By: Alvin Huang, Chloe Kim, Kristen Ng, Selena Lam
Introduction
Every year, the National Basketball Association (NBA) holds an annual player draft. Through this annual draft, teams can increase their chances of winning a championship by drafting players who can help the team. As franchises meticulously select emerging talents, the question arises: does a player's draft ranking correlate with their subsequent performance on the court? In the following analysis, we embark on a journey through the annals of draft history, spanning multiple seasons to unravel the intricate relationship between draft ranking and player performance.
Transitioning from the NCAA, EuroLeague, or G-League to the NBA can be a monumental challenge for many young players. Yet, across multiple seasons, players such as Anthony Edwards, Luka Doncic, LeBron James, and Anthony Davis emphatically demonstrate why they were the first picks in their respective draft classes. However, the narrative extends beyond the early selections. Players like Nikola Jokic, Jimmy Butler, and Giannis Antetokounmpo, drafted in the mid or late rounds, have convincingly shown us that draft ranks do not determine success, proving themselves to be some of the best in their draft classes. With this in mind, let's examine the rank, position, and NBA performance to uncover the captivating stories woven into the fabric of basketball history.
Methodology
Our decision to consider players with a minimum of 5 seasons aims to mitigate the impact of early career variability and better capture the essence of a player's overall contribution to the NBA. This approach aligns with the belief that experience plays a crucial role in understanding a player's capabilities, and averaging performance over a more extended period provides a more stable representation.
To evaluate player performance, we devised a scoring system incorporating various factors, each assigned a weight summing to 100%. These factors were chosen to reflect individual and team success, acknowledging the importance of a player's impact on both fronts. Specific weights and factors were chosen and assigned based on subjective judgments about the relative importance of different performance metrics. This method was used to understand the subtle ways players contribute to the game. By looking at how well a player performs individually and their role in the team, the scoring system aims to give a complete picture of a player's impact on the basketball court.
Player Success Score = (Weight1 Factor1 + Weight2 Factor2 + ... + WeightN * FactorN) / Total Weight
Factor 1: Championships Won 5%
While winning championships is a significant achievement and reflects a player's ability to perform under pressure, it's important not to overemphasize it. Basketball is a team sport, and individual contributions can be limited by team dynamics.
Factor 2: Shooting Efficiency Percentage 20%
Efficiency in scoring is crucial, as it directly contributes to a player's overall impact on the game. This includes field goal percentage, three-point percentage, and free-throw percentage.
Factor 3: Usage Percentage 15%
Usage percentage measures the percentage of team plays used by a player while he is on the court. While it's important to consider a player's involvement in the game, excessive usage might not always translate to efficiency.
Factor 4: Assist Percentage 15%
Assists reflect a player's ability to create opportunities for teammates. This is crucial in evaluating a player's playmaking and team contribution.
Factor 5: Rebound Percentage (Defensive + Offensive) 15%
Rebounding is a versatile skill that contributes to both offense and defense. Offensive rebounds can lead to second-chance points, while defensive rebounds help end the opponent's possession.
Factor 6: Average Number of Points 20%
A player's average number of points per game is a classic metric for evaluating scoring consistency. This factor considers a player's ability to consistently contribute to their team's point production, emphasizing the importance of scoring output in overall performance.
Factor 7: Number of All-Star Selections 10%
All-Star selections indicate a player's consistency and recognition by peers and fans. However, the All-Star selection process may be influenced by popularity and market size, so it shouldn't carry too much weight.
For championships won and All-Star selections, we collected our data from Basketball-Reference. For the other 5 factors, we obtained the data from Kaggle.
To clean the Kaggle dataset, we selected the relevant variables (player_name, draft_round, draft_number, pts, oreb_pct, dreb_pct, ts_pct, ast_pct, usg_pct). Then, we created a data frame by subsetting all players who have played 5 or more seasons. We replaced occurrences of “Undrafted” in the data frame with 0 and changed the data type of the draft_round and draft_number columns to numeric. In order to obtain just one occurrence for each player, we averaged the columns of each player and concatenated columns for their championships won and All-Star selections.
To obtain a score for each player, we normalized each factor by calculating the z-score as follows:
zi = ((xi – mean(x)) / sd(x)) * 100
where:
zi: The ith normalized value of the respective factor
xi: The ith value of the respective factor
mean(x): The mean value of the respective factor
sd(x): The standard deviation of the respective factor
This approach was adopted to prevent any single variable from being overly influential.
We focused on various statistics to gain a comprehensive understanding of player performance. The z-score of each factor for each player was summed up in order to obtain their total score, and these scores were graphed on a scatter plot. A correlation map was also constructed to unveil the strength and nature of relationships between different factors. The color intensity in each cell served as a visual cue, highlighting correlations. Lastly, we utilized a histogram to illustrate the frequency distribution of draft ranks among players selected for All-Star games, providing insights into the draft positions most associated with All-Star selections.
Results and Analysis
Analyzing data from various drafts, we used a heatmap to see if there's a connection between when a player is drafted and how well they perform. To dig deeper into this, we considered other factors alongside draft position. It turns out that there's a clear link between points scored and assists, meaning if a player scores more, they're likely to draw more defensive attention, creating chances for teammates to score. However, we didn't find strong connections between draft position and other performance measures. This suggests that a player's draft ranking doesn't consistently predict their performance.
Illustrated by this bar chart spanning the years 2000 to 2016, a notable pattern emerges: 16 players chosen as the first overall draft picks during this period went on to achieve NBA All-Star status. While the prevalence of All-Star players emerging from the first overall pick aligns with expectations, it is intriguing to observe a substantial number of All-Stars drafted late in the first round and even into the second round. Of particular interest is that the 9th overall pick produced the third-highest number of All-Stars, suggesting a noteworthy trend favoring players selected in this position. Notably, Isaiah Thomas, drafted 60th overall, defied conventional expectations by becoming a 2-time NBA All-Star. His standout 2016-2017 season, marked by an impressive average of 28.9 points, 5.9 assists, and a 3-point field goal percentage of 0.38, further exemplifies the dynamic nature of talent discovery in professional basketball.
Who doesn’t love an underdog? Following the culmination of the two rounds and 60 picks of the NBA draft, there’s still hope for players who weren’t drafted. They can get picked up by a team as an undrafted free agent and make a huge impact in the game, leaving many people questioning how they had managed to fly so far below the radar before and during the draft. Our list of the top undrafted players is according to our success metric. There were ten players with success scores over 12, the y-intercept we found when fitting a regression line to a scatterplot of all player scores. This means these ten players performed above average despite being undrafted, proving that even the most unexpected stars can emerge in the NBA.
For the draft round 1 player scores, we regressed the player score against the draft number and found a correlation coefficient of -0.3451212. This indicates a negative relationship between the player’s draft number and their score. As draft number increases, the player’s score tends to decrease. However, the magnitude of the coefficient of just -0.345 suggests only a weak correlation between the two variables. The correlation is not strong enough to predict scores based solely on their draft number. This aligns with what we found in the heat map where we did not find a strong correlation between the draft position and other performance measures. This suggests that a player's draft ranking doesn't consistently predict their overall performance in the NBA.
Similarly, for the draft round 2 player scores, we also regressed the player score against the draft number and found a correlation coefficient of 0.02730532. This coefficient is close to 0, indicating almost no correlation between the two variables. This implies that draft number is not a strong predictor of round 2 player scores. Other factors that are not captured by the draft number play a more significant role in determining the scores of round 2 players.
Limitations and Improvements
It is important to recognize variability in how players are scouted and evaluated. Our exploration into the dynamics of player development underscored the significance of factors beyond what could be seen in our data. Injuries, team dynamics, coaching strategies, and the ability to adapt to the rigors of professional basketball all play major roles in shaping a player's career and success in the NBA. The unpredictability of these variables, coupled with the evolving landscape of the sport, adds an element of complexity beyond our statistical analysis of game data.
To enhance our analysis, we can consider injuries in our data since that is one of the biggest factors hindering player success. Talented players can suffer a season-ending injury and be completely out for one or more seasons. We can also factor in average playing time because the opportunity to play can increase a player’s points per game, assist percentage, rebound percentage, etc., increasing their player success score.
Conclusion
Overall, the dynamic between player draft ranking and performance in the NBA reveals a deep, multifaceted talent evaluation and development story. As we dissected the data, we found a pattern of success in players beyond just the first few picks in the drafts, with underdogs drafted late or even those undrafted, highlighting the successes and pitfalls of the drafting process.
While in a constant quest for the next generation of stars, teams in the league must consider selecting for immediate impact and investing in long-term potential. While high draft positions often yield franchise-altering talents, the NBA's history is full of tales of diamonds in the rough—players who defied their rankings and developed into elite performers throughout their years. This could be due to the “underdog mentality,” where players selected later in the draft have a stronger desire to prove themselves and have greater motivation to excel and showcase their capabilities. Just like individual talent is important, a player’s fit with their team is equally as important. Success often depends on how well a player’s skills align with the playing style and needs of the team that drafted them.
Ultimately, the NBA draft remains a thrilling event where dreams come true, and lives are forever changed. As we have seen, for some players, it is only the beginning of their rise to worldwide recognition and stardom. While the analytics provide invaluable insights, the humanistic elements, including resilience, work ethic, and other intangibles that defy quantification, continue to shine bright in the league. At the end of the day, beyond the numbers are the stories of triumph, resilience, and the relentless pursuit of greatness that truly make the sport what it is.
Comments