How We Trained a Genetic Algorithm to Fly in Paper Mario: TTYD @Malleo
How We Trained a Genetic Algorithm to Fly in Paper Mario: TTYD  @Malleo
Uploaded November 2021 | Updated September 2026, 2 weeks ago
This video discusses how we reverse-engineered the flight physics to run frame-by-frame simulations, as well as how we made use of a genetic algorithm to achieve optimal plane flight input sequences.

View the code used for this project here: github.com/malleoz/ttydPlaneSim
Jdaster's TTYD Utils repository: github.com/jdaster64/ttyd-utils

TwoMinutePaper's video: youtube.com/watch?v=Lu56xVlZ40M
The Open AI Paper: openai.com/blog/emergent-tool-use

Special Thanks to:
mmtrebuchet - Discord: mmtrebuchet 7937
Monado - twitter.com/Monado_ and twitch.tv/monado
Sjorec - twitter.com/SjorecSRL and twitch.tv/sjorec
Sway - twitter.com/Sway_NC and twitch.tv/swaync
Muz - twitter.com/MuzYoshi and twitch.tv/muzyoshi

Interested in TTYD speedrunning? Join the community Discord here: discordapp.com/invite/YB23azC
_____________________________
Twitter: twitter.com/tasmalleo
Twitch: twitch.tv/tasmalleo
Join my Discord! discord.gg/VRk59mY
Intro/Outro by TheSneakySpy: twitter.com/TheSneakySpy and youtube.com/TheSneakySpy
Check out Mario Kart Wii TASes: mkwtas.com
How We Trained a Genetic Algorithm to Fly in Paper Mario: TTYD[MKW TAS] DS Yoshi Falls Lap 1 - 19466 by Malleo(TAS) Paper Mario TTYD Levitate% - 47:42.53 by Malleo1x Goombella Buffer to Bypass Loading ZonesKinetolic Gaming vs Forgotten Legends - War HighlightsSonic Heroes (TAS) - Seaside Hill (Team Sonic) in 1:23.44 {WR} [16:9]
Malleo |

How We Trained a Genetic Algorithm to Fly in Paper Mario: TTYD

SHARE TO X SHARE TO REDDIT SHARE TO FACEBOOK WALLPAPER