An Analysis on NBA Team Strategy and Its Implications for the League
By: Jonathan Haroun
Basketball is an ever-changing game. While NBA rules have remained the same over the past two decades, team play has evolved greatly. In this article, I seek to explore these changes through the lens of those teams who have mastered the current iteration of league strategy — the Finals champions.
I found particular interest in NBA strategy because each player has a tremendous impact on the outcome of the game. With only 5 players on the court, each one contributes a huge piece to his team’s success. Players with unique skillsets can completely disrupt existing schemes and birth offensive strategies that are novel to coaches and executives.
This analysis will cover NBA games from the 2003–2004 season to the most recent 2019–2020 season. My initial interest was to explore a possible correlation between a major statistic and an NBA champion across all seasons. I decided to examine the following major team statistics: Rebounds, Assists, Steals, 2-Pointers Made, 3-Pointers Made, Offensive Rebounds.
Below are the t-values of correlation between these major statistics and an NBA championship team.
(Intercept) AST REB `2PTM` `3PTM` OREB STL
-0.03 1.73 0.91 -0.85 -1.34 -2.14 0.64This regression tells us that teams with high assist, rebound, and steals totals are more likely to win championships. While teams with high 2FG, 3FG, and offensive rebound totals are less likely to win championships. However, this analysis is naive. This regression assumes that the ‘value’ of these statistics does NOT change from season to season. In reality, we know this not to be true. Let us step back and evaluate how these statistics have changed from year to year by The Finals champion.
Rebounding and 2-Point Field Goal numbers appear to have oscillated but remained similar from year to year. This is in stark contrast to assists and ‘3-Pointers Made’, with assists increased 1.5 times and 3-Pointers Made increasing more than two times since the 2004 season. These are stats that change markedly from year to year and highlight the inadequacy of the initial regression.
It’s particularly remarkable to zoom in on the 3-Point FGs by Finals Champions over time. The number of made 3 pointers a game has more than doubled since 2004. This graph highlights why the initial regression was inadequate — while the 3 three-point shot played little role in the Pistons 2004 Title, it played a very large role in the 2019 Raptors Title. The Golden State Warriors Team of 2015 was the first to construct their game plan around the shot. They were led by the Splash Brothers, 3-Point Sharpshooters Steph Curry and Klay Thompson, who not only shot the 3-pointer accurately but also took the shot a lot — together averaging over 15 attempts from behind the arc a game.
With that in mind, it makes the most sense to conduct this analysis by year. What defines a great team changes from season to season, as league play adjusts. Therefore, we should do an analysis on a season-by-season basis. Below is a table displaying the season and the statistic most correlated with wins during the regular season across the league.
There is a very interesting trend that emerges from this table. From 2004-2013 Rebounds were most commonly correlated with winning and from 2014–2020, 3-Point FG made were most commonly correlated with wins.
In the graph to the left, the t-value of correlation with wins is plotted against these two major statistics. The larger the t-value the more the statistic is correlated with success. It is clear that in the early 2010s the value of rebounds decreases while the value of three-pointers increases. There are primarily two reasons for this sharp change starting in 2014: The growth of ‘position-less players’ and statistical appreciation for the 3-point shot.
Growth of Position-less Players
The league is increasingly seeing players with unique skillsets enter and change the way that traditional positions — point guard, shooting guard, small forward, power forward, center — are defined. The earliest and most memorable example of such a player is Kevin Durant. Standing at 6"10, one would expect that Durant would spend his time in the low post, playing center and power-forward and finding most of his production in the paint. Yet despite his height, Durant is a prolific dribbler, driver, and shooter, and while he still has the height to get involved down low, finds most of his production starting from behind the arc.
“He’s 7 feet tall and he’s running around like a two-guard. He can handle the ball, he can take you off the dribble, he can post you up, he shoots over you. You can’t trap him because he sees right over you. I don’t think in the years that I’ve coached and played . . . there’s ever been a more difficult guy to prepare for. You really feel like you’re wasting your time doing it. He’s going to probably score anyway.” — Doc Rivers, Clippers Head Coach
Recently, we have seen an influx of these ‘position-less’ players in the NBA. These players allow teams to shift their offenses in dynamic and previously unseen ways. Their entrance has led to new offensive schemes and has contributed to the evolution of the game.
Increasing Appreciation for the 3
When the 3-Point line was established in 1986, many were skeptical of its actual use. The New York Times ran an article proclaiming the newly established shot as a “gimmick”, and coaches across the league waved off its adoption. When asked about the new shot resident Suns coach John MacLeod stated:
“I’m not going to set up plays for guys to bomb from 23 feet. I think that’s very boring basketball.”
Recently, professional NBA teams have turned to statisticians in hopes of gaining strategic insights on how to maximize their play. These analysts have discovered that the most efficient shots as measured by expected points per shot are those close to the rim and 3-point shots. This conclusion is logical — if you were to take a 2-point shot it should be close to the basket so it’s easier to score, and if you should shoot a shot far away it should be from beyond the three-point line so that you receive additional possible points per attempt.
With the influx of position-less players, the path towards increased 3-point shot attempts has only accelerated. These players force defenses to spread across the floor to defend against them, as they can shoot, drive, and pass. This has led to starting lineups where each player can score from three, a notion very foreign to the NBA in the early 2000s when points were scored in the paint, mostly by big men.
In short summary, what makes a championship NBA team depends upon the players in the league and the philosophy behind scoring. A successful team in today’s NBA is one that possesses multi-talented players, or those that can shoot, pass, and score, and focuses a majority of their shot attempts from behind the arc or inside the painted area.
With this knowledge in mind, let us return to our analysis of NBA championship teams and explore how they have and continue to shape the trajectory of the league. Specifically, let us look at a statistic not found on the scoresheet — the number of top-ranked players on a championship roster. In order to do this analysis, I decided to look at the top 20 players as stated by The Washington Post in their yearly ranking of players.
Interestingly, starting in 2016 the number of top players on a championship roster seems to increase. This coincides closely with the years when the 3-pointer best-predicted wins. The data suggests that these two statistics are related to one another. In light of the previous discussion on increasing 3-point popularity, this makes perfect sense — top-ranked ‘position-less’ players are able to stretch the floor and shoot threes. Similarly, top shooting players are able to score efficiently, taking shots with a high expected point value. These players can work hand in hand with one another to create a team that finds the most efficient shot on most trips down the court. On a team with four top-ranked players who can all shoot the three (as the Warriors were in 2017 and 2018) this was certainly true.
Super-Teams
The Warriors were referred to as a Super-Team after signing Kevin Durant to a roster of three existing All-Stars in 2016. With Durant, the Warriors became Finals champions in 2017 and 2018. It became clear to other teams that it would be challenging to beat the Warriors without assembling a super-team of their own, and thus began the era of the modern super-team in the NBA. With each passing season, it seems, another organization attempts to draw one or two additional stars to their program. These additions can turn a mediocre squad into a top team. Just ask the New York Nets, who after going below .500 last season, added Kevin Durant and James Harden and are now 2nd in their conference.
I argue that a major contributing factor to the era of the modern super-team was the changed style of play in the NBA. A 3-point heavy offense calls for an arsenal of players that can both drive and shoot — this was the 2016–2017 Golden State Warriors and has been the goal of every super-team since.
Super-Teams in the NBA appear to be here to stay. The philosophy of concentrating shots in the paint and behind the arc continues, and an arsenal of stretch players is more admirable than ever for NBA teams. The question therefore remains, how does the Super-Team impact the NBA from an image perspective. Do fans enjoy watching teams with numerous stars, or does it make games too predictable or unfair? Below is the viewership of the NBA finals by Season.
In order to make this conclusion, it would be best to examine how the number of top-ranked players on the championship team impacts Finals viewers. After computing a regression between ‘Number of Top 20 Ranked Players’ and ‘Viewers in Millions’, one discovers that there is a positive correlation between the number of top-ranked players on the champion team and the number of fans that tuned in to watch the series.
Estimate Std. Error t-value Pr(>|t|)
`# Top 20 Players` 0.9784 0.9377 1.043 0.313However, TV viewership does not tell the whole story. In order to get a true analysis of how fans felt about super-teams, it was best to analyze their own thoughts on the matter. I decided to approach this by analyzing fans that tweeted about the formation of recent super-teams. Specifically, I looked at tweets that contained the phrases “NBA” and “super-team”. Using sentiment analysis I discovered that a majority of tweeters reacted positively to the formation of these super-teams, though the reaction was not overwhelmingly strong (sentiment_value=0.418).
Implications for the NBA
Many basketball purists have shaken their fists at the recent formation of super-teams across the league, demanding the league update its rules to prevent a mass gathering of talent on one organization. I could not disagree with these commentators more — basketball’s growth as a sport is a sign that the game is still alive and growing. It has become more high-scoring, fast-paced, and exciting, and the data shows that fans have reacted positively. The NBA should continue to foster the growth of the game of basketball and monitor its support from fans. I’m excited to see what the future has in store for the game of basketball.
Data Sources
- https://www.basketball-reference.com/
- https://developer.twitter.com/en/docs
- https://www.washingtonpost.com/graphics/2020/sports/nba-top-players-2020-2021/
- statista.com/statistics/240377/nba-finals-tv-viewership-in-the-united-states/#:~:text=How%20many%20people%20watched%20the,viewers%20in%20the%20United%20States
