Lecture Notes: T20 Cricket Data Analytics Series

Overview of the Project

• Review of T20 Cricket World Cup (England vs Pakistan)
• Project focus on cricket data analytics using T20 World Cup data
• Steps involved: scraping data, data cleaning and transformation, building dashboards in Power BI

Project Steps

1. Data Scraping

• Source: ESPNcricinfo website
• Tools: Web scraping techniques to extract relevant data
  • Use of Bright Data for scraping with proxy networks
  • Data types to capture:
    • Match results table
    • Detailed scorecards for batting and bowling
    • Player specific information

2. Data Cleaning & Transformation

• Python and Pandas: Tools used for data cleaning and transformation
• Objective: Transform JSON data into a flat CSV format for easier analysis in Power BI
• Steps include:
  • Renaming columns
  • Creating new columns (e.g., match ID, out/not out indicator)
  • Handling missing values and data types

3. Building Dashboards in Power BI

• Dashboard Features:
  • Categorized player selection (openers, anchors, fast bowlers)
  • Display player statistics: runs, strike rate, batting average
  • Filtering capabilities for selection criteria
  • Visualizations like scatter plots for performance comparisons

Problem Statement

• Assembling a cricket team to defeat Planet Sporta
• Required team performance metrics:
  • Average runs scored: 180
  • Runs to defend: 150

Parameters for Player Selection

Openers

• Criteria:
  • Batting average
  • Strike rate
  • Boundary percentage
• Target: 50 runs in the first 5 overs

Middle Order (Anchors)

• Criteria:
  • Ability to shift gears and bat for a longer duration
  • Higher batting average
  • Average balls faced

Finishers

• Criteria:
  • Ability to chase down and stabilize innings
• Preferably all-rounders with batting focus

All-rounders

• Criteria:
  • Should be capable of hitting hard and also bowling effectively

Fast Bowlers

• Criteria:
  • Bowling economy below 7
  • Wickets taken every 16 balls

Data Collection Process

• Bright Data: Utilized for efficient data collection
• Created multiple collectors for different data types
• Example of data extraction process shown using JavaScript code for web scraping

Data Modeling in Power BI

• Establishing relationships between tables based on match IDs and player names
• Creating DAX measures for calculations (e.g., total runs, innings batted)
• Calculated columns for derived metrics (e.g., boundary runs)

Dashboard Creation

• Mock-up provided for initial design
• Creating visuals based on player categories and performance metrics
• Emphasized on the importance of visualization aesthetics and layout

Final Team Selection

• Analysis of player performance using the created dashboard
  • Selected players based on statistical performance and roles needed
  • Emphasized the importance of pairing players effectively for optimal results

Challenge and Further Learning

• Participants given an exercise to improve the dashboard and add new insights
• Encouraged to share progress on platforms like LinkedIn for networking and visibility
• Resources and codes provided for further exploration and practice

Conclusion

• The project aims to provide hands-on experience in cricket data analytics through practical application of data scraping, cleaning, and visualization techniques.
• Encouraged students to leverage the knowledge gained for real-world applications and potential career opportunities.