NBA: Theme week Analytics: Statistics are like a bikini

The figures that mean the world: Statistics play an increasingly important role in reporting on the NBA, but also for managers and coaches.In a theme week, SPOX looks at the “triumphant advance of analytics”, explains the strengths and weaknesses of various advanced stats and talks to a German computer scientist who has become a stats guru in the USA.

“Just because your statistics look good doesn’t make you a good defensive team, Charles Barkley started shortly before the 2015 All-Star-Break:” They scored 118 points.No good team scores 118 points.”

The Houston Rockets had just secured third place in the Western Conference with a 127:118 victory.And apart from the fact that the Phoenix Suns of the 1992/93 season – the team that won Barkley’s Most Valuable Player award and made it to the NBA finals – scored a total of 118 or more points eleven times – this opinion seemed understandable.

Analytics don’t work at all,”says Barkley:” It’s just some kind of crap that smart people have come up with to get into the NBA, even though they don’t have talent “.

The prototype of these Analytics Guys, who according to Barkley “never played basketball and didn’t get any girls in high school,” has been Daryl Morey for years.However, the general manager of the Rockets is by far not the only one of his guild, more and more teams are relying on fresh personnel without a classic NBA past.

Since 2011,30 new General Manager positions have been filled in the league (in some franchises, of course, there has been more than one change) – in 25 cases, the job went to an applicant who had not previously been active in the league as a player or head coach.A trend towards “modern” methods can also be observed further down in the hierarchy.

Just under ten years ago, only five organizations employed at least one data analyst, and there are now 70 full-time Analytics Guys throughout the NBA.The pioneers seem to be already benefiting from this development, while other franchises are lagging behind. Of the ten teams that did not have statisticians in their ranks before 2012, eight are among the most unsuccessful of the past five years.

This group includes among othersthe Orlando Magic, Los Angeles Lakers, Sacramento Kings and New Orleans Pelicans, whose small number of victories also go hand in hand with disappointing draftpicks, lost trades and antiquated play.However, it would of course be too easy to blame this on the sleepiness of the statistical trend.

Charles Barkley has another, not less simple explanation:”It’s all about talent,” but statistical data can be helpful when looking for that talent.An impressive example of this was John Hollinger, who swapped his job as a data analyst at ESPN for the position of Vice President of Basketball Operations at Memphis Grizzlies.Shortly after assuming office, Hollinger made a decision there that surprised many of the team’s supporters.

In January 2013, he shipped Rudy Gay to Tayshaun Prince, Ed Davis, Austin Day and a second round pick to the Toronto Raptors.At that time, the 26-year-old Gay was the top scorer of the Grizzlies with 17.2 points per game and was accordingly popular with the fans.But Hollinger trusted other numbers.

The Player Efficiency Rating (PER), which he developed, saw Gay as a below-average offensive player, whose effective field throwing rate (eFG%) was one of the worst in the league.However, the forward was the most expensive player in a squad with Zach Randolph, whose salary costs were already five million dollars above the luxury tax limit.

The assumption that Gay’s departure would hardly weaken the team in sporting terms was confirmed.While Memphis also had the second-best defence in the league thanks to Prince’s new addition, young Mike Conley was the most active defender.The team finally won 56 games, scored the top seeded Oklahoma City Thunder in the playoffs, failed in the conference finals at the San Antonio Spurs – and saved several million dollars in salary costs.

One thing, however, could not be said of the Memphis Grizzlies under John Hollinger at first: a radical focus on Moreyball.This offensive system, dominated by threesomes and baskets, has been driving Daryl Morey in Houston to the top for years.In Tennessee, on the other hand, the trio’s triumphs did not move into the midfield of the league until an aging Z-Bo slipped into a new role as a backup last season.

Nevertheless, Moreyball is probably the clearest indication of the “triumphant advance of analytics”.Ever since cameras have been installed in the halls of the NBA, transforming every ball and player’s movement into data points, the teams have had unprecedented opportunities:”Analytics are simply information,”said former Sixers manager Sam Hinkie,”We use them constantly and everywhere.If you wanted to know fifty years ago whether rain was coming, you had to look west.Today you turn on your mobile phone and millions of sensors give you a weather forecast.That’s much better than looking west.”

From the information about the average value of different litters, Morey came to a central conclusion:”Tests from the middle distance are the worst litters of the NBA.They are harder to hit than basket-goods and have fewer points than attempts to hit beyond the three-line.The goal of a successful team should therefore be to avoid these “bad” throws if possible.

This simple insight seems to have taken the league by storm in recent years.By 2012, teams had taken an average of around 18 distance throws per game for years, and a good one-fifth of the field throws came from beyond the three-line line.By the 2016/17 season, this figure had risen rapidly, and now almost every third roll in the NBA is a threesome (27 per game).In Houston, almost half of all experiments were fired from a distance in search of maximum efficiency.

But of course, even the best data analyses do not guarantee success.You need talented players – no Morey, Hinkie or Hollinger Barkley would disagree.You need trade partners who are willing to give up such players.These talented players need to be healthier than Chandler Parsons in Memphis and understand each other better than Dwight Howard and his fellow players in Houston.

A man who is as aware of this as he is of the importance of statistics in the modern NBA is Bob Myers.At the MIT Sloan Sports Analytics Conference, the annual “Dorkapalooza” (Bill Simmons), talented nerds, analysts and managers, the GM of the Golden State Warriors said in 2016:”Analytics are like a bikini.They show a lot, but they don’t show everything.”

nearby

The SPOX theme week is intended to shed some light on the dark side of the payment jungle.An interview with Jeremiah Engelmann is the first step.ESPN’s stats guru tells the story of his journey from Karlsruhe University to the best basketball league in the world, his time with the Phoenix Suns and explains the benefits of data for teams, players and fans.He also gives an insight into the emergence of his Real Plus-Minus (RPM), talks about undervalued role-players and overvalued stars and forecasts for the coming season.

From Tuesday onwards, everything will be under the sign of the numbers.Whether it’s the individual offensive performance of players, their role in the team structure, their contribution to defence or the performance and game idea of entire teams: in today’s NBA, all of this is evaluated on the basis of a wide variety of numbers and metrics.We explain what these statistics tell us – and what not.After all, many of them aren’t “shit”but bikinis.