Skip to content

Latest commit

 

History

History
133 lines (114 loc) · 6.71 KB

PROPOSAL.md

File metadata and controls

133 lines (114 loc) · 6.71 KB
Raw
editor_options
markdown
wrap
72

Motivation and Purpose

Our role: Data scientist consultancy firm.

Target audience: People who are enthusiastic NBA fans and are interested in assessing NBA player performances.

NBA or National Basketball Association is the most famous professional Basketball league in the world. It consists of 30 professional basketball teams with approximately 500 players playing per season.

The motivation behind this project is to help NBA fans better understand their favorite players and visualize how the players have performed historically over time and across players and teams. Since the NBA has a long history and team changes are a regular occurrence, it is necessary to create a data visualization that could show the player's performances.

Further, it helps NBA fans make better decisions when betting, as the visualization of player data helps fans give basic judgments on player performance levels. Fans can determine a player's style such as what kind of player he is, how many games he plays, how much he scores per game, the number of assists, etc., and use this as a starting point to get to know a new player better.

On the other hand, we can also honor and learn those former NBA superstars, such as Micheal Jordan and Kobe Bryant, by visualizing their performances. For instance, Kobe Bryant, we can learn about the wonders the legendary superstar created with the Lakers and the glory of the purple and gold dynasty with scored 33,643 points, 7,047 rebounds, and 6,306 assists during the entire career. Salute to the eternal Lakers #24.

Description of the data

The NBA player stats data for the regular season will be retrieved from in real-time from The Basketball Reference website when the application user searches the NBA player name in the search bar.

The returned data will contain a total of 54 features for each year of this player in the NBA along with some particular summary stats based on the team he played for, as well as a career summary.

For the usage of the returned data, we won't be using all of the features.

Before diving into the details of the data, it would be necessary to explain what each used feature column means: - Season : The regular season played. - Age : Player's age on February 1 of the season - Tm : Team - Pos : Position - G : Games - FG% : Field Goal Percentage - TRB : Total Rebounds Per game - AST : Assists Per game - STL : Steals Per game - BLK : Blocks Per game - PTS : Points Per game

The data features will be used for this application are as followings:

  1. For player information section, we used
    • The Pos column for Position
    • The Age column of the last year (from Season column) player in NBA for Age
    • The Season column (the rows for each year the player player) for Experience
  2. For interaction/filter section, we used
    • The Season column (the rows for each year the player played) for Year slide bar
    • The Tm column for Team check boxes, by using the unique values in the Tm column
  3. For the game information plot section, we used
    • The PTS column divided by the G column (the rows for each year the player player) for Point per game
    • The FG% column (the rows for each year the player player) for Shooting accuracy (this can be further breakdown into 3 Pointer shooting accuracy, 2 Pointer shooting accuracy and Free Throw shooting accuracy for furture development)
  4. For the game played plot section, we used:
    • The G column (the rows for each year the player player)
  5. For the player skill section, we used:
    • The PTS column divided by the G column (the rows for each year the player player) for Point per game
    • The TRB column divided by the G column (the rows for each year the player player) for Total Rebound per game
    • The AST column divided by the G column (the rows for each year the player player) for Assist per game
    • The STL column divided by the G column (the rows for each year the player player) for Steal per game
    • The BLK column divided by the G column (the rows for each year the player player) for Block per game

Primary research question and usage scenarios

The project aims to answer the following primary question:

How can we understand a NBA player's historical performance and skills during their career?

Below is an example of a usage scenario from a member of our target audience.

Tom is an enthusiastic NBA fan and would like to assess his loved NBA players based on the real performance data in previous NBA games. There are some options available on the internet to obtain the information, such as searching particular players through Wikipedia pages, checking historical data from NBA official website, sporting news websites, and The Basketball Reference website, etc. However, he found it was less effective to find his interested information in a well-organized way from these websites. Either he had to read paragraphs or a massive table to extract key data by himself, or there was less control to customize the visualizations to make his own analysis conclusions. By using NBA Player Stats Application Dashboard, now Tom can effectively and efficiently search players' NBA performance data and draw insights from the app. When he enters one of his loved player's name, Vince Carter, in the search section of the app, firstly, he will see Vince's head photo and basic information including position, age, years of experiences. Then he is able to see three plots to show the player's performance history, such as points per game, number of games played, and some skill indicators (Point per game, Total Rebound per game, Assist per game, Steal per game, and Block per game). The app has provided filtering features for him to select the particular year range and NBA team to output the above plots matching his desired conditions. This is a convenient way for Tom to play with the plotting and help him to analyze the player's performance in particular year(s) and in a specific team. By setting the year range between 1998 and 2008, it is clear to see Vince had earned above 20 points on average for most of the years; but after 2008, his capability to gain points had been dropped year after year till 2019 when he left the NBA although the total number of games he played annually did not decrease for the later years. In addition, the app indicates Vince had played with eight NBA teams during his career. It seems he had the best performance when he was a team member with Toronto Raptors (TOR) from 1998 to 2004.