external biases. That's what "analytics" is all about. Each of those three may have a slightly different approach to get there, but at the end, once the truth is uncovered, it's hard to argue with it.

Inventing Statistics
One concept, and it may just be semantics, often mentioned by Bill James that I disagree with is that of "inventing" statistics. Inventing a statistic is the same thing as creating a new word (like "analytics" or "google"). It's not that no player in the history of basketball blocked a shot until 1973 (when the NBA started tracking blocked shots in its box scores), or that defensive players started sacking quarterbacks before they started hurrying them or even that, in people's minds, when making roster decisions, (perceived) value over a replacement player was never considered. Statistics are words we use to tell a story of what happened or what can happen. At some point soon, we will likely be publishing the "statistics" that we use in simulation, which are simply the per-event likelihoods of an outcome (like (walks)/(plate appearances - hit-by-pitch) or (team defensive sacks)/(opponent pass attempts + opponent sacks allowed)). We are not inventing these numbers. They recount what has happened in the past. And we employ them to predict what should happen when various entities interact in specific situations. It's not revolutionary (though I am not sure that I believe that anything is), but we have integrated what we know from the stories that have been told in the data with technology that can process millions of factors every second and several years of refinement founded in empirical performance and advances in both data and technology.

The Future
Just as no language is bound by the quantity of words it could use, "analytics" is not bound by finite factors that can be evaluated. The example regarding Jeff Van Gundy above was not the norm at the MIT SSAC (that it exists at all is more my concern). The vast majority of attendees, presenters and panelists are focused on utilizing technology to mine and manipulate information to improve the creation of teams, the user experience, pricing of tickets, consumption of the product, future technology, expansion of new endeavors and one's ability to exploit markets (we're part of that last one, though the sports books would have a very different definition of "improve" than we would). I believe strongly in the importance of every variable input that exists in our engines. I also believe strongly that simulation will always be the best approach predicting future events that involve interacting entities. However, it would be silly of me to (ever) believe that we have completed our evolution of "analytics" to the point that it is impossible to improve the incorporation and manipulation of data in our simulation engines. I am not averse to exploring on-going work and research on psychological factors, physiological advancements, optical tracking of player action and movement, off-field evaluations and more. In fact, we have already begun to incorporate research and data discussed at the conference. Everything that is relevant to output should be able to be quantified (i.e. my assertion that "intangibles" do not exist) as an input. The key lies in consistently assessing the relevance of newly uncovered information. Our single most important goal as a business that produces/sells bankroll management recommendations is to align the confidence and performance of our picks. This is a process that never stops.

Other Highlights
Here are some other highlights from the convention:

  • Research Papers/Presentations - I anticipated spending the majority of my time at the conference either watching panel discussions from team front office personnel and media members or networking. While I did a significant amount of both, I also found myself gravitating far more to the research paper presentations than I had anticipated. Several panels were stale and awkward in the sense that few people really wanted to say anything and many felt compelled to downplay "analytics" (by now you fully understand that I don't get or appreciate this mentality). In the research papers, it was exactly the opposite. Consequently, I did at least as much networking with researchers and presenters that could help us with inputs and engines as I did with others who are already established in the industry and may help us in other ways.
  • Sports Bookmaking and Gambling Panel - This was my first SSAC and, thus, the only time I have seen this panel (apparently, it was very different last year). Jeff Ma (of "21" fame) moderated with handicapper "Dr. Bob" Stoll, ESPN's Chad Millman and Cantor Gaming's Matthew Holt serving on the panel. Stoll and Holt were naturally the most interesting to me. Holt conveyed (strongly) that they believe (strongly) in their "algorithms" (so strongly that anyone can bet "a million dollars" at their books any time - though he acknowledged that they may force you to take a different line if you want to do so). It's not as though Holt could have ever said, "our algorithms have loopholes" or "we don't know what we are doing so we will only let you bet $500," but the approach to emphasize "algorithms," "a million dollars" and even "potential smoothing" (an approach that we incorporate and will utilize even more to help with accounting for previous performance) was excessive. Meanwhile, Dr. Bob may be one of the few people who knows exactly what it is like to be me (for lack of a better way of phrasing). Our approaches are different and I obviously believe (strongly) in's methodology, yet I could relate to almost everything that he said (including the tightening of lines in recent years and the difficulty in explaining variance to gamblers).
  • eSports - I did not know much about eSports (professional video gaming) before Saturday. From what I can tell, i's best and brightest were on its panel. I know more now. Chances are that everyone will very soon. If there is an entity that can lead the charge to completely alter the way we consume sports media (for the better), I think this is it.
  • The Price is Right - Drew Carey, who has become one of the more forward-thinking owners in American professional sports (as owner of the MLS' Seattle Sounders) attended the conference to talk about Soccer Analytics and Franchises in Transition. I will never forget exiting the gambling panel to witness Carey breaking down Price is Right economics with an MIT student. Never.
  • Bill James - Bill James, as alluded to above, sports' equivalent to Columbus, Newton (Isaac, not Cam or fig), Einstein or Thoreau, is the most iconic figure in "sports analytics" due to his willingness to question everything about baseball convention and back up his conclusions with facts. I view him as the ultimate story teller for the game. He views himself very humbly. Without saying it directly, I also think that he viewed this conference as a culmination of his life's work and is ready for/looking forward to others driving the advancement of his industry/cause. Never was this more evident than when he earned the first ever "Alpha" lifetime achievement award and stated, "It is old people who win awards, but young people who change the world." Bill James changed his world by questioning everything and seeking the truth. May we never stop doing that.
  • BS Reports - The two live BS reports with Bill Simmons interviewing Bill James and Mark Cuban respectively were fantastic. All three participants were genuine, candid and "get it."

As usual, if you have any of your own comments about this article or suggestions about how to improve the site, please do not hesitate to contact us at any time. We respond to every support contact as quickly as we can (usually within a few hours) and are very amenable to suggestions. I firmly believe that open communication with our customers and user feedback is the best way for us to grow and provide the types of products that will maximize the experience for all. Thank you in advance for your suggestions, comments and questions.