←
NBA Player Stats - Who's the Real MVP?
Presented by: Bladimir Nunez
Last Updated: 11/8/2025
Project Overview
Basketball debates often hinge on subjective opinions, especially when
it comes to the GOAT conversation. This project brings data to the
forefront. Using MySQL, I analyzed over 12,000 player-season records
spanning 27 NBA seasons (1996 - 2022) to uncover performance trends,
compare eras, and identify statistically driven MVPs. Here is a link
below to my SQL file that contains my code and comments.
SQL File
Objectives
-
Identify season leaders in scoring, rebounding, and playmaking and
evaluate their efficiency
-
Compare player size, style, and performance across four eras: 1990s,
2000s, 2010s, and 2020s.
-
Determine which teams, positions, or player profiles consistently
produce top performers.
-
Use a weighted performance index to crown MVPs and compare them to
official NBA selections.
Skills
Demonstrated
- Data Cleaning & Inspection
- Window Functions (RANK,LAG,PARTITION BY)
- Views & Aggregations
- CTE's
Methodology
Data Loading
The dataset contains 12,844 rows of data that shows players stats
across 27 seasons of the NBA. I used Excel to first clean and
standarize the data then imported the cleaned data into MySQL
Workbench for analysis.
Data Cleaning
- Search for duplicates and none were found
- Change column names to be more in line with values
- Standarize data within college, country columns
-
Fill in NULL values within college column with the college name the
player went to, if player had no college I filled in cells with "No
College"
- Converted player_weight column from Kilos to Pounds
-
In draft_year,draft_round,draft_number columns I changed "Undrafted"
values to 0 to indicate the player did not get drafted and casted
those columns to INT.
- All statistics columns I converted to Integers/Double
Exploratory Analysis
Inital EDA includes counting unique amount of players that played in
the NBA over 22 seasons and showing ammount of colleges in the dataset
alongside NBA teams. We found the number of players each college
produced to go to the NBA and showed the percentages of players origin
country's across the 22 seasons.
Key facts from the data:
I found that across the NBA from 1996-2022 83.4% of players are from
the USA. Given that the sport was created and played here in the US
makes sense for that number to be high. The second most NBA players
are from Canada and France covering 1.60% & 1.48% of the league
respectively. Kentucky, Duke & North Carolina produced the most NBA
players out of any college/University. From this time Kentucky
produced 451 players, Duke 410 & North Carolina 355. There are 337
colleges that players came from and only 36 current NBA teams. This
shows the fierce competiveness of a college athlete trying to make it
to the big leagues.
Key Findings
1. Player Performance Analysis
- Efficiency vs. Volume:
- Most Improved Players:
Scoring:
MarShon Brooks (2017-18,+15.5 PPG),Louis King (2022-23, +15.5 PPG),
JaKarr Sampson (2018-19, +15.3 PPG)
Rebounds:
Assists:
Leading players in each season for points,rebounds & assists:
2. Era & Team Comparisons
3. MVP & Dream Team