Cavaliers

Playing the Odds: Everything we know about the Cavaliers and analytics

Bennet Rosen Header

Bennet Rosen Header

Where numbers and charts collide with the “eyes and ears” of NBA decision making

Have you ever watched the movie 21? The one about the MIT blackjack players who got rich quick? It was mostly real, if you didn’t know. The movie was based on a 2003 book titled Bringing Down the House, written by Ben Mezrich. Back then, Mezrich was not the incredibly successful writer he is today.1. The main reason he even wrote his first bestseller was the persistence of the story’s real-life lead character: Jeff Ma.

Ma is an interesting fellow. He received the pseudonym Kevin Lewis in Mezrich’s book. He was described as your run-of-the-mill card-playing collegian – spry, cynical and spirited – who also happened to be smart enough to have a mechanical engineering degree from MIT.

Years after his well-known poker-playing escapades, Ma also became the reason why Ben Alamar, the Cleveland Cavaliers’ Senior Analytics Consultant, works in the sports analytics world today. Alamar never intended to be in the business. The economics Ph.D. out in California was just a casual fan and had written just one grad school paper about sports, on the NFL and playoff probabilities. He had his own separate life. Heck, he had never even heard of Moneyball when it first came out. But that all changed when Protrade came along in 2004.

Protrade was a unique entity, a product of Ma’s personal passions and the mid-2000s gambling boom. The website was a sports stock market where Internet users bought or shorted stock on certain players. It was a fascinating tool that rode a brief wave of popularity that included a rebranding to the name Citizen Sports, a strategy change to being more fantasy sports-focused and, finally, a 2010 sale to Yahoo.

Before its eventual restructuring and death, Protrade gained some business notoriety as professional sports teams gained awareness of the site’s analytics potential. Those deals enabled Ma to have a brief career as a sports consultant, although he’s now the CEO of a tech company called tenXer, his fourth start-up. One Protrade project in particular, with the Portland Trailblazers on projecting college basketball stats to the pros, was made possible by the economic analyses of Alamar.

Ben Alamar  (via The Booksmith)

Ben Alamar
(via The Booksmith)

With this one taste of the sports world, Alamar was hooked. He has admitted since that it was a stumbling and unplanned path into the industry, but he’s never looked back. Since this first foray with Protrade, he’s been a leading author, researcher, public speaker and consultant across multiple sports leagues – a number-devouring Superman. The oddity: His current day job is as Clark Kent as it gets, serving as a professor of sports management at Menlo College in the Silicon Valley.

Inspired by the Protrade experience, in 2005, Alamar founded the Journal of Quantitative Analysis in Sports. An official journal of the American Statistical Association, it was the first ever sports analytics academic publication. He transferred editorship of the journal to Bowling Green math professor Jim Albert in 2011.

On the consulting end, Alamar’s impressive economics background and brief Protrade days led to a nine-month gig as a statistician for the San Francisco 49ers. He has worked with ESPN and The Wall Street Journal on various projects and articles. NBA general managers were especially fascinated by his continued public draft research.

Sam Hinkie, then a Boston Celtics executive and now the oft-discussed Philadelphia 76ers GM, recommended Alamar’s name to Sam Presti, who was named as the Seattle Supersonics GM in June 2007. Just a few months later, Alamar was officially added to the fold as a team consultant.

*****

“Baseball analytics is like a full-grown adult with a pretty good job. Basketball analytics is the soon-to-be high school valedictorian voted most likely to succeed. Pro football analytics is the second-grade son of the famous businessman who nobody’s really sure about yet.” – SB Nation’s Bill Connelly, March 2014.

*****

Basketball is undoubtedly several years behind baseball in the data revolution, but that doesn’t mean incredibly talented statistical minds aren’t already working within the frameworks of countless NBA organizations. Some of the most commonly known names are Houston’s Daryl Morey and Memphis’ John Hollinger. Other individuals that have been reported on include Dallas’ Roland Beech (also a former Protrade consultant), Boston’s Mike Zarren and Drew Cannon, and Portland’s Ben Falk.

Countless others rose from the ashes in the previously much more robust corners of the APBRmetrics forums.2 In the early days of basketball analytics – with other leaders like Dean Oliver, Kevin Pelton and Neil Paine, now all members of the ESPN hierarchy in various roles – this was where up-and-coming “basement number-crunchers” posed questions, tested hypotheses and shared research.

Over time, as in baseball, NBA teams seeking any possible type of competitive advantage eventually poached the stats writers criticizing their daily work on public forums. That natural evolution has somewhat quieted the open industry.

Jon Nichols, current Cleveland Cavaliers Director of Analytics, is one such former avid public researcher who now works on the private side. That description, however, dates Nichols a bit too much: He finished his University of Michigan undergraduate degree of sports management in only 2009 — he’s about the same age as the average NBA player today.

Jon Nichols  (Via Twitter)

Jon Nichols
(Via Twitter)

A native of south Florida, the inquisitive Nichols ironically is a childhood fan of the Miami Heat. Throughout his undergrad days, he was extremely active in the blogging and analytics communities. Even though his first sports jobs were sales internships, he tested his hard-core analytics mettle on a variety of sites: Hardwood Paroxysm, NBADraft.net, 82Games.com, Orlando Pinstriped Post (née Third Quarter Collapse) and even The New York Times.

Eventually, in late 2008, he created his own website, Basketball Statistics.com, where he shared his original statistical creations and packaged together all of his other Internet work. Nichols was open about the site’s intention: He was lobbying for a career in basketball analytics. This was his passion, no doubt. He was going to do everything he could to make it easy for an NBA organization to spot his work.

On Jan. 29, 2010, the long-awaited post finally was published at Nichols’ website. “Due to some recent developments, I will not be posting updates to Basketball-Statistics.com for the indefinite future,” he wrote, adding that the site will remain intact and apologizing for having to leave so abruptly. Eventually, fans learned that Nichols had taken an analytics position with the Milwaukee Bucks. Nichols hasn’t been active in the public sphere since.

*****

“For ‘competitive reasons,’ Mr. [David] Griffin wouldn’t disclose the size of the Cavs’ beefed-up analytics department, but it’s evident the organization, in the words of Mr. [Tad] Carper [Senior VP of Communications], is going ‘all in’ with the data.” – Kevin Kelps in Crain’s Cleveland, September 2013.

*****

“[Chris Grant] beefed up Cleveland’s analytics department, turning it into one of the largest in the league.” – Grantland’s Zach Lowe, February 2014.

*****

The origin of the Cleveland Cavaliers’ analytics narrative is unknown. It’s been hovering around for a while, mostly interconnected to the workings of the now-former general manager Chris Grant.

Grant’s direct background, actually, is in psychology. After completing a master’s degree in educational leadership, the 6-foot-10-inch former college basketball player joined the Atlanta Hawks front office. After nine years in Atlanta and five more in Cleveland under Danny Ferry, he then had a largely unsuccessful 3.5-year run as the Cavs GM which resulted in his mid-season firing last month.

Sure, the since-scapegoated Grant worked as a video coordinator and an advance scout before his rapid ascent to the GM position during the fateful summer of 2010. But Grant simply did not have a noted analytics background himself. It is possible then that the Cavs’ roster moves alone – specifically, their curious draft picks – maybe were responsible for this long-running storyline.

In 2011, Tristan Thompson was an undersized, offensively limited power forward who was a likely lottery pick that very few saw as a top-5 selection. One person who did put weight into Thompson’s stock? PER inventor John Hollinger, who was still working as a statistician for the Disney-owned cable company known as ESPN. Hollinger’s Draft Rater actually had the Canadian by way of Texas as the draft’s top forward.

After another bad season, the Cavs again had the No. 4 draft pick in 2012. This time around, Dion Waiters reportedly had a draft promise and his camp shut down all in-person meetings. Hollinger’s Draft Rater again liked Waiters, giving him the best rating for a small wing since Dwyane Wade. The Cavs shocked many by picking him so high, again selecting a player they were largely unconnected to until the final days.

Age plays a large factor in statistical draft analyses, and Thompson and Waiters were both younger than most of their peers. They had oddly fitting games, especially offensively, but racked up stats in important athletic categories such as blocks and steals. Perhaps those selections are responsible for the eventual narrative of Grant’s analytics focus? We’ll likely never know for certain.

What is known is that neither Alamar nor Nichols had any impact on those picks; neither were hired until the 2013 calendar year. There was no listed analytics professional in the team’s 2012-13 media guide, although that’s not necessarily a definitive answer. Dan Rosenbaum, also a noteworthy economics Ph.D. and one of the original basketball analytic minds, served as a team consultant starting in 2005, but his end date is uncertain3.

During the trade deadline press conference three weeks ago, new Cavs GM David Griffin began with a unique roll call of several notable front office staffers who he felt played an integral role in the team acquiring a seven-foot, floor-spacing center named Spencer Hawes. Griffin name-dropped Jon Nichols early on. Later, when asked by WFNY to what extent Nichols and the analytics team were involved in the trade for Hawes, Griffin stated:

“Quite a bit, quite a bit, and Jon’s had a big role in a lot of things over the year. Again, none of these things happen in a vacuum. You know, I got my job in the beginning in Phoenix to a great degree because of analytics. I believe in them and their usefulness and I understand the limitations of them as well. He (Nichols) played as big a role as the eyes, ears and numbers can play in conjunction with one another. It’s something we believe very strongly in.

“Again, because ownership supports us to such a huge degree, we can invest in those processes. And we’ve got Ben Alamar, who’s another consultant on the statistical side that does great work for us. And we’ve got SportVU data programmers who do tremendous work. I mean, if you understood the bandwidth that went into the process of making every decision here, you’d understand that I’m just the guy that gets to come in front of you and share the passion that they all have for what we do.”

WFNY asked for additional access to Nichols and Alamar. We were denied, informed that their work is proprietary and kept behind the scenes. The team was understandably hesitant to share further details, as most NBA squads closely guard the inner workings of their analytics department. The Cavs do not want to be seen as an analytics-first organization.

*****

“I believe that anything that has an impact on the court can be quantified. I think absolutely anything. That’s not saying it’s easy or we can measure it perfectly, but we can get a measure and an indication of how different things have impact.” – Ben Alamar in September 2013, in a video with Coach Nick (@BBallSource)

Tristan Thompson and Analytics

“If you wish to persuade me, you must think my thoughts, feel my feelings and speak my words.”

Without any hesitance, Ben Alamar controlled the room on the third floor of the Hynes Convention Center. He stood tall, with confidence, and spoke clearly about the communication problems affecting the sports analytics industry. The Cicero quote above was the beginning of his quick PowerPoint presentation.

Alamar was set to give a 30-minute presentation at the 2014 Sloan Sports Analytics Conference on the topic of procuring new data from a haystack. The title promised attendees that they would learn how to “Find Value in New Data.” Alamar, a conference veteran, clearly had the attention of the packed room, which was filled to the brim with wide-eyed stats enthusiasts and hush-hush team executives.

The session’s title ended up being a bit of a misnomer. Alamar’s presentation – which ended promptly after only about 15 minutes – ended up being one of the rare Sloan talks that focused incredibly little on analytics and, given Sloan’s glaring weaknesses, that was refreshing. Instead, he shared his four-part process inspired by old-fashioned design thinking, aimed to focus on assisting organizations with analytics.

The presentation felt like an advertisement for his own method of consulting although it wasn’t exactly new to his long-time readers. The process: Perspective, Creativity, Prototype and Iterate. The focus was on bringing useful information into the hands of an organization’s actual decision-makers, saving them time. Don’t be afraid to fail, he urged. Keep innovating. Add value.

His stated goal of sharing this tried-and-tested process, both at Sloan and in his non-stop public writings on Twitter, his website and countless other entities: End the perpetually discussed communication issues between analysts and decision-makers. This simple methodology narrowed down how to deliver a competitive advantage within the framework of an established organizational culture. It works better that way.

For Alamar has shared many times before: Analytics is not a strategy. And more pointedly, analytics is not synonymous with Moneyball, which was simply one select strategy (finding undervalued gems) that utilized analytics. The Moneyball narrative was unfairly polarizing. In the long-run, Alamar urges organizations to not just “do analytics.” His point instead: “Analytic systems can require a significant investment in tools and personnel, so it is the strategy for employing those systems within the organization that determine how successful the organization [is] with their analytics.”

Yet, uncertainty reigns. As any statistician would say, the numbers only help to explain the degree of uncertainty that exists. That’s something Alamar shared under the spotlight during a panel discussion at the 2013 Sloan conference. He shared the stage with some of the best and the brightest: his old buddy Jeff Ma, conference co-founder Daryl Morey, Cleveland Browns president Alec Scheiner, FiveThirtyEight czar Nate Silver and football analyst Phil Birnbaum.

The panel began with a discussion of the randomness of each individual’s path into the sports world. Alamar shared that he worked as the Director of Analytics for the Seattle Supersonics and Oklahoma City Thunder for over five years. Then, sometime before the 2013 Sloan conference, he joined up with the Cavaliers. That means he was finally starting to share specifics of his days working alongside his usually former boss secretive Sam Presti. This was a main detail in his well-received book released last summer, which was entitled Sports Analytics: A Guide for Coaches, Managers, and Other Decision Makers.4

In a video with Grantland’s Zach Lowe during the 2013 conference, Alamar discussed the decision-making process leading up to drafting Russell Westbrook, not Brook Lopez, with the No. 4 pick in 2008. The most fascinating line he shared: “(Centers) are the scarce resource, right? That’s what they always say. But the truth is that’s not necessarily the case and even if they were the scarce resource, that doesn’t mean they have the biggest impact.”

In his first real-world NBA Draft experience, Alamar’s research and proprietary analysis assisted with Presti’s decision to take Westbrook, despite the UCLA product’s limited experience at the point guard position. Multiple times, the consultant has emphasized: The analytics were not the reason for the pick. The numbers did help to reduce some of the uncertainty. The importance of analytics, however, was already imbedded deep within the franchise’s overall organizational strategy. They were sticking with that strategy.

*****

“Instead of attributing team success to things that don’t have much to do with the game of basketball, we should instead focus on analyzing what we do know. … Tools such as statistics are becoming more advanced and better at picking up on the little things, and in time they may be able to tell us a whole lot.” – Jon Nichols, March 2007 at NBADraft.net

*****

“Not everything that can be counted counts, and not everything that counts can be counted.”

That famous George Gallup quote was the introductory slide for the first-place Second Spectrum research paper on the “Three Dimensions of Rebounding” at the 2014 Sloan conference. Fittingly, it also is the greeting quote at Jon Nichols’ website.

That synchronization epitomizes this wonderful fact: A wannabe basketball analytics enthusiast could know very little about the current state of the industry, read all of Jon Nichols’ work, and know just enough to have a detailed conversation with 95% of the active community today. And yet, Nichols hasn’t published an article in over four years.

It’s not that the basketball analytics revolution has exactly died. It has evolved and mostly moved to the private side; the same thing happened in baseball after the Moneyball explosion, although perhaps to a less damaging extent. One only has to credit Nichols’ forward-thinking curiosity for his work standing the test of time. Because four years is practically an eternity in sports analytics.

To really get a sense of Nichols’ diverse basketball interests, one can explore in two ways: Look through his 37-article archive at the NBA blog Hardwood Paroxysm5 or scroll down his immense personal archive at his own site.

Nichols’ usual trick was to analyze published play-by-play data, mostly from Basketball Geek’s Ryan Parker, and note differences between actual per-possession statistics and the more commonly published estimates. His various other noted fields of interests: in-game shot type changes, diminishing effects of certain statistics, lineup data, defensive metrics, NCAA plus-minus stats, age correlations and numerous NBA Draft studies.

But his biggest contributions to the field came in the form of three original statistics: Composite Score, Value Rating and Position-Adjusted Classification. The three served widely purposes: From a combined NBA ranking of offensive and defensive metrics, to a salary-relative value percentage, to a scouting tool based on position-based rankings in various key identifying statistics. They’d be great resources if seen in the public eye today. Records are still available as of the end of the 2008-09 season.

More importantly, he also understood that statistics can’t explain everything. In a personal website post about the Moneyball polarization of stats against scouting and shortly after the 2009 Sloan conference, he wrote this: “But in my mind, the human element will always play a large role. To many teams, it still may factor into 95% of their decision-making. While I do think those teams are missing out on a great trend that’s now been around for a few years, there is some logic to trusting your eyes. The big thing is knowing when to use your numbers and when to use your scouts.”

Again, this is coming from a young stats innovator and researcher. Even he was saying that scouting tools will always be essential. Nichols likely further refined his nuanced understanding of the role of basketball analytics during his time as a basketball manager for the Harvard University basketball team. How did he get such a role at Harvard? That’s because he had enrolled in a master’s program in informational technology. He finished the degree in 2013.

It’s scary enough to consider completing such a degree while working full-time, let alone in the rigorous professional sports world. It’s perhaps even scarier to wonder how much better the young savant has gotten as an analyst since he was last seen in the public eye besides his semi-occasional posts on Twitter.

In Milwaukee, he was known to have assisted with their in-depth video charting system, tracking usually unmarked defensive statistics such as deflections or helps. He gave presentations to the Bucks coaching staff, led by Scott Skiles, on the team and opponents. He was made as available as possible to the coaches, the organization’s assistant GM shared with an SB Nation blog. He then joined the Cavs front office in September 2013, filling a position that had been posted publicly this past summer.

*****

“The Cavs have been one of the teams to ‘embrace analytics’ so the question is: are they ignoring the advice, or getting bad advice?” – Box Score Geeks’ Andrés Alvarez in February 2014.

*****

Ben Alamar is the skilled economics researcher who lucked into a sports-focused path and now serves as a team consultant while continuing to push the sports analytics industry forward. Jon Nichols is the young stats creator who lobbied for an NBA job, got it, and now is on his second full-time, behind-the-scenes gig. These two are the leading analytic people who you didn’t know about in the Cleveland Cavaliers organization.

Within the context of the Cavs current situation, it’s again vital to note that Alamar and Nichols had no impact whatsoever in the fated 2011 and 2012 drafts. Alamar was on board before the largely disastrous 2013 offseason, while Nichols only has been in tow since the preseason.

Overall, the Cavaliers had seven listed representatives at the 2014 Sloan conference. Of the six known names, three were in business operations and three were in basketball operations6. Griffin was not scheduled to be in attendance. Alamar led a session, per usual. Nichols easily could hide among the grade of lookalike 20-somethings.

At Sloan, as always, basketball coaches and executives shared their skepticism of using too much scripted mathematical jargon on the court. Daryl Morey shared that when he arrived in Houston, he discovered that “random” offensive sets actually were the team’s most efficient. Perhaps there is a yet-to-be-publicly known skill set related to instantaneous athletic decision-making. Is that something that the Cavs have emphasized in the past or are continuing to do so now?

That’s possibly where Krossover can help. The Cavs announced their deal with the video-indexing sports tech start-up back in November. They were listed as an early adopter last spring, with Alamar a huge proponent and a company adviser. The Cavs have reportedly used the service to easily categorize video from across the NBA and college basketball. The service instantly creates efficiency data, too.

Krossover also includes a gaming app that tests individual’s real-time basketball choices. Alamar discussed it in his video with Zach Lowe last year. The company’s founder, 5-foot-9-inch Vasu Kulkarni, a native of India, was an unheralded star of last year’s Sloan conference. Many were buzzing about the startup’s potential to help train and improve on-court decisions.

But all that Krossover can do at this point in time is assist with the players currently on roster. And boy, the Cavs sure don’t play like an analytically inspired basketball team at all. Offensively, they take the worst shots in the NBA, on average, inefficiently preferring the dreaded non-restricted area two-pointers. On defense, they allow the most three-pointers in the NBA, the opposite of what teams like Portland are focusing on each day.

Personnel-wise, the team currently lacks able long-distance shooters, wing defenders and post presences. Based on that alone, and not even recognizing their mostly forgettable play on the court in 2013-14, it’s hard to consider how “analytically” focused this organization has been or why they have received that narrative in the recent past.

Rebuilding is not easy in any professional sport. There are no guarantees. In fact, Alamar’s 2013 Sloan panel was titled “True Performance and The Science of Randomness.” Alamar recalled a time when a GM, obviously Presti, texted him about Brook Lopez beating up on his team. Shortly thereafter, he shared this caveat about using analysis within an organization’s strategy: “And baked in, there’s an understanding that there’s risk involved. There’s no certainty in the data. It’s not fact. It’s just a probability.”

So when will the storyline be about the Cavaliers, the team that get rich quick? It’s all in the odds. With brilliant minds like Ben Alamar and Jon Nichols working for the team, you should probably stand pat for now, not necessarily double down. There won’t be just one decision or one jaw-dropping hat that changes everything. But over time, assuming the organization holds the course, it’s likely the success probability will be tilted just a little bit better.

*****

Author: Jacob Rosen | Inspiration: Will Garbe | Editor, Production: Scott Sargent | Photos: AP

___________________________________________________

Footnotes:

  1. He also wrote the book behind The Social Network []
  2. The acronym stands for the Association of Professional Basketball Research. []
  3. This TrueHoop article from March 2009 was the most popular one about Rosenbaum’s days with the Cavaliers. There have been precious few public reports of Rosenbaum’s work since. []
  4. The book can be found on Amazon. []
  5. Nichols wrote all his content at Hardwood Paroxysm and the New York Times during his brief transition period from Michigan graduation in spring 2009 to joining the Bucks in January 2010. []
  6. Besides Alamar and Nichols, the other name was Taotao Zhang, the team’s Basketball Operations Assistant. []
n');
  • Ezzie Goldish

    Great piece.

  • The_Real_Shamrock

    ugh!

  • mgbode

    completely agree. most noteworthy line:

    As any statistician would say, the numbers only help to explain the degree of uncertainty that exists.

    add in the age-factor for these prospects. while age helps them have a higher ceiling in the prognostications it also adds a higher variability to where they will end up. also, that whole team/philosophy fit with prospects is as important in basketball as any sport and needs to be weighed.

    absolutely fantastic piece balancing out where we came from, what we are doing (as much as can be known), and what it might all mean. a ton of research here (obviously) and a ton that I did not previously know. loved it.

  • Harv 21

    Very nicely done, Jacob. Writing is so much better when you start by knowing the subject matter and exactly what you want to say, and obviously you did.

    My question: was bringing in Alomar and Nichols the idea of Chris Grant or Dan Gilbert? In other words, might they soon be gone before their input makes a meaningful impact?

  • Harv 21

    yeah, the age/developmental stage of the top prospects is a huge issue. Players used to develop dramatically so much in their sophomore and junior years, or stall. Tristan is a good example: it’s so tough to predict that a good, energetic kid willing to work hard enough to change his dominant shooting hand will not be able to develop any real scoring move or sustain his game effort throughout a full season. They might have seen that lack of development as they scouted him a few more years. And while those things yet might happen, they have to make a decision on how much to invest on his future as his rookie contract ends. Seems to me that, except for the few transcendent talents, the NBA draft will turn into more of a crap shoot than ever. And GMs will get canned because of it.

  • steve-o

    So basically the current incarnation of the Cavs is the result of a failed math experiment.

  • Pat Leonard

    “while age helps them have a higher ceiling in the prognostications it also adds a higher variability to where they will end up.”

    It’s funny how this seems so much more readily understood by baseball fans than by basketball fans.

  • Steve

    My guess is that higher upside is so much more appreciated in basketball. Get that high ceiling player, and you have the most important piece for the next 5+ years. He busts? Well you’re back at the top of the lottery, with a great chance to get a high ceiling player again. You add some mid range talent, and you’re looking at around 40 wins, not a true contender, but not bad enough to pick at the top of the lottery, and you’re stuck in a rut.

    In baseball, obviously high upside still is a big plus, but you can be the Indians and win 92 games with your best player as Jason Kipnis if you put enough mid range talent on the rest of the roster.

  • Cynic

    LOL … I know there were alot of words up there, steve-o. Even though that makes it seem scary, I know you can read them all if you really try! If there’s some that seem hard, just try to sound them out, ok buddy?

  • mgbode

    I agree, but also add in that those young players spend 4-6 years in the minors even if they are really, really good. so, the delayed gratification helps temper the enthusiasm in baseball as well IMO.

  • Steve

    Absolutely agree with that.

  • Chief Blahoo

    Interesting article; there’s so many people in front offices and on coaching staffs that look askance at analytics, as if using stats precludes them from factoring in things that aren’t stats. Like intangibles, chemistry, etc. Of course you can’t look at a box score and measure a team’s chemistry, but you measure what you can and learn as much about it as possible. So i’m glad that the Cavs have embraced the analytics movement, despite the record it gives me a measure of hope for the future. Even if Mike Brown is our coach.

  • WFNYJacob

    Thanks Ezzie. Appreciate that.

  • WFNYJacob

    Thanks bud. A lot went into this. It’s been the main reason for my recent quietness on the site.

  • mgbode

    time well spent