Ball4life proposal 2

After consultations with Prof KAM, we felt that our previous scope of the proposal was too board and we might not be able to complete it in time. Therefore, we revised our project topic to analyse the game of James Harden.

Proposal

Poster

Application

Research Paper

Version 1

Version 2

<--- Go Back to Project Groups

INTRODUCTION

The National Basketball Association (NBA) is a men’s professional basketball league in North America. Comprising of 30 teams (29 in the United States and 1 in Canada), it is one of the 4 major professional sports leagues in the United States and Canada. Furthermore, it is widely regarded as the premier basketball league in the world, with many fans touting it as the hardest league with the world’s best players competing in it.

With an average NBA team being valued at US$1.9 billion, the NBA continues to generate earnings in the hundreds of millions. This, coupled with growing fan bases in China, Philippines, Vietnam and Europe ensures that the NBA will be able to provide entertaining and competitive basketball games for all their fans.

PROBLEM & MOTIVATION

Despite the wide array of sports analytics visualisations on the NBA, most are done with enterprise software such as Tableau and Microsoft PowerBi. However, there is little to no visualisation done on the NBA using self-implemented coding applications such as R Shiny. This presents a significant untapped opportunity for the team to develop a visualization web application that serves two main pain points:

1. Allow users to modify or expand on existing base code using R Shiny to visualise NBA applications.
2. Provide users with instant information not discernible by looking at raw figures on the NBA website.

On a macro scale, the visualisation is able to provide users who are interested in NBA analytics with an initial understanding on the different metrics being tracked by the NBA, and its effect on determining a player’s overall efficiency. Using James Harden, the industry dubbed statistical anomaly, the team aims to build a visualisation tool centred around James Harden’s shot selection, his field goal percentages and his performance when paired with different players.

OBJECTIVE

This visualisation aims to provide insights into the following:

1. How has James Harden's shot selection change over the years
2. How does his performance varies with different lineup paired with him
3. How effective is his performance compared with similar players

SELECTED DATABASE

The Data Sets we will be using for our analysis and for our application is listed below:

Dataset/Source	Rationale Of Usage
James Harden Shot Data Details	This will help us to properly plot out his shot distribution on the floor. This plot will help to understand his preferred and most effective scoring areas.
James Harden Lineup Per Game Statistics	This dataset will allow us to analyse how his performance varies when pair up with different teammates. This will allow his coach to determine the best lineup to pair him up with.
NBA Player Per Game Statistics	By having this information, we can carefully analyse and compare James Harden's performance in a few key areas with adjacent players with similar build and playing style.

BACKGROUND SURVEY

We did basic background research on some existing visualizations or dashboards we could drive inspirations from or make it better. Below are a few visuals we found:

References to Existing Visualizations	Key Takeaways
Title: NBA Shot Tracker Link: http://www.estherbaek.com/NBAShotTracker/	Feedback Clear illustration of insights on player’s field goal percentage Simple and intuitive UI to better engage audiences Too little information, would require more information to better understand individual’s game
Title: Player Analysis Link: https://shotline.peterbeshai.com/p/1415	Feedback Detailed breakdown of player’s efficiency rate in the 2 key offensive tools Leveraged on time-series data to better understand player’s growth or decline Too little information, would require more information to better understand individual’s game
Title: Best Draft Class Link: https://towardsdatascience.com/basketball-analytics-the-best-draft-class-13a6eac0cdb5	Feedback Could have included trend line to predict next round of best draft class There is a lack of filter to show what are the factors being considered in efficiency

PROPOSED STORYBOARD

With a clearer idea of what we want after looking at a few visualization examples, we came up with a few storyboard ideas.

Proposed Storyboard	Description
	Homepage This page will display a brief overview of our motivation and list down the key areas which the visualisation will address. This provides a first touch point for users and sets the user journey in an organised manner.
	Understanding Efficiency of James Harden This dashboard will consist of a shot distribution chart and bullet chart to compare James Harden's critical statistics with the players playing the same position. In addition, users will get to choose the season and which lineup to pair up with James Harden, to see how the statistics vary.
	Comparing James Harden to Top Shooting Guards This dashboard contains multiple charts. One key thing to take note is that the data used in this dashboard is the latest game data set. Hence, a web scraping script has been embedded in a button to allow users to click and scrape the latest game dataset. Firstly, it has a side by side bar chart to show the comparison of James Harden's critical statistics with adjacent elite shooting guard. User get to choose which elite shooting guard to compare with James Harden. There will also be a pie chart to brief show the shot distribution of James Harden over the past few games. Lastly, there will 4 different box plots to show James Harden's performance in critical statistics of a shooting guard.

CHALLENGES

Challenges	Mitigation Plan
Data Extraction from Web Sources/Pages.	Refine web scraping techniques to write automated scripts to scrape from NBA web pages
Data Cleaning and Transformation	Assign each team member with a data source to clean Organize team meeting to collate it into desired form
Lack of Experience with R and ShinyR	Go through available DataCamp courses Refers to online documentation and past examples to learn

TIMELINE

TOOLS/TECHNOLOGIES

COMMENTS

Feel free to leave us some comments so that we can improve!

No.	Name	Date	Comments
1.	Insert your name here	Insert date here	Insert comment here
2.	Insert your name here	Insert date here	Insert comment here
3.	Insert your name here	Insert date here	Insert comment here