By: Edward Egros

conference

2017 Sloan Sports Analytics Conference

Pasted Graphic
Another installment of the Sloan Sports Analytics Conference has come and gone. More than 3,500 were estimated attending the proceedings, learning and offering their latest research in the sports analytics world. While football and basketball are often the most popular sports here, there seemed to be a noticeable effort to highlight the quantitative strides made in other sports.

One panel featured golf analytics, led by Golfweek's
David Dusek, who highlighted the success stories of these quantitative tools. Jeff Price, Chief Commercial Officer of the PGA, offered an example of Team USA at the Ryder Cup. At Hazeltine National Golf Club, long par 5's meant emphasizing wedge play. It's this discovery that helped the Stars and Stripes to a decisive 17-11 victory.

On a more personal level, current professional golfer Jason Gore explained how to turn research into actionable results.

"When I talked to a sports psychologist, Fred Astaire would [practice and] put chalk on the floor," said Gore. "But once he grabbed Ginger's hand, he never thought about the chalk on the floor."

There is still room for growth.

"We're in the first inning of the data revolution in golf," said
Arccos Golf CEO Sal Syed.

Dusek pointed out some major tournaments like the Masters and United States Open still do not provide the media with advanced statistics.
15th Club CEO Blake Wooster says the potential is there to analyze how golfers perform under pressure. Lastly, the group seemed to agree lasers should be used to measure distance more accurately. Even Gore believed lasers used by caddies could speed up pace of play.

Pasted Graphic 1

The seminal football panel of this year's conference was unabashedly endorsing the concepts of its own sport's analytic revolution. It was even subtitled "Please Stop Punting", a concept where going for it on 4th down
yields more expected points and discounts a more traditional idea valuing field position.

Almost immediately, Baltimore Ravens offensive lineman John Urschel, who is pursuing a Ph.D. in mathematics from MIT, discussed a common situation he says coaches get wrong. When a team trails by 14 late in a game and score a touchdown, he says it is better to go for two than kick the extra point. The reasoning is, two points essentially give you the win with another touchdown, but even if unsuccessful, you can go for two again and achieve a tie, and because most teams convert two-point attempts 50% of the time, you are at least giving yourself a better chance at winning, with a small chance at needing a third score of some kind.

Mike Lombardi, former football executive and current analyst for Fox Sports, says analytics help with time allocation throughout the week, knowing what coaches should communicate with players and which statistics are important in determining the outcome of a game, such as 3rd down red zone defense.

"You don't establish the run, you establish the lead," said Lombardi. "Teams with the lead at halftime frequently go on to win," citing last year's Super Bowl champion Patriots as the top team with wins after leading at halftime, then citing the second-place team, the Super Bowl runner-up Falcons. The players, which included former Patriot Tedy Bruschi, explained how halftime is all about adjustments, but that they should take fewer than five minutes to implement.

From a front office perspective, analytics can help decipher if trading players and draft picks make fiscal and qualitative sense.

"The toughest thing to do in sports is to know what you're trading. It's why the Patriots won't trade [backup quarterback] Jimmy Garoppolo," said Lombardi.

Football discussion was not confined just to that panel. A couple of talks featured fantasy football and if there are things to give analytic players an advantage. Here are some tips from Tauhid Zaman, the KDD Career Development Professor in Communications and Technology at MIT, and Renee Miller, a neuroscientist at the University of Rochester:

- When picking a quarterback, get one or two of his receivers as well.

- Avoid players who cancel each other out, like a defense against one of your offensive players.

- We weigh football players' performances at the start of the year too heavily. Instead, looking at the bigger picture of their performance.

- Be careful of overconfidence: "The more data we have, the more confident we become in our decision making."

Pasted Graphic 2

Lastly, in basketball, while guys like Luis Scola seemed to get most of the attention from hoops fans, maybe the most direct knowledge given came from Seth Partnow, Director of Basketball Research for the Milwaukee Bucks. In his talk, "Truths and Myths of the Three Point Revolution in Basketball," Partnow offered the following bulletpoints:

- Defensive three-point shooting percentage is a useless stat because of the noise involved (good defenses prevent the shot).

- Long range shots in the NBA do not lead to fast breaks, it's shots around the rim that cause these.

- Ten of the last 12 NBA champions ranked in the top ten in three-point shooting.

As robust as this research might be, it does not offer a glimpse into the future of basketball analytics. However, one panel discussed solely how the sport will evolve thanks to quantitative tools. There may still be blowback from coaches and those who approach the sport more traditionally.

"When you're working with the [NBA] Draft…you end up trying to convince coaches," said Dean Oliver, a statistician who worked in the front offices of the Sacramento Kings, Seattle SuperSonics and Denver Nuggets. "You don't expect to win 100% of these arguments and that's fine."

Using analytics, a couple of panelists offered simple suggestions for improving the game. Former NBA player and coach Vinny Del Negro wants the league to add a fourth referee because the pace of the game has gone up and it is getting tougher for officials to keep up. WNBA point guard Sue Bird wants to get rid of the shootaround because of the rest players need and the lack of proof shooters develop a rhythm because of this routine. She also wants the analytics to assist in the psychology of a team.

"If I were a general manager, I'd want to know if [players] retain information well and how they handle things under pressure," said Bird.

The flexibility of these tools, spanning different sports and perhaps different fields of expertise, perhaps proves why this conference has lasted as long as it has.

Pasted Graphic 3

(All photos courtesy of Sloan Sports Analytics Conference).

A Recap of the 2016 MIT Sloan Sports Analytics Conference

Sloan 1For the 10th time, sports analytics enthusiasts of all kinds came to Boston to attend the annual MIT Sloan Sports Analytics Conference. I was one of close to 4,000 attendees, though this was my first. Coaches, general managers, players, journalists, academics and just about anyone else in-between gave their takes on the industry and shared their research to the masses.
The following stream-of-consciousness features the panels I attended and some of my bigger observations.


Sloan 2

The War on Analytics

Goose Gossage isn’t the only one profanely fighting analytics. If you believe some of the speakers at the MIT Sloan Sports Analytics Conference, there exists a countermovement to the quantitative revolution.

Perhaps it was most appropriate the 10
th anniversary of this meeting began with a “Moneyball Reunion” panel, including the author of “Moneyball” Michael Lewis, the Godfather of sabermetrics Bill James and an assistant for the Oakland A’s, Paul DePodesta. That team’s general manager, Billy Beane, found a reason for using analytics when scouting players.

“Billy used to tell our scouts…’I have all of this experience’”, said DePodesta, referring to Beane’s 25 years of working in some capacity in Major League Baseball. “I can’t walk into a high school game and say ‘This guy is going to be a star.’ If I can’t do it, I don’t know how anyone can do it…we have to come up with a different way,” said DePodesta.

The team combated old school thinking by finding players who were devalued in some way by others. Sometimes it was due to their physical stature. Lewis recalled the story of the A’s considering Alabama catcher Jeremy Brown, who many considered overweight: “He’s so fat, his thighs would rub together and set his jeans on fire.”

These stories happened more than a decade ago. Just like analytics, the criticisms and concerns have evolved. The second panel of the day focused on basketball and featured former NBA forward Shane Battier. He originally resisted analytics for a more personal reason.

Teams can quantitatively gauge a player’s health when it comes to sleeping habits, nutrition, etc. On the surface, it seems franchises would only need to know this information to maximize a player’s health, thereby making him/her more effective. But Battier’s concern was that teams would find some data to devalue him and have reason to pay him less and/or offer fewer years on a contract.

“It’s called capitalism,” said Battier.

Personal reasons or otherwise, Battier does believe there is a stigma within NBA locker rooms about what he called, “the math”. Though he claims it extended his career as he aged, it’s “still not cool to be hip to the math”. He did add if a player found analytics to be useful, they might find subtle ways to learn to how to improve.

The conflict between believers and non-believers rages on. Safe to say this conference preaches to the choir. When asked about Goose Gossage’s comments that baseball is now run by nerds, Bill James’s response received one of the louder ovations of the morning: “Back in 2002, you had to pay attention those guys. Now, you can just ignore them.”

Sloan 3

Talking About Playoffs

Taking a personal tone with this blog entry, one of the more interesting panel discussions of the day involved playoff analytics. Specifically, how do we devise the best system for determining a champion for each respective sport? It’s a philosophical question as much as it is analytical because leagues could simply have one-game championships for every sport; and though it would be exciting, it would also be inherently unfair for teams that would win a series but lose the opener.

Each sport has its own set of challenges. While the NFL cannot play as many games as other professional leagues, college athletics must deal with other factors. NCAA executive Oliver Luck points to class time, money for travel and time commitments that, if abused, would be unrealistic for student-athletes.

However, at the forefront of these conversations is attracting the most loyal fans. They may not want to see a nine-game World Series (something I have argued for) because it is too long to retain interest. Nine games might be a truer way of determining the best team in a series—especially with expanded starting rotations—but in the end it is what the fans want, and that is something analytics can help with. NASCAR Vice President of Strategic Development Eric Nyquist pointed to how analytics helped his sport redo the Chase for the Sprint Cup so that a champion is not already determined by season’s end but it is not entirely haphazard as to who earns honors as the top driver.

Playoffs can also have other benefits when done correctly. Luck said the College Football Playoff has helped teams schedule more competitive non-conference games. It has also helped college basketball in spotlighting conference tournaments and conference games (though admits non-conference games could be more popular than they are).

This panel also agreed on an underlying truth that analytics highlights: there are many more games that would have to be played in all sports to determine the best team, at least thousands. Because this notion is unachievable, the next best thing is to come up the playoff format in the sport’s best interest. Who does it best? Neil Paine of fivethirtyeight.com says the NFL because it preserves uncertainty but the winner is often in the conversation of one of the top teams that season. The NBA, meanwhile, has too much certainty and only a handful of teams, if that, have a chance at a championship.

It would be ponderous for me to go through each sport and say whether I think they conduct playoffs properly. I also understand why uncertainty must exist to keep fans interested so there are fewer things to point to that would dissuade fans from following the playoffs. Still, I would hope leagues avoid caving too much to all of the whims of fans and perhaps provide a product that is fairer to the teams competing for championships than those rooting for them. I have found it is in the long-term best interest of a sport to maintain an unaffected, traditional system and not make determining a champion seem so capricious.

As a postscript, I found professional bowling to be the worst in determining a champion. In the tournaments I covered, early rounds would be a matchup of two bowlers in a best-of-seven series of matches, but once you reach the final rounds—which are televised—it is one match determining who advances and who wins the whole thing. To prove my point, I would like to believe this is why the sport is not as popular as it once was. I am probably mistaken, and if you are adamantly opposed to this idea, might I suggest a winner-take-all debate.

Sloan 4

Evolution of Sports Journalism

Of all of the panels at this conference, this was the one I was most looking forward to (surprising, isn’t it?). While it took a circuitous route to discussing sports analytics, it was a journey worth taking. For you young journalists, pay attention closely.

One of the more dominant voices on the panel was Jaymee Messler’s, President of the Players’ Tribune. Her company describes itself as “a new media company that provides athletes with a platform to connect directly with their fans, in their own words”. Founder Derek Jeter says he hopes the site will “transform how athletes and newsmakers share information”.

“We’re not following the news cycle,” said Messler. “We complement the media really well…driving stories that are compelling and are not getting covered by the [traditional] media.”

Here’s how it works: an athlete has a message they want to deliver. The Players’ Tribune offers a platform replete with resources to make sure it is exactly what they want to say. While traditional media might lose the ability to break the story, they gain material for questions the next opportunity they have for an interview.

The criticism involves the last part of this sequence. Why would the athlete grant an interview? Why would they talk about something if they feel everything about it has already been said? If they spend less time with reporters and more with the tribune, how do you build trust? (
My thesis alluded to many of these problems).

“The barrier to entry is zero,” said David Dusek of Golfweek. “You can, with a few clicks, get your voice out there…the players are much more controlling in that way and they have a way to react directly to fans (sometimes the media) and to have their voice heard…it’s interesting to see how it’s becoming more challenging.”

Reporters already had challenges talking to athletes before the Players’ Tribune thanks to athletes’ social media accounts. They already have a way to communicate to the public so a reporter may seem like a middleman. Traditional media also has to compete with new media that can provide scores and highlights more quickly than they can present. Lastly, clichés have become even more tired than ever.

What’s a reporter to do? One solution: analytics.

“Analytics is just one avenue to get a creative solution around limited access,” said Carl Bialik of fivethirtyeight.com. “We do want to talk to people in the sports world about what we find…some of the best interviews I’ve had are with people who are rarely asked about certain things.” These things include data trends, advanced statistics and specific forecasts.

Not all reporters can (and perhaps should) research their own analytics. It may not even be the unique route they should take to become more creative. What matters here are the conflicting forces that make the journalist’s job more challenging. Fortunately, there are solutions, hence the evolution.

Sloan 5

Conclusions

Virtually every hour of this two-day event, there are six different panels and lectures to choose from. I attended as much as I could while still covering the event and was not present for 49 different events, and that was just on Friday. That’s not to mention the many sports science exhibits, software presentations and other technological displays I was unable to see readily.

Perhaps one of the things that has attracted more than 3,000 people to this conference is the depth of sports analytics presented. Poster presentations and white papers are available for the deeply analytical. Other events like panels speak of analytics in broader, general terms. Even if a sports fan only wants to see players and coaches discuss their craft, there is a place for that person too. There is also a variety of subjects covered, from business analytics to athletic performance measurements to sports journalism and even to the future of how we will watch and listen to games.

While sports like football, hockey and soccer were covered, there were not as many baseball presentations as one might expect. Analytics have progressed more within that sport than any other. One reason might be a national sabermetric conference happening the same week in almost the other end of the country. It is also Spring Training with many MLB teams preparing for the season. Still, it might be a positive development for sports analytics to stress other sports so it can branch out and attract different fans. On at least two occasions, panels discussed how the NBA and basketball have the most room to grow internationally in terms of popularity.

The conference also took on developing stories. The Steph Curry phenomenon of making so many lengthy basketball shots had its share of supporters. Away from sports, Nate Silver of fivethirtyeight.com updated his political findings of who will be the major party nominees for President. Even conversations I had with presenters and attendees involved sports stories happening in the moment.

If analytics do not whet your appetite, this conference may not change your mind. After all, the pro-analytical comments were often received with at least some fanfare, a kind of “preaching to the choir”. For anyone who does have the slightest interest in sports analytics, chances are there will be at least one lecture or exhibit that will make for an informative weekend.


(All photos courtesy of MIT Sloan Sports Analytics Conference)