Pokémon Attack Analysis

An exploratory analysis of Pokémon “Attack” values using a dataset of 1,008 Pokémon. This project computes summary statistics, visualizes the distribution and sampling behavior of the Attack stat, and evaluates normality assumptions.

Table of Contents

1. Project Overview
2. Features & Outputs
3. Prerequisites
4. Installation
5. Usage
6. Script Breakdown
7. Interpreting the Results
8. Extending & Customizing
9. Data Source & Citations
10. License

---

Project Overview

This analysis focuses on the “Attack” statistic from a Pokémon dataset (pokemon.csv, 1,008 entries). We:

• Load and inspect the full Pokémon table.
• Visualize the distribution of Attack values.
• Compute mean and standard deviation for the population.
• Count how many Pokémon fall within 1, 2, and 3 standard deviations of the mean.
• Assess normality via empirical proportions vs. the 68–95–99.7 rule.
• Draw a random sample of 200 Attack values (with replacement), visualize and summarize it.
• Compare sample vs. population to discuss sampling accuracy.

All code is contained in Pokemon_Attack_Analysis.R, and the key visualizations are saved as attack-population.png and attack-sample.png.

---

Features & Outputs

• Population Histogram (attack-population.png): Attack distribution across all 1,008 Pokémon.
• Population Summary: Mean ≈ 77.17, SD ≈ 29.76.
• Empirical Rule Counts:
  • Within 1 SD: 669 Pokémon (≈ 66.3%).
  • Within 2 SD: 980 Pokémon (≈ 97.2%).
  • Within 3 SD: 1,007 Pokémon (≈ 99.9%).
• Normality Assessment: Visual and numeric comparison to expected ~68–95–99.7%.
• Sampled Histogram (attack-sample.png): Distribution of 200 random draws of Attack.
• Sample Summary: Sample mean & SD reported, with discussion of sampling error.

---

Prerequisites

• R (≥ 4.0) installed on your system.
• RStudio (optional) for interactive development.
• Internet access to install packages and download the dataset.

R Packages

• ggplot2
• dplyr

The analysis script will auto-install any missing packages.

---

Installation

1. Clone this repository:

   git clone https://github.com/yourusername/pokemon-attack-analysis.git
   cd pokemon-attack-analysis

2. Place pokemon.csv (the full Pokémon dataset) in the project root.

No further build steps are needed.

---

Usage

1. Open R or RStudio, set the working directory to the project folder:

   setwd("path/to/pokemon-attack-analysis")

2. Run the analysis script:

   source("Pokemon_Attack_Analysis.R")

3. View the two plots and console output:

   • attack-population.png
   • attack-sample.png And review the printed summary statistics and counts.

---

Script Breakdown

• Load Data

  pokemon <- read.csv("pokemon.csv")

• Population Visualization

  • Histogram of Attack with binwidth = 10.
  • Save as attack-population.png.

• Population Summary

  • Compute meanAttack, sdAttack.
  • Count Pokémon within ±1, ±2, ±3 SD.

• Normality Check

  • Compare empirical proportions to 68–95–99.7 rule.

• Sampling

  • Draw 200 Attack values with replacement.
  • Histogram with same binwidth, save as attack-sample.png.
  • Compute sample mean & SD.

• Interpretation

  • Print conclusions regarding normality and sampling adequacy.

---

Interpreting the Results

• The population histogram is right-skewed (long tail of high-Attack Pokémon), not perfectly normal.
• Within 1 SD: ~ 66.3% vs. expected 68% — close but slightly low.
• Within 2 SD: ~ 97.2% vs. expected 95% — slightly higher.
• Sampling of 200 shows variability in mean and SD; small sample sizes yield sampling error.
• To better approximate the population, consider larger samples (e.g., ~ 600).

---

Extending & Customizing

• Other Stats: Repeat for Defense, HP, etc.
• Binwidth Tuning: Adjust histogram binwidth.
• Statistical Tests: Apply Shapiro–Wilk or Kolmogorov–Smirnov.
• Multiple Samples: Automate repeated sampling to study sampling distribution.
• Interactive Plots: Convert to plotly for hover details.

---

Data Source & Citations

• Pokémon Dataset: pokemon.csv — contains stats for 1,008 Pokémon.
• R & Packages: R Core Team (2023); Hadley Wickham et al. for ggplot2, dplyr.

---

License

This project is released under the MIT License. See LICENSE for full details.