Data Analysis on The Value of an NBA Draft Pick

Posted on May 4, 2020
The skills the author demoed here can be learned through taking Data Science with Machine Learning bootcamp with NYC Data Science Academy.

Throughout its rich history, data shows sports have been the ultimate release when looking for a distraction from the stresses of the real world. Amid the COVID-19 pandemic, people assumed watching, talking and breathing sports would be the perfect forget-the-pandemic-for-a-second elixir.

Fast forward to today, and it's been nearly two months without any live sports. The agony! 

In order to alleviate any sports-watching withdrawal, I decided to look at the history of the NBA Draft and analyze the countless number of successes and failures that have accompanied it. This means looking at which teams have drafted the best and worst and with which selection has those teams been the best and worst, which college have produced the best and worst talent, and which draft pick numbers have seen the most return in value. 

A little background...

The first NBA draft, when the NBA was called the Basketball Association of America (BAA), was in 1947. At the time, there were only 10 teams and draft had 10 rounds. As the years went on, the number of teams increased, but so did the number of rounds, with the number of selections maxing out at 238 in 1970. To put that number into perspective, this year in the NFL draft, there were a total of 255 selections; each NFL team (32 teams) has around 53 players, while an average NBA team has 15 players, so let that 1970 selection number sink in. 

The draft became an important cog in building winning teams, but eventually, in 1989, the NBA draft evolved into only having two rounds, and has continued with that format to this day. This is where my NBA draft analysis began. I believed it only made sense to analyze data from drafts of similar nature. While the number of teams in 1989 has increased from 24 to 30 today, I felt that this dataset was the most uniform and would the best representation in looking for draft successes and failures.

The Data

In order to understand this NBA draft obstacle, I found a dataset from that included draft classes from 1989-2016. The dataset included each player picked, the pick they were selected at, the team that drafted them, where they went to college (or N/A if they came straight of out high school or didn't go to college), their career statistics up to 2016, and a few advanced statistics: Win Shares and Value Over Replacement.


Win Shares (WS): a single number assigned to player showing their contributions to their team in a year. It is calculated from their offensive and defensive contributions.

Value Over Replacement (VORP): similar to win shares, it's a single number assigned to player showing their contributions to their team in a year, but in comparison to a replacement-level player (an average/below-average player). It is calculated from their offensive and defensive contributions.

Average Value over Replacement: Value Over Replacement divided by number of seasons played.

Expected Value Over Replacement: based on a draft position, a number is calculated to find the Value Over Replacement a team expects to get when drafting a player at the pick.

Formula: 21.930875 - 5.414771 * log(Draft Pick)

Value Over Replacement Differential: Expected Value Over Replacement subtracted from Value Over Replacement.

The Kareem of the Crop can Draft

A nice little pun to start off the analysis! Kareem Abdul-Jabbar isn't part of this dataset, as he was drafted in 1969, but thought I'd start off with a dad joke.

Anyway, to little surprise, many of the winningest teams, based on win percentage (number of wins divided number of games) in the 1989-2016 span have also drafted the best. In terms of total VORP and VORP differential, this applies to teams like the San Antonio Spurs, Seattle SuperSonics, Los Angeles Lakers, Phoenix Suns, and Boston Celtics. To some surprise however, some of the worst teams in terms of record have also seen high total VORP and VORP differential, including the Minnesota Timberwolves, Orlando Magic, and Charlotte Hornets.

Alternatively, however, the expectation still holds true that some of the worst teams in terms of wins have also drafted the worst in terms of VORP and VORP differential, including teams like the Washington Wizards and Los Angeles Clippers. However, teams like the Portland Trail Blazers and the Philadelphia 76ers, who have been among the winningest teams, haven't drafted all that well, which makes you wonder how they have been able to turn many not-so-great drafts into a substantial number of wins.

Data on The Blue Bloods 

Colleges like Duke, Kentucky, Kansas and North Carolina have been known to recruit the best players out of high school and ultimately win the most, earning the classification of "The Blue Bloods of College Basketball." Based on this, the expectation would be that these colleges would send many players to the draft who would lead long and extremely successful NBA careers. This wasn't the case.

*It's important to note that when analyzing which colleges have produced the best NBA talent that you look at Average Value Over Replacement instead of total Value Over Replacement. This is because many of these colleges have sent a different total number of players to the draft over the years, so it wouldn't be fair to compare a college who has sent 40 players between 1989-2016 to a college who has only sent 10 players.*

**It's also important to note that when analyzing these colleges, I filtered out teams who sent 6 players or less to the NBA draft. Teams like Davidson, who have only sent one player to the NBA draft in the 1989-2016 span, which happens to be Stephen Curry, one of the best players in terms of VORP and VORP differential, can skew the data. Outliers are important to understand, but can undermine data when trying to make certain generalizations.**

Back to the analysis. To much surprise, colleges that haven't been considered to be the juggernauts of the NCAA, like Wake Forest, Alabama, California, Marquette and Clemson are some of the best at sending players to the NBA and having them become perennial stars, in terms of Average VORP and VORP differential, while Duke, Kentucky, Kansas, and North Carolina lie close to or at the bottom. The Blue Bloods reign ends at the college level!

It's also interesting to see that in terms of VORP differential, players who have come straight from high school or from international leagues, rank 4th in production (high schoolers are deemed not ready for the jump in competition and international players play a much different kind of basketball).

The Bulls get the 7th pick every year!

As a diehard Chicago Bulls fan (I hope you're all watching The Last Dance!), it feels like in recent history, they keep getting the 7th pick in the draft. And while they haven't done terribly with that pick solely based on an eye-test, it made me curious to see how NBA teams have faired when drafting at the same pick over many drafts.

To no surprise, teams like the Miami Heat, Golden State Warriors, and Atlanta Hawks, who have had multiple top ten picks in the draft, have seen much return in value based on VORP and VORP differential. However, it's interesting to see that the San Antonio Spurs and Utah Jazz, who have had multiple picks at the end of draft have seen a similar return in value to those with high draft picks.

When looking at teams who have seen the least return in value, it's fascinating to see that every team in the bottom 10 picked no lower than 15th!

Kudos to those teams with masterminds in their front offices! However, some re-evalution is needed for those teams who's front offices have continued to mess up their high picks.

The Picks, The Picks, The Picks

What I was most curious about when choosing this dataset to analyze was to see the value of each draft position. And what I found is both expected and utter shocking.

When looking at total VORP, it was expected that the top 10 in this criteria would be nearly every top 10 draft position. The only two top 10 picks who fell out of the top 10 VORP were the 6th pick and the 8th pick, but these draft positions didn't fall far behind. It's interesting to see, however, that the 24th pick, where most would expect average NBA talent to be drafted, came in at number ten in terms of total VORP. And the alternative that the bottom picks in the draft would be at the bottom of total VORP holds true. This total VORP hypothesis has been settled.

When you scroll over to VORP differential, however, things start to get interesting. In a short and general explanation, when sorting by VORP differential, the total VORP table is essentially flipped, in that the bottom picks in the draft see the most return in value while some of the tops picks see the least return in value. Wild! If we go more in depth, picks 45. 47. 56, 57, and 60 were some of the best in terms of VORP differential. And at the bottom we see picks like 6, 7, and 8.

Finding Homegrown Talent isn't Easy

Hindsight is 20/20. It's easy for me to bash teams for drafting players who didn't pan out and passing on players who took over the league. In each team's mind, at the time of the draft, their high draft picks seem like no brainer All-Stars and their later picks seem like average rotation players who may not play for more than five seasons.

Additionally, the thought process is that the stud from Duke is going to dominate and that the guy from New Mexico may never play a game. Unfortunately, the future is extremely unclear and is often very cruel. A player can get injured, their game may not translate to the next level, or they get traded for a bona fide star, with a team not willing to wait for greatness.

What this data indicates is that drafting isn't a straightforward process. While, more often than not, a team will draft a star with a higher pick, it's crucial that they do extensive research and perform due diligence in terms of a player's injury history, their competitiveness and mentality, and if they're truly ready for the jump to the NBA. Alternatively, this shows that teams shouldn't phone-in their later picks, either. There's hidden talent in each draft every year, and that hidden talent can turn around a franchise.

The NBA Draft can be daunting and lead to a lot of head-scratching, but understanding the value of a pick can truly make-or-break a team for years to come.

For more insights, here's a link to my app.

About Author

Jack Goldman

A recent college graduate with a certification in Data Sciences, having a background in biopsychology, cognition and neuroscience as well as statistics, blended with experience in sports marketing and public relations.
View all posts by Jack Goldman >

Leave a Comment

No comments found.

View Posts by Categories

Our Recent Popular Posts

View Posts by Tags

#python #trainwithnycdsa 2019 2020 Revenue 3-points agriculture air quality airbnb airline alcohol Alex Baransky algorithm alumni Alumni Interview Alumni Reviews Alumni Spotlight alumni story Alumnus ames dataset ames housing dataset apartment rent API Application artist aws bank loans beautiful soup Best Bootcamp Best Data Science 2019 Best Data Science Bootcamp Best Data Science Bootcamp 2020 Best Ranked Big Data Book Launch Book-Signing bootcamp Bootcamp Alumni Bootcamp Prep boston safety Bundles cake recipe California Cancer Research capstone car price Career Career Day citibike classic cars classpass clustering Coding Course Demo Course Report covid 19 credit credit card crime frequency crops D3.js data data analysis Data Analyst data analytics data for tripadvisor reviews data science Data Science Academy Data Science Bootcamp Data science jobs Data Science Reviews Data Scientist Data Scientist Jobs data visualization database Deep Learning Demo Day Discount disney dplyr drug data e-commerce economy employee employee burnout employer networking environment feature engineering Finance Financial Data Science fitness studio Flask flight delay gbm Get Hired ggplot2 googleVis H20 Hadoop hallmark holiday movie happiness healthcare frauds higgs boson Hiring hiring partner events Hiring Partners hotels housing housing data housing predictions housing price hy-vee Income Industry Experts Injuries Instructor Blog Instructor Interview insurance italki Job Job Placement Jobs Jon Krohn JP Morgan Chase Kaggle Kickstarter las vegas airport lasso regression Lead Data Scienctist Lead Data Scientist leaflet league linear regression Logistic Regression machine learning Maps market matplotlib Medical Research Meet the team meetup methal health miami beach movie music Napoli NBA netflix Networking neural network Neural networks New Courses NHL nlp NYC NYC Data Science nyc data science academy NYC Open Data nyc property NYCDSA NYCDSA Alumni Online Online Bootcamp Online Training Open Data painter pandas Part-time performance phoenix pollutants Portfolio Development precision measurement prediction Prework Programming public safety PwC python Python Data Analysis python machine learning python scrapy python web scraping python webscraping Python Workshop R R Data Analysis R language R Programming R Shiny r studio R Visualization R Workshop R-bloggers random forest Ranking recommendation recommendation system regression Remote remote data science bootcamp Scrapy scrapy visualization seaborn seafood type Selenium sentiment analysis sentiment classification Shiny Shiny Dashboard Spark Special Special Summer Sports statistics streaming Student Interview Student Showcase SVM Switchup Tableau teachers team team performance TensorFlow Testimonial tf-idf Top Data Science Bootcamp Top manufacturing companies Transfers tweets twitter videos visualization wallstreet wallstreetbets web scraping Weekend Course What to expect whiskey whiskeyadvocate wildfire word cloud word2vec XGBoost yelp youtube trending ZORI