Dean Oliver of the Denver Nuggets and Kevin Pritchard of the Portland Trail Blazers participate in the 2010 Sloan Sports Analytics Conference's Basketball Analytics panel. Photos: Ben Golliver | Blazersedge
Disclaimer: The following is 4,000+ words about a sports analytics conference. Buckle up.
The Sloan Sports Analytics Conference was, appropriately, held in an archaic city's sterile Convention Center. While the historic Massachusetts State House, built in 1798 on the heels of the American Revolution, loomed within walking distance, Sloan attendees -- sports obsessives and mathematical eggheads -- were exclusively concerned with the future. Modern buildings crop up alongside historic ones, new ideas replace obsolete ones.
Sloan was exactly what you might expect from a group of MIT geniuses who are both passionate about the subject and uber-motivated to get their foot in the industry's door. The day began early and it ran late. It was, from start to finish, ruthlessly efficient, pioneeringly techy, overwhelmingly substantive, and low on aesthetic frills, a laboratory for ideas about how to make sports teams and organizations better through the study of numbers.
Sloan's main conference room featured a single stage with a slightly corny locker room backdrop and five simple chairs. The stage was flanked by two gigantic projection screens that served the wonky audience of more than 1,000, many of whom took notes and tweeted diligently. The minds and the words were the center of attention; insanely quick witted one-liners provided the fireworks, their flash magnified by the lack of distractions. Although there was a corporate presence -- ESPN and EA Sports were major sponsors -- the gross commercialism that you might associate with the average trade show or industry conference was nonexistent. If you momentarily stopped paying attention to a speaker's point, it wasn't because a large neon sign was blinking madly nearby, it was because Bill Simmons quietly snuck in the back door and took a seat two rows behind you.
MIT volunteers and attendees from previous years marveled at the conference's growing size. "We couldn't even fill a classroom three years ago," one staffer joked. But as a first-timer expecting lines out the door and swarms of people, I was pleasantly surprised by the event's intimacy and its attendees' collective restraint. How often do NBA General Managers and bloggers sit side-by-side snacking on boxed lunches? How often does Deputy Commissioner Adam Silver give out his email address to a packed room and then spend at least 20 minutes addressing individual questions in a small group? An average scene of NBA life might find a player deluged by dozens of autograph seekers or a General Manager surrounded by 20 microphones, eventually a numb sets in and despondency results. Here, an NBA owner like Mark Cuban matched the excitement level of the basketball junkies that hung on his every word.
There remained a definite pecking order, in descending order of importance: Sloan Founder and Rockets GM Daryl Morey, best-selling author Michael Lewis, ESPN's Sports Guy Bill Simmons, the panelists that were in management, the panelists that were media members, regular media members, and, finally, attendees. But the gap between those last two groups and everyone else was about as narrow as you will find anywhere. At Sloan, access to the NBA's decision-makers and trend-setters, the sharp minds that are leading a statistical revolution that is forcing old-school, ex-player NBA executives to adapt or relinquish control, was as good as it gets. The Brain Trust of the Portland Trail Blazers, led by General Manager Kevin Pritchard, took up their position near Sloan's front lines.
Click through for a full account of Kevin Pritchard's appearance on Sloan's Basketball Analytics panel, the massive Blazers presence at Sloan, how and when the Blazers use statistical and video analysis to guide their decisions, the formalized processes that combine to create winning organizations, the Blazersedge reader who might represent the future of basketball and a whole lot more in between.
-- Ben Golliver | firstname.lastname@example.org | Twitter
From Exhausted to Enthusiastic
It was here at Sloan that Portland Trail Blazers General Manager Kevin Pritchard looked as comfortable as I've ever seen him. Pritchard has had a tough year. Ironically, his year has been derailed by, more than anything else, the one thing that eludes even the most advanced thinking by the smartest minds in the basketball analytics community: injury.
Since Greg Oden went down with a fractured patella in December, Pritchard hasn't always seemed like himself. Immediately after Oden's injury, he was near tears. In interviews and press conferences, he became increasingly laconic. He was, more often than I remember during prior seasons, absent. When Greg Oden faced the media regarding embarrassing naked photographs, he sat alone on a stool; Pritchard was nowhere to be seen, although he conducted interviews by telephone. When a low-key press conference was held after Brandon Roy was selected as an All Star, Roy sat alone and introduced himself while Pritchard watched from way in the back, wearing a turtleneck, chatting cordially but briefly with a writer or two. When Sergio Rodriguez popped off to the Spanish media last year, Pritchard was out in front, arm around Sergio for the photo op, projecting confidence and establishing control of the situation and story. When Rudy Fernandez did something similar in February, Pritchard seemed to barely acknowledge the situation, mostly leaving Head Coach Nate McMillan and Fernandez's improved play to answer the questions. The 82 game NBA season can be a drag but, in prior years, Pritchard's relentless optimism carried himself, his organization and, often, the media and fans along through the inevitable mid-season drudgery. That happened less this season.
None of the above is meant as criticism. It is clear where Pritchard has been: on his Blackberry, in a meeting or staring at a computer screen with the intention of improving his depleted basketball team. Hardship exemptions were applied for. D-League prospects were worked out, tested, called up, sent down. A free agent was signed, waived and then re-signed. Surgeries and rehabilitations were scheduled and updated. Out-of-state medical consultations were arranged and cancelled. And, of course, trades were explored and one was consummated.
When the roster juggling was mostly completed -- when those who were going to be healthy had returned and the trade deadline had passed -- Pritchard found his team set on the course he had hoped for. After acquiring a difference-making, veteran Center and virtually assuring a playoff spot, Pritchard presented Marcus Camby to the media and then stood wearily in a Rose Garden back hallway taking questions. I had a full list of topics prepared, saved up for a week or more: Steve Blake's year-over-year deterioration in statistical output and consistency, whether Blake or Travis Outlaw would be a summer free agent target, whether it might be a good idea for the Blazers to aggressively court Camby for next season as he can play both the 4 and the 5 and fill out a fearsome frontcourt rotation. I made it through only the first Blake question. As is typical with Pritchard, he was making consistent eye contact and showing his emotions. Both his eyes and his slumped shoulders seemed to say, "I'm exhausted." It felt akin to urgently asking a marathon champion, hands on knees sucking wind, "How do you feel?" I switched off the recorder. There would be other days.
The Pritchard who attended and participated in the Basketball Analytics panel at Sloan couldn't have provided a greater contrast. Relaxed, loose, energetic and confident, Pritchard was eager to both share and learn. So eager, perhaps, that he and Vice President of Basketball Operations Tom Penn engaged in a little "Good Cop, Bad Cop" routine over how much of the Blazers' proprietary information Pritchard might share when he took the stage. Penn, as always, seemed to be urging discretion. Pritchard was also noticeably tan, especially compared to the pasty assemblage of conference attendees. Splayed out in a chair an hour or so before his panel, he cracked jokes with members of upper management from opposing teams and looked, more than anything else, really happy. He had every reason to be.
Rolling Deep, Opening Up
For starters, Pritchard was in familiar company. Not only was this his fourth Sloan Conference but, this time, he was rolling deep. By my count the Blazers had more healthy bodies in Boston -- Pritchard, Penn, Head of Scouting Mike Born, Ben Faulk, Sean Kiely, Justin Kubatko, Jeff Ma and Ryan Parker -- than they had suit up for multiple games in December. "You're the one hiring everyone, Kevin," Cuban cracked during the panel. "That's true," Pritchard replied with a grin. "It's a little bit of an arms race." When you consider the resources required to make that presence possible -- salaries, flights, hotels, meals, conference fees, etc. -- you quickly realize that any writer trying to sell the idea that the Blazers don't treat statistical analytics and video charting as foundational elements for making both in-season lineup, playing time and strategy adjustments and off-season personnel moves simply doesn't know what he is talking about. Period.
Second, and perhaps more important: it became quite clear on Saturday that advanced statistical analysis is a passion of Pritchard's, far more so than was previously apparent. "As a player I liked it," he said during the Basketball Analytics panel. "As I got into coaching in the minor leagues and the NBA, I really liked it. I tried to use it as much as I could." Really, it was hard for any stats geek or developing stats geek not to feel an instant kinship with Pritchard. During his panel, he bemoaned the traditional box score as "incomplete" and stated with absolute certainty that "there will never be a uniform language for adjusted plus/minus." He amiably chatted about a recent John Hollinger column. And no moment prompted a more forceful double-take than when Pritchard casually referenced his "quant," geek short hand for the organization's quantitative analysis of advanced statistical metrics. I'll take "quant" over "calm waters" every day of the week.
Pritchard's evident comfort level led to some truly laugh out loud moments. When discussing Synergy Sports, the video and statistic logging service that I've referenced many times this year thanks to The Invisible Ninja, Pritchard joked, "It's phenomenal. I'm so distraught that Mark [Cuban] owns that because I know there's a way for him to see what I'm watching." The mental image of Cuban peering over Pritchard's shoulder while he rewinds plays had the room cutting up. Later, Pritchard dropped an elegant double entendre when discussing Blazers Owner Paul Allen's interest and support of the franchise's statistical analysis. "My owner is into data," Pritchard said with a smile. "He's like the data king of the world." Of course Allen, who was "into data" enough to help found Microsoft, would be interested in quantifying the activities that occur when his prized basketball team takes the court.
One of the most memorable exchanges came when an audience member asked for the panel's thoughts on building around a "Big Three" to form a championship team. The Big Three's simplistic, marketing-style phrasing drew laughter from the stats-heavy crowd even before any of the panelists had a chance to respond. So it came as no surprise when Hollinger directly attacked the premise of the question, mercilessly listing off recent NBA champions that hadn't relied on three strong players. His rat-a-tat, encyclopedic delivery of counterexamples -- the 1990s Rockets, the early 2000s Lakers, the "Big Zero" of the Detroit Pistons - might have continued forever had Pritchard not gazed over at Hollinger and slyly interjected, "Can you be a little more specific?" Hollinger took a breath and grinned. Both his calm and the pride of the inquirer were restored instantly.
And finally, near the end of the panel discussion, Pritchard tossed his empty water bottle in mock frustration over receiving so many text messages from ESPN writers regarding his trade plans over the past month. If the young, engaged crowd wasn't able to drink from the empty bottle, they were happy to settle for eating out of his hand.
Speaking the Same Stats Language
The panel, however, was about far more than jokes and laughs. Pritchard, Cuban, Hollinger, Denver Nuggets executive Dean Oliver and Boston Celtics executive Mike Zarren addressed a wide variety of topics, beginning with one I raised earlier this year: How do Blazers coaches and players incorporate the study of advanced statistics on a daily basis?
You might remember Nate McMillan told me that advanced statistics are something that he has access to but that he prefers breaking down video tape. When asked whether he used stats to guide offseason personnel moves or in-season adjustments, Pritchard leaned more heavily towards the former. But he did say, "I do believe there's huge value in applying [stats] to the coaching staff" and he spoke of "educating coaches" regarding the tendencies of their own players and opponents.
As for the players, I've mentioned this year that both Brandon Roy and Nicolas Batum seem particularly interested in the pregame scouting reports prepared for them by the team. Discussing that interest, Pritchard said that it varied from player to player but that the biggest concern is ensuring that management, coaching staff and players are all on the same page when it comes to the potential offensive and defensive adjustments that might be suggested by the numbers. "I think it's all individual dependent. Some players love as much data as they can get," Pritchard said. "The data is available, our assistant coaches know it. The biggest key going forward is making sure everyone is speaking on the same level." This thought echoed what a number of other executives and observers -- including Simmons -- mentioned throughout the day: that statistical analysis only has value if it is communicated cleanly from source to players.
An interesting and overlooked hurdle in this communication process that Pritchard raised was the year-to-year turnover of the coaching staff. "Every 3-5 years you have a different coach," Pritchard said, apparently speaking generally about the NBA. Perhaps he forgot that Nate McMillan is currently in year 5 of his tenure in Portland. It's probably best to give him the benefit of the doubt although the statement immediately drew some quiet chuckles.
Pritchard was also careful to note that there are limitations to the best in-game analysis and that advanced numbers are far less valuable in certain game situations than others. Two examples Pritchard cited were the Utah Jazz's overall offensive scheme and the Dallas Mavericks' late-game offense with Dirk Nowitzki and the high pick-and-roll (a setup that I video charted here). Pritchard's point: a lot of times in the NBA you know what's coming, you just can't stop it and all the numbers in the world won't help. And, of course, lady luck always looms. "It can come down to one play. If Courtney [Lee] makes that layup [in the NBA Finals]. What happens?," Pritchard wondered aloud.
Formalizing Processes for Success
Game-to-game adjustments seem much less important to Pritchard than mastering team building and the processes that go into that: scouting, drafting, trading, and managing the salary cap. Each of these processes, it seems, has been standardized, refined and is constantly reanalyzed for potential improvements.
In a panel earlier in the day, Maurizio Gherardini, the Senior Vice President of Basketball Operations for the Toronto Raptors and architect of the Benetton Treviso basketball dynasty in Italy, stated that globalized scouting was the single greatest key to his success in Italy. He described expending significant resources to scour the globe for new talent and how the efforts to acquire improved personnel paid almost immediate dividends. Without question, the Blazers have taken up this best practice for organizational growth: they have a full-time scout stationed overseas and multiple scouts crisscrossing the United States on a regular basis.
According to a member of the management team, the Blazers now regularly supplement that in-person scouting with a review of a player's statistics and, if possible, video made available by Synergy as well. "Eyes, ears and stats," the management member stated, in summarizing the team's current approach. In other words, the team scouts each prospect in person multiple times, consults with coaches and trusted sources to find out more about the player's background and mental makeup, and then reviews the player's numerical output and development. Only after all three factors have been considered are decisions made.
Look no further than Blazers rookie Dante Cunningham to see the benefits of this systematized process. The organization scouted him in person multiple times and brought him in for a pre-draft workout last summer, which they unanimously praised. Members of the organization consulted with Cunningham's college coaching staff to get a full picture of his basketball intelligence, team defensive abilities and family background. And, finally, they recognized distinct year-to-year improvement in his numbers as he progressed through 4 years at Villanova. At the end of the day, Cunningham aced all three aspects of the "eyes, ears and stats" test. While their scouts didn't predict an all star impact and still had questions about exactly what position Cunningham might play in the NBA, they were more than convinced that he had enough tools and intelligence to make an impact sooner rather than later. Less than a season into Cunningham's career he has proven to be an able member of the team's rotation, standing out on defense and consistently knocking down a face-up jumper. The team's evaluation has played out as well as they might reasonably hope for a player taken in the draft's second round.
Of all the panelists, Pritchard seemed to place the most value on the draft process as a way to improve his team. "The draft has been very important for us," he stated emphatically. The main reason? Rookie scale contracts. Because rookies are paid according to a specific, relatively modest scale, there is a great benefit to having a number of contributors on their rookie deals simultaneously. As I mentioned yesterday, the Blazers have 5 members of their rotation -- Cunningham, Jeff Pendergraph, Jerryd Bayless, Rudy Fernandez and Nicolas Batum -- who will combine to make less than a single veteran, Joel Przybilla, who might not actually suit up at all due to injury. The combined impact of those players, both on the court and on the books, is almost immeasurable.
It might be helpful to view players on their rookie deals as the prototypical "undervalued asset" from Michael Lewis's Moneyball. In that book, Lewis writes that 10 years ago the Major League Baseball market generally undervalued the ability to draw walks. Players that could regularly draw walks, get on base and eventually be driven in by their teammates were significantly easier and cheaper to acquire than their impact on wins and losses suggested they should be. In the current NBA market, where a team might sell off a second round draft pick and think nothing of it, that draft pick and the salary owed to it has become undervalued. If you can find a rookie who is ready to contribute immediately and is locked into a certain salary structure, his contributions per dollar spent will almost certainly exceed an average NBA veteran who requires a substantially higher salary. "Sometimes you have a great talent but it doesn't make sense with the contract," Pritchard lamented. This helps explain why the team has opted to pass on veteran acquisitions in the past and instead honed in on NBA-ready, 4-year college players such as Cunningham and Pendergraph when filling out its roster.
That search for value has guided the Blazers' recent free agency and trade moves too. Put simply, the Blazers are engaging in a form of risk assessment: The more a potential acquisition might cost, the more potential damage he can do to the team's salary cap flexibility. Mark Cuban praised the Blazers moves over the last few years, stating simply, "Who you are not picking up is as telling as who you are picking up." Cuban's statement brought to mind this chart I created earlier this season, tracking the performance of various summer targets relative to their salaries.
The Blazers had been mentioned in rumors regarding high-dollar but inefficient and underperforming players like Richard Jefferson and Vince Carter but have found better value in players like Andre Miller and Juwan Howard. Management applied that same "eyes, ears and stats" approach to both players and remained confident through Miller's early season struggles that he would come on strong as the year progressed, thanks in no small part to the stats and analytics component.
The same thing can be said for the recent Marcus Camby trade. With a solid knowledge base on Camby assembled during his long NBA career, a member of the management team said the Blazers were able to turn to videotape breakdowns of Camby's recent play as a Clipper, honing in on specific situations to clarify their understanding of his unique skillset. They were able to watch looped tape of Camby's tipped out offensive rebounds or weakside blocked shots to reinforce their previous scouting.
Following Best Practices; Reaffirming Personal Practices
As the panel began to draw to a close, Pritchard repeated two points made earlier in the day by Bill Polian, the highly-respected architect and President of the Indianapolis Colts. First, that statistical analysis can be very effective in reinforcing an organizational philosophy. Polian pointed out that he might evaluate a nose tackle totally differently than the New England Patriots and that the two teams' systems were so different and complicated that trying to swap players from one team to another would be an almost meaningless exercise. "We all see certain [statistical] things that are very important to us. It may be different for Denver," Pritchard repeated. "If we're about A, B,C, D and E [statistically] and those things trickle down to what we do [on the court as a team and off the court as an organization], that helps us have the best data. It doesn't mean that's the best data for Denver too. It just means that data is best for us."
Second, that statistical analysis cannot guarantee victories but it can go a long way to establishing a winning organization. Polian stated the best talent evaluators in the NFL really only succeed about 55% of the time despite all of the elaborate measurement techniques they use. Pritchard stated that he believed the success rate was similar in the NBA and that building a team built for long-term success versus building a title team was comparable to playing the stock market. "It's sort of like the S&P 500. You can beat the S&P 500 every year but if you're trying to be #1 in our sport, you have to take huge risks."
There have been times this year when I have doubted whether the Blazers were willing to take those risks and whether they were content with putting a lower-tier playoff team on the floor. Those questions were answered for me initially when the team completed the Marcus Camby trade, cutting loose two players -- Blake and Outlaw -- that had been community mainstays. "Whether it's a star player or the 13th or 14th man, it's hard to trade a player away," Pritchard flatly admitted to the audience. And those questions were firmly put to bed after seeing and hearing first-hand how meticulously Pritchard and his staff approach their organizational structure, how they have listened and openly borrowed ideas from champions like Polian. Asked what he had enjoyed most about this year's conference, a member of Blazers management said simply, "the reaffirmation." Meaning: We work hard and think we're doing things the right way and the exchange of ideas and practices at this conference has helped confirm that.
It's more than fair for Blazers fans to feel a similar reaffirmation. TrueHoop's Kevin Arnovitz wrote yesterday, "If I'm a fan, I want my team to be one of those 16 in attendance [at Sloan]. That would demonstrate a commitment not only to smart thinking, but to winning. " Not only were the Blazers in attendance, they were at or near the head of the pack. Injuries and bad luck will happen; Intelligence, determination and zeal for improvement are the best recipe to overcome them. This Blazers brain trust is not lacking in any of these attributes. And it probably hasn't received the credit it deserves.
As the conference closed, the one-day vacation and celebration over almost before it started, the inevitable question of "What's next?" arose. For the Blazers brain trust, the immediate answer was a return to the college scouting beat in preparation for this year's draft. Some members of the organization would travel to the Big East tournament while others would attend the Big 12 tournament.
The longer term answer, though, weighed more heavily. "What's next?" for Pritchard and company will very likely be an NBA world that is starting to catch on, a league that increasingly turns to numbers and advanced statistics to inform personnel and coaching decisions. The competition at Sloan increased greatly this year -- more than half of the league's 30 teams were present, a sharp increase over last year -- and there is no sign of it slowing down. The brains are coming.
As curators and devotees of this sport, we should, all of us, embrace this trend towards the increased analysis of statistics and videotape. It has made decision-makers smarter and coaches more prepared. It has, over time, made the game itself more efficient and, dare I say, better.
As I prepared to leave Boston's modern, sterile Convention Center and step again into its old cold, I spoke briefly with Douglas Hwang, an MIT graduate student and one of Sloan's organizers. Hwang happens to be a Blazersedge reader and suggested that I attend this year's conference way back in December. I asked Hwang, who spent 4 years as a jet engine engineer after obtaining his undergraduate degree and spent hundreds of hours organizing this year's conference, what was next for him now that Sloan was over. Hwang just shook his head and smiled at the question. "I do n't know," he finally concluded.
Limits on what's next for Hwang and those similarly intelligent and passionate about the analytics of sports are evaporating. Should he desire a career in basketball, Hwang could become a team consultant, a video analyst and, perhaps, even a General Manager like his fellow MIT alum Daryl Morey. This thought would have been an impossibility no more than 15 years ago. But if the growth of basketball analytics is a revolution and its most fervent revolutionaries are engaged in an arms race, who better to bring along than a jet engine engineer?
-- Ben Golliver | email@example.com | Twitter