Uploaded June 2021 | Updated September 2026, 2 hours ago
I'm finally back with more Pokemon AI, this time focusing on Pokemon Showdown! I made four different Showdown bots, each more complex than the last, and in this video I not only explain how I made them, but pit them against each other in a tournament to see which bot reigns supreme! I also have a big announcement at the end, so be sure to stick around for that!
Blog / Transcript: remptongames.com/2021/06/27/programming-ai-for-pokemon-showdown-bot-battle-royale
Chapters :
00:00 Introduction
01:12 Why Pokemon Showdown?
01:52 Getting Started
03:23 Rules-based bot
04:41 Minimax bot
07:55 Problems with Minimax
10:11 Why not machine learning?
11:23 Bot tournament!
13:44 Answering Emerald AI Questions
16:20 Future Project Announcement
17:19 Conclusion
Github Repo: github.com/RemptonGames/Pokemon-Showdown-Agent
Algorithms Explained – minimax and alpha-beta pruning - Sebastian Lague - youtube.com/watch?v=l-hh51ncgDI
Poke-env Github Repository: github.com/hsahovic/poke-env
Poke-env docs: poke-env.readthedocs.io/en/stable/index.html
Pokemon Battling Research Papers:
Showdown AI Competition - julian.togelius.com/Lee2017Showdown.pdf
Competitive Deep Reinforcement Learning over a Pokémon Battling Simulator -ieeexplore.ieee.org/abstract/document/9096092
Optimal Battle Strategy in Pokemon using Reinforcement Learning - https://web.stanford.edu/class/aa228/reports/2018/final151.pdf
Heuristics for sleep and heal in combat - gram.cs.mcgill.ca/papers/xu-16-heuristics.pdf
IMPROVING COMPANION AI IN SMALL-SCALE ATTRITION GAMES - file:///C:/Users/Caleb/AppData/Local/Temp/c8522750-3795-40dc-8f77-ac84c9e6d4a3.pdf
Implementation and Evaluation of Information Set Monte Carlo Tree Search for Pokemon - eprints.lib.hokudai.ac.jp/dspace/bitstream/2115/72345/1/ihara-smc2018.pdf
Ontology-based Knowledge System and Team Verification Tool for Competitive Pokemon - journals.flvc.org/FLAIRS/article/view/128544/130039
Comparison of Artificial Intelligence Algorithms for Pokémon Battles - https://odr.chalmers.se/bitstream/20.500.12380/300015/1/Linus%20Norstr%C3%B6m.pdf
Gotta Train ‘em all: Learning to Play Pokemon Showdown with Reinforcement Learning - https://cs230.stanford.edu/projects_fall_2018/reports/12447633.pdf
I'm finally back with more Pokemon AI, this time focusing on Pokemon Showdown! I made four different Showdown bots, each more complex than the last, and in this video I not only explain how I made them, but pit them against each other in a tournament to see which bot reigns supreme! I also have a big announcement at the end, so be sure to stick around for that!
Blog / Transcript: remptongames.com/2021/06/27/programming-ai-for-pokemon-showdown-bot-battle-royale
Chapters :
00:00 Introduction
01:12 Why Pokemon Showdown?
01:52 Getting Started
03:23 Rules-based bot
04:41 Minimax bot
07:55 Problems with Minimax
10:11 Why not machine learning?
11:23 Bot tournament!
13:44 Answering Emerald AI Questions
16:20 Future Project Announcement
17:19 Conclusion
Github Repo: github.com/RemptonGames/Pokemon-Showdown-Agent
Algorithms Explained – minimax and alpha-beta pruning - Sebastian Lague - youtube.com/watch?v=l-hh51ncgDI
Poke-env Github Repository: github.com/hsahovic/poke-env
Poke-env docs: poke-env.readthedocs.io/en/stable/index.html
Pokemon Battling Research Papers:
Showdown AI Competition - julian.togelius.com/Lee2017Showdown.pdf
Competitive Deep Reinforcement Learning over a Pokémon Battling Simulator -ieeexplore.ieee.org/abstract/document/9096092
Optimal Battle Strategy in Pokemon using Reinforcement Learning - https://web.stanford.edu/class/aa228/reports/2018/final151.pdf
Heuristics for sleep and heal in combat - gram.cs.mcgill.ca/papers/xu-16-heuristics.pdf
IMPROVING COMPANION AI IN SMALL-SCALE ATTRITION GAMES - file:///C:/Users/Caleb/AppData/Local/Temp/c8522750-3795-40dc-8f77-ac84c9e6d4a3.pdf
Implementation and Evaluation of Information Set Monte Carlo Tree Search for Pokemon - eprints.lib.hokudai.ac.jp/dspace/bitstream/2115/72345/1/ihara-smc2018.pdf
Ontology-based Knowledge System and Team Verification Tool for Competitive Pokemon - journals.flvc.org/FLAIRS/article/view/128544/130039
Comparison of Artificial Intelligence Algorithms for Pokémon Battles - https://odr.chalmers.se/bitstream/20.500.12380/300015/1/Linus%20Norstr%C3%B6m.pdf
Gotta Train ‘em all: Learning to Play Pokemon Showdown with Reinforcement Learning - https://cs230.stanford.edu/projects_fall_2018/reports/12447633.pdf
