The article, called Applying ‘Moneyball’
Methods to the NBA, is a reference to the book written by Michael Lewis
called Moneyball: The Art of Winning an Unfair Game, as well as the film
Moneyball, which told the story of the 2001 MLB Oakland Athletics using
a more statistical approach to value players instead of the more traditional methods
being used at the time. In 2001, the Athletics were one of the least funded
teams in the MLB. After having a great previous year, wealthier teams, such as
the Boston Red Sox and New York Yankees, bought most of their all-starts
because the A’s simply could not afford them. As a result, the General Manager
Billy Beane looked for other ways to find value in inexpensive players. With the
help of a Harvard graduate with a degree in economics, Paul DePodesta, the team
relied on Big Data and statistical analysis to draft new players. For instance,
the team drafted players disregarded by the rest of the MLB, on account of them
being of old age or having injuries, because they have statistically proven
they can get on base more so than other famous players at the time. Beane took
a lot of criticism for his decisions from the rest of the baseball community,
but after a very poor start, the Athletics went on to beat the American League
record for most games won in a row with a total of 20 games. Although the A’s
later made it to the American League post-season, they lost early to the New
York Yankees. However, what’s very interesting is that during the regular
season, the Athletics won 7 more games than the Yankees, but where the Yankees
spent almost $1.2 million per win, the Athletics only spent a little more than $330,000
(Source).
The Boston Red Sox, becoming inspired by the Athletics’s new business model,
decided to implement it themselves and in 2004, they won their first World
Series since 1918.
Today, a lot more professional sports rely on a more statistical
approach, but it is still not yet fully embraced by the community. Entering an age where
Big Data is becoming more accessible, easier to manage, and easier to analyze,
I think it can be extremely beneficial in further analyzing a player’s
performance. In the article, a Stanford medical student, Mutha Alagappan,
attempts to use Big Data to determine the different play styles of NBA players.
In the NBA, there are a total of 5 positions, which are point guards, shooting
guards, small forwards, powers forwards, and centers. On one end of the
spectrum, point guards are known to take a lot more 3-point shots, while
centers on the hand are very tall, and stay closer to the rim. While working
for a Big Data startup, Alagappan decided to test its software by using
basketball statistics. Expecting to see a statistical map highlighting 5
different positions, he instead saw a total of 10 different positions, with
each having their own unique trait.
Alagappan further explains his analysis by discussing the
shooting patterns of 3 different point guards, Chris Paul, Steve Nash, and
Jason Kidd.
By looking at the charts, you can see that Jason Kidd takes
a lot of 3-point shots, Chris Paul likes to charge the rim, and Steve Nash is a
mix of the two. What this proves is that even though all of these players share
the position of point guard, their play styles differ drastically. By showing
this data, Alagappan hopes to emphasize the need for a better classification of
player positions. In doing so, NBA managers and coaches can better build their
teams by determining what types of offenses and defenses to run, as well as
what players to start against certain teams. For instance, when facing a team
specializing in Paint Protectors that have exceptional defense close to the
rim, it might be wise to use 3-Point Specialists who have no need to get close
to the rim. It opens the door for so many more options, and having more options
on the court is never a bad thing.
The term Big Data is becoming increasingly popular in today’s
market, and I believe the need for it will continue to increase drastically
within only a few years. We learned in class how Big Data can be used against
us in the case of NSA surveillance, but there are still a number of other
situations where Big Data can be extremely beneficial. I think as time goes on though,
an imaginary line will need to be drawn that specifies exactly how far can we
go in terms of Big Data. There is obviously no relation with the use of Big
Data by the NSA and the method I have mentioned above, but I’m sure there will
be plenty of gray areas that will eventually have to be handled as ethically as
possible.
No comments:
Post a Comment