Visualizing the Great NBA Point Guards
This is a data visualization project many years in the making. I had taken a D3.js class in an effort to add another arrow to the business intelligence quiver. I had gotten to a certain point but could not get over a couple bugs and have let it sit for years on the backburner.
The premise of this post, when I first conceptualized what I wanted to do, was to not forget about John Stockton as one of the all-time great point guards in the NBA. Humans are prone to recency bias. It’s been over 20 years since he has retired and emerging talents like Luka Doncic and Shai Gilgeous-Alexander will continue to come into the NBA causing us to forget about some illustrious careers of yesteryear.
I can’t fault anybody for taking Magic Johnson or Steph Curry as their all-time point guard due to each of those individuals having won multiple championships, multiple MVPs, and the way they each impacted the game. I would say how amazing is it that John Stockton was able to statistically differentiate himself in so many categories.
There are only so many times a basketball can be shot during a game and to compose a team strictly of scorers has repeatedly proven to lead to discontent. A player like John Stockton helps alleviate the potential tension of needing to ensure points are disseminated across the team due to his pass-first mentality. Yet adjusting the chart to show “True Shooting Percentage” and “Effective Field Goal %” will clearly convey that John Stockton is more than capable of providing you with a bucket if needed.
Some additional notes on the above chart regarding who is and isn’t included. The NBA didn’t keep track of steals before the 1973-74 season. Historic greats like Oscar Robertson and Bob Cousy would not have a record of not only that particular statistic, but more complex metrics like Value Over Replacement Player (VORP) as well.
There are a lot of very talented players in the NBA that have not had the longevity of career yet to include on this data visualization. Magic Johnson played in 906 regular season games. So I made the cutoff 900 regular season games played to qualify. The data represented in the cart was accurate as of the 2024 All-Star break.
Enhancing with ChatGPT
One of my favorite things about this project was utilizing artificial intelligence to resolve some issues I was having. Originally the dots on the chart were a solid red color. You had to click on the circle for the tooltip to help you know who the dot represented. I could have assigned a color based on primary jersey color, but that does not provide as smooth of an experience as the images does.
In the dataset I included an image URL for each player profile on Basketball-Reference.com.
{"Player":"John Stockton","Assists":15806,"Steals":3265,"Games":1504,"eFG":0.546,"Efficiency":21.83,"TS%":0.608,"Win Shares":207.7,"VORP":106.5,"image":"https://www.basketball-reference.com/req/202106291/images/headshots/stockjo01.jpg"}
I then engineered the following prompt:
“As a computer engineer specializing in web development with HTML, CSS, and JavaScript scan the following code. Make CSS adjustments to the svg circles so stroke equals black and stroke-width is 5 pixels. Use the dataset “image” url to fit the photo of the player inside the circle.”
I then plugged in my code and ChatGPT manipulated my code adding this piece of functionality making the end result much more aesthetically pleasing.
There are a few more enhancements I’m working through. For example utilizing the .transition() function to animate a smooth transition for the circles as the values adjust. However, I’ve learned that artificial intelligence is not quite creating bug-free code just yet.
Final Thoughts
I contemplated changing this to a different topic for a couple reasons.
First, in a recent interview a former scorekeeper turned team executive, Alex Rucker, disclosed a practice of inflating stats during the 1990s, even citing specific examples by name including John Stockton. I cannot refute the probability of this claim, but have a hard time completely discounting the correlation of total games played as being a larger indicator of how John Stockton became the all-time leader in both assists and steals. Also, you cannot deceptively boost made field goals and efficiency as a result of not carelessly turning the ball over. I also can’t help but draw the comparison of the baseball steroid era; yes, Barry Bonds was taking steroids, but so was the vast majority of other players during his time and yet he still differentiated himself as THE slugger of his generation.
Second, his public conduct in regards to COVID restrictions has made him somewhat of a controversial figure. I do appreciate hearing as many viewpoints as possible on issues to try to become as informed as possible. Yet, I have concerns with some of the arguements of his perspective, however that does not have to diminish from his accomplishments as an athlete.
In conclusion, happy birthday to one of the top 5 influences on how I patterned my playing style. I doubt I will see either of his records break in my lifetime.