pokeLab
Competitive pokémon battle-oriented research, correlation studies and more. Any collaborations are welcome! Mainly focused on Smogon's Showdown community. https://www.smogon.com/stats/ is a nice and incredibly huge database to work with.
Team Generator
Generates teams from usage data.
Optimiser / damage lab
Optimal spread for defensive Effort Values (EVs) for a given base power move. Further stochastic analysis should be performed according to base power distributions (e.g. base power of ghost-type moves in OU metagame have a X distribution, so Wobbuffet's optimal EV spread in OU metagame will be A, while in Uber metagame will be B). This will take big efforts to be accurately studied. ¿Is it better to average all possible incoming damage scores? ¿Perhaps is it better to have an EV spread that minimises noise in expected damage?
Correlation studies
¿Are steel types significantly more defensive than the other ones? ¿Are there any correlations between base stats among any group of mons? ¿Could a genealogy be built from these data?
Genealogy
"genetic" distance from pokemon taking into account their typing, stats, moveset and more. From phenotype (stats, typing, etc), try to infer genomes.
Generator
Generative adversarial network, which generates complete pokemon (and possibly complete evolutionary families), including sprites, moveset, base stats, etc. image set here: https://github.com/smogon/sprites/tree/master/newsrc/dex
BattleBot
A pokemon showdown bot using Selenium and firefox webdriver.