This dataset provides detailed information on 1007 Pokémon from the popular mobile game Pokémon GO. It includes a wide range of attributes such as base stats, move sets, rarity, and acquisition methods. The data was collected using the RapidAPI Pokémon GO API, offering researchers and data enthusiasts a rich resource for analysis, machine learning projects, and game strategy development.

Key features of this dataset include:

• Comprehensive coverage of 1007 Pokémon
• 24 attributes for each Pokémon, including battle stats, type, and rarity
• Information on acquisition methods (wild, egg, raid, etc.)
• Move set details for both fast and charged moves
• Game mechanics data such as capture and flee rates

This dataset is ideal for:

• Analyzing Pokémon strengths and weaknesses
• Developing machine learning models for Pokémon classification or prediction
• Studying game balance and design in Pokémon GO
• Creating tools for players to optimize their gameplay strategies

Whether you're a data scientist, game developer, or Pokémon enthusiast, this dataset offers a wealth of information to explore and analyze the world of Pokémon GO.

🔍 The Case of the Disguised Villains: Predicting Team Rocket with Data

In the bustling world of Kanto, where Pokémon battles shape destinies, crime lurks in the shadows. Detective Kotso, the sharpest mind in Pokémon crime investigations, has been tasked with an urgent mission. The mayor suspects that Team Rocket has infiltrated the city, disguising themselves as ordinary citizens.

But Kotso doesn’t work alone—he relies on you, a brilliant data scientist, to uncover the truth. Your job? Analyze the data of 5,000 residents to predict which of the 1,000 unclassified individuals are secretly part of Team Rocket.

Can you spot the hidden patterns? Can Machine Learning crack the case where traditional detective work fails? The fate of Kanto depends on your skills.

📊 Dataset Structure & Features

This dataset holds the key to exposing Team Rocket’s operatives. Below is a breakdown of the features at your disposal:

🕵️ Can You Crack the Case?

This dataset is not just about numbers—it’s a criminal investigation. Hidden patterns lurk beneath the surface, waiting to be uncovered.

• Are certain Pokémon types more common among Team Rocket members?
• Do suspicious financial transactions hint at illegal activities?
• Does their battle strategy betray their allegiance?

This isn’t just another classification task—it’s a race against time to stop Team Rocket before they take control of Kanto!

Detective Kotso is counting on you. Will you rise to the challenge? 🕵️‍♂️🔎

🔎 10 Key Questions & Suggested Analysis Techniques

1️⃣ Do certain Pokémon types indicate suspicious behavior?
- 📈 Graph: Stacked bar chart comparing Pokémon type distribution between Rocket & non-Rocket members.
- 🎯 Test: Chi-square test for correlation.

2️⃣ Is economic status a reliable predictor of criminal affiliation?
- 📊 Graph: Box plot of debt and economic status per Team Rocket status.
- 🏦 Test: ANOVA test for group differences.

3️⃣ Do Team Rocket members have a preference for specific PokéBalls?
- 🎨 Graph: Heatmap of PokéBall usage vs. Team Rocket status.
- ⚡ Test: Chi-square test for independence.

4️⃣ Does a high battle win ratio correlate with Team Rocket membership?
- 📉 Graph: KDE plot of win ratio distribution for both classes.
- 🏆 Test: T-test for mean differences.

5️⃣ Are migration patterns different for Team Rocket members?
- 📈 Graph: Violin plot of migration counts per group.
- 🌍 Test: Mann-Whitney U test.

6️⃣ Do Rocket members tend to avoid charity participation?
- 📊 Graph: Grouped bar chart of charity participation rates.
- 🕵️‍♂️ Test: Fisher’s Exact Test for small sample sizes.

7️⃣ Do Rocket members disguise themselves in certain professions?
- 📊 Graph: Horizontal bar chart of profession frequency per group.
- 🕵️‍♂️ Test: Chi-square test for profession-Team Rocket relationship.

8️⃣ Is there an unusual cluster of Rocket members in specific cities?
- 🗺 Graph: Geographic heatmap of city distributions.
- 📌 Test: Spatial autocorrelation test.

9️⃣ How does badge count affect the likelihood of being a Rocket member?
- 📉 Graph: Histogram of gym badge distributions.
- 🏅 Test: Kruskal-Wallis test.

🔟 **Are there any multi-feature interactions that reve...

This dataset provides detailed information for Pokémon from 1 to 1045, as listed in the National Pokédex. It includes fundamental Pokédex entries such as their names, types, and physical attributes, alongside more in-depth data like move sets, type effectiveness, abilities with full descriptions, and battle strategies sourced from Smogon. Additionally, the dataset contains brief descriptions from Bulbapedia. A distinct text corpus file is also included, offering a textual representation for each Pokémon, compiled from all the details present in the main Pokédex file.

Columns

The main Pokémon file features 56 columns, providing extensive details for each creature. Key columns include: * pokédex number: The official National Pokédex identification number. * name: The English name of the Pokémon. * japanese name: The Japanese name of the Pokémon. * generation: The generation number the Pokémon originates from. * status: Indicates if the Pokémon is Legendary. * species: The specific species of the Pokémon. * type number: How many elemental types the Pokémon possesses. * type 1: The primary elemental type. * type 2: The secondary elemental type, if applicable. * height: The Pokémon's height in metres. * weight: The Pokémon's weight in kilograms. * abilities number: The count of abilities it can have. * total points: The sum of all base stats. * stats: Individual columns for key battle statistics: HP, attack, defence, special attack, special defence, and speed. * catch rate: The Pokémon's catch rate. * base friendship: The base friendship value. * base experience: The base experience yield. * growth rate: The growth rate category. * egg type number: The number of egg groups it belongs to. * egg type 1: The primary egg group. * egg type 2: The secondary egg group, if applicable. * percentage male: The likelihood of the Pokémon being male. * egg cycles: The number of steps required to hatch an egg. * type effectiveness: Columns detailing effectiveness against various types (e.g., normal, fire, water, grass, electric, flying, ground, rock, fighting, psychic, dark, ghost, dragon, ice, fairy, poison, bug, steel). * Smogon description: Battle strategies primarily from SM Pokédex, or other generations if more relevant. * Bulba description: Initial sentences from the Pokémon's Bulbapedia page. * moves: A dictionary detailing moves the Pokémon learns by levelling up, including name, type, damage type, power, accuracy, PP, level learned, secondary effect chance, and description. * ability 1, ability 2, hidden ability: The names of the Pokémon's abilities. * ability 1 description, ability 2 description, hidden ability description: Descriptions for each of the Pokémon's abilities.

The accompanying Poké corpus file contains a text corpus for each Pokémon, generated by consolidating all the information from the Pokédex file.

Distribution

This dataset encompasses information for Pokémon numbered 1 through 1045. The primary Pokémon data file contains 56 distinct columns for each entry. While specific row counts are not provided, there are 1045 unique Pokémon entries detailed. Data files are typically provided in CSV format.

Usage

This dataset is ideally suited for a variety of applications, particularly in the fields of artificial intelligence, machine learning, and data analysis related to gaming and entertainment. * Building AI Chatbots: Useful for creating conversational agents, such as a Pokémon chatbot, through retrieval-augmented generation (RAG) pipelines. * Game Development: Provides extensive data for developers creating Pokémon-inspired games or applications. * Data Analysis: Researchers and enthusiasts can analyse Pokémon stats, moves, and abilities for competitive strategy or general insights. * Natural Language Processing (NLP): The text corpus can be used for text generation, entity recognition, and other NLP tasks related to Pokémon lore.

Coverage

The dataset covers Pokémon from number 1 to 1045 in the National Pokédex. Its scope is global, providing information relevant to all regions where Pokémon are known. There are no specific notes on data availability for certain groups or years beyond the stated Pokédex range.

License

CC BY-SA

Who Can Use It

• Data Scientists and AI/ML Developers: For training models, building recommendation systems, or developing chatbots and other AI applications using the detailed Pokémon attributes and text corpus.
• Game Developers: To integrate accurate and detailed Pokémon information into their projects.
• Researchers: For academic studies on game design, character attributes, or data structures in entertainment.
• Pokémon Enthusiasts and Community Developers: For fan-made applications, wikis, or statistical analyses

Overview

PokemonGo is a mobile augmented reality game developed by Niantic inc. for iOS, Android, and Apple Watch devices. It was initially released in selected countries in July 2016. In the game, players use a mobile device's GPS capability to locate, capture, battle, and train virtual creatures, called Pokémon, who appear on the screen as if they were in the same real-world location as the player.

Dataset

Dataset consists of roughly 293,000 pokemon sightings (historical appearances of Pokemon), having coordinates, time, weather, population density, distance to pokestops/ gyms etc. as features. The target is to train a machine learning algorithm so that it can predict where pokemon appear in future. So, can you predict'em all?)

Feature description

• pokemonId - the identifier of a pokemon, should be deleted to not affect predictions. (numeric; ranges between 1 and 151)
• latitude, longitude - coordinates of a sighting (numeric)
• appearedLocalTime - exact time of a sighting in format yyyy-mm-dd'T'hh-mm-ss.ms'Z' (nominal)
• cellId 90-5850m - geographic position projected on a S2 Cell, with cell sizes ranging from 90 to 5850m (numeric)
• appearedTimeOfDay - time of the day of a sighting (night, evening, afternoon, morning)
• appearedHour/appearedMinute - local hour/minute of a sighting (numeric)
• appearedDayOfWeek - week day of a sighting (Monday, Tuesday, Wednesday, Thursday, Friday, Saturday, Sunday)
• appearedDay/appearedMonth/appearedYear - day/month/year of a sighting (numeric)
• terrainType - terrain where pokemon appeared described with help of GLCF Modis Land Cover (numeric)
• closeToWater - did pokemon appear close (100m or less) to water (Boolean, same source as above)
• city - the city of a sighting (nominal)
• continent (not always parsed right) - the continent of a sighting (nominal)
• weather - weather type during a sighting (Foggy Clear, PartlyCloudy, MostlyCloudy, Overcast, Rain, BreezyandOvercast, LightRain, Drizzle, BreezyandPartlyCloudy, HeavyRain, BreezyandMostlyCloudy, Breezy, Windy, WindyandFoggy, Humid, Dry, WindyandPartlyCloudy, DryandMostlyCloudy, DryandPartlyCloudy, DrizzleandBreezy, LightRainandBreezy, HumidandPartlyCloudy, HumidandOvercast, RainandWindy) // Source for all weather features
• temperature - temperature in celsius at the location of a sighting (numeric)
• windSpeed - speed of the wind in km/h at the location of a sighting (numeric)
• windBearing - wind direction (numeric)
• pressure - atmospheric pressure in bar at the location of a sighting (numeric)
• weatherIcon - a compact representation of the weather at the location of a sighting (fog, clear-night, partly-cloudy-night, partly-cloudy-day, cloudy, clear-day, rain, wind)
• sunriseMinutesMidnight-sunsetMinutesBefore - time of appearance relatively to sunrise/sunset Source
• population density - what is the population density per square km of a sighting (numeric, Source)
• urban-rural - how urban is location where pokemon appeared (Boolean, built on Population density, <200 for rural, >=200 and <400 for midUrban, >=400 and <800 for subUrban, >800 for urban)
• gymDistanceKm, pokestopDistanceKm - how far is the nearest gym/pokestop in km from a sighting? (numeric, extracted from this dataset)
• gymIn100m-pokestopIn5000m - is there a gym/pokestop in 100/200/etc meters? (Boolean)
• cooc 1-cooc 151 - co-occurrence with any other pokemon (pokemon ids range between 1 and 151) within 100m distance and within the last 24 hours (Boolean)
• class - says which pokemonId it is, to be predicted. Data dump ------------

All pokemon sightings (in JSON file, without features) can be found in Discussion "Datadump"

📖 Description:

Step into the Golden Era of Pokémon and relive the magic from Kanto to Hoenn — the first three generations that defined childhoods, sparked rivalries, and ignited a global phenomenon. Whether you started with Bulbasaur, bonded with Cyndaquil, or battled through Hoenn with Mudkip, this dataset is a tribute to the most iconic 386 Pokémon that shaped an era.

As a lifelong Pokémon fan and a data enthusiast, I created this dataset to bridge nostalgia with numbers, giving fellow data scientists, analysts, and fans a structured, rich, and clean dataset that’s more than just stats, it's a journey through the classics.

🔍 What's Inside:

A complete profile for 386 Pokémon from Gen I–III (Kanto, Johto, Hoenn) featuring:

• Battle Stats

  • Base HP: Health Points
  • Base Attack: Physical attack strength
  • Base Defense: Resistance to physical attacks
  • Base Special Attack: Strength of special moves (e.g., Flamethrower)
  • Base Special Defense: Resistance to special attacks
  • Base Speed: Determines move order in battle
  • Total Base Stats: Sum of all base stats (useful for comparing overall strength)

• Type Information

  • Type 1: Primary elemental type (e.g., Fire, Water, Grass)
  • Type 2: Secondary type if applicable (e.g., Flying, Poison)

• Evolution Chain

  • Evolution Stage: Stage in the evolution line (1 = basic, 2 = mid, 3 = final)
  • Hidden Ability: Special ability that’s not always visible or available by default

• Abilities

  • Describes special traits (passive skills) a Pokémon can have in battle, including its Hidden Ability which can only be obtained under certain conditions (like via breeding or events)

• Breeding Data

  • Gender Ratio: Distribution of male/female genders (or Genderless)
  • Catch Method: How the Pokémon is typically acquired (wild, starter, evolution, etc.)
  • Base Friendship: Starting friendship level, used in evolution or move effectiveness

• Capture Mechanics

  • Capture Rate: Integer from 3–255, higher means easier to catch
  • Is Legendary: TRUE/FALSE flag to identify rare, one-of-a-kind Pokémon

• Visual Traits (Great for Clustering & Modeling)

  • Color: UI categorization from Pokédex (used in official classifications)
  • Shape: Visual silhouette group (bipedal, quadruped, etc.)
  • Height (m): Official height in meters
  • Weight (kg): Official weight in kilograms

• Meta Info

  • ID: Unique internal identifier (can be dropped in most analyses)
  • Name: Pokémon's English name
  • Generation: Number indicating the generation (1 = Kanto, etc.)
  • Region: Region the Pokémon originated from (Kanto, Johto, Hoenn)
  • Number: National Pokédex number (missing in a few cases, may need fixing)

🛠️ Suggested Projects:

• Predict a Pokémon’s type or legendary status using ML models
• Cluster Pokémon by battle roles (tanks, sweepers, etc.)
• Build a team recommendation engine
• Explore evolution patterns through stats and stages
• Visualize stat distributions across generations

📚 Source & Curation

Data compiled from trusted community resources:

• Pokémon Database
• Bulbapedia

If you're a fan of Pokémon and data, I hope this brings you joy and insights. Let’s catch 'em all — one data point at a time.