Robot gigachad
Open Source Python robot by entropicdrifter
My fourth bot.
Gigachad (originally named hannibal) was number 1 on the Alpha Board from November 2022 till December 2023 (to my recollection at least). It was designed to be a deep improvement on seven-of-nine. Where Seven cleaned up and refined the logic for we-are-borg, the idea of Gigachad was 2-fold: simulate the enemy's turn as well as possible and use recursion in order to resolve movement hierarchically with low-health retreats receiving the highest priority.
There are likely still a few bugs left in the implementation, at this point likely due to my sloppy architectural non-philosophy of just hacking on fixes to one problem at a time. My next bot will have a much more intentional design now that I have some experience at thinking through the problems of more advanced bots.
I'm open-sourcing gigachad now so that hopefully other bots will grow stronger off of training against it.
Battles
The Alpha Board
11/21 07:25 UTC | #13907 | (+0) | neuralbot2-6h | 752 | |
11/21 07:10 UTC | #13906 | (+0) | neuralbot1-1h | 538 | |
11/21 06:55 UTC | #13904 | (+0) | neuralbot2-24h | 956 | |
11/20 19:10 UTC | #13901 | (+7) | crystal | 1368 | |
11/20 17:10 UTC | #13897 | (+1) | neuralbot4-3h | 1102 | |
11/20 04:56 UTC | #13891 | (-2) | poggers | 2188 |