HOW DATA SCIENCE CHANGED THE NBA
According to Forbes, ‘Data Scientist is the sexiest job of 21st century’ and broader uplifting curve of the same is surely yet to come.
Data Science has the potential to give astonishing results in every human thinkable domain, and the signs of such eye-dropping results are already visible to the human eye. One such domain is the Sports. With more and more research happening in DATA SCIENCE, several sports team owner and HEAD COACHES continue to increase their trust in data analytics, due to the magical insight it gives the team to prepare and understands not only what’s best for them, but also what’s worse for their opponents and how they can get an edge over them. Here, we will be covering about NBA, which was one of the first major sports leagues to bring data science into the sports.
NBA
The National Basketball Association (NBA) is an American men’s professional basketball league. It is composed of 30 teams (29 in the United States and 1 in Canada) and is one of the four major professional sports leagues(and certainly the richest) in the United States and Canada. It is the premier men’s professional basketball league in the world.
Traditional basketball includes big man dominating the game with his size and power near to the basket, and of course the basketball skills to actually put the ball inside the basket. But, the way the basketball is played now has changed drastically the way in comparison to the way it was played before.Modern (or “dead NBA” as per Stephen A. Smith), comprises a lot more open game as compared to just dominating near the basket, and uses a strategy “THE-ALL-CAN-SHOOT-3's”, regardless of their position that they play.
THE 90’S NBA:
90’s NBA was a physical, more exhausting and much more fierce display of the strength of the strong-man. Most Point Guards used to think twice before even going down the lane, to get a layup, as the rules were more lenient towards the defender. Contact was considered a foul, (no matter how hard it was), until the defender touches some part of the ball. This was the reason there was so many 2-point mid-range jump shots taken by the players.
As it can be clearly noticed, the use of 2-pointer in 80’s was around 96%, and only 3% of the shots taken was 3-pointers, as compared to the stats in 2019, 25% of the total shots taken were 3-pointers. (The code and dataset will be in Github).
THE MODERN NBA:
The modern NBA, as Max Kellerman says, “It’s a wise-man’s league.” The NBA players certainly have become smarter and learnt to play without getting injured as they used to get in the 90's.Also, the improvement in health facilities play a major role in the significant improvement.But there is lesser and lesser physical play in each and every season that comes. The 3-point shot attempt has increased significantly and usually the team that’s hotter from beyond the arc, on a given day, wins the game.
HOUSTON’S EARLY BELIEF:
Houston Rockets was one of the first major leagues sports team to have a seperate department for data and analytics. This was introduced in 2013, and most people didn’t understand it’s potential at that time. The team, got some fair success in the early years, but they caught the world’s attention by shock in 2018 against the GREATEST TEAM OF BASKETBALL EVER ASSEMBLED(Golden State Warriors).
As all NBA playoff series are, it was a best of 7-game series and frankly, just as good as James Harden and the Houston Rockets were,no one expected them to even win a game against the mighty Warriors. The Houston Rockets forced a game-7 and lost in final moments of the game.
How did they force the game-7? What different did they do?How did they stop such a mighty force?
Well, they did 2 things differently, that if they would said before doing, they would have been a jok in the media.
- NO BIG GUY in rotation for more than 10 minutes a game.Force the fast-break as the bigs cant run quick. Force the opponents as well, to go all small.
2. Everyone takes a three if left wide-open. No matter the regular season 3-point %, no matter the career statistics, just shoot.
Houston, in 2018 series, calculated eFG%(effective Field Goal %), instead of the traditional FG% and experimented a theoretical game-play which not only gave a chance to stay in the game, but also changed the way game is visualised by the statisticians and the coaching staff.
They data and analytics department introduced the below mentioned:
NEW VISION IN BASKETBALL STATISTICS:
eFG% = (2pt FGM + 1.5*3pt FGM) / FGA
USG% = 100 * ((FGA + 0.44 * FTA + TOV) * (Tm MP / 5)) / (MP * (Tm FGA + 0.44 * Tm FTA + Tm TOV))
Usage percentage is an estimate of the percentage of team plays used by a player while he was on the floor.
It was a shocking discovery that even with so many changes occuring in the way that the league is being played, the eFG% of the players according to players hasnt changed at all. Houston took advantage of this by playing all the small in their rotation so all could spread out the floor and be wide open to shot the three, whose eFG% has to be equal to 1.5 times less than the regular basket being made.
The USG% of point guards is comparatively higher, and the hustles ,dives to get the 50–50,plays that essentially doesn’t count in statistics, favours the playing of point guards. This stats are counted in USG%.
The github link to my code is as below:
