dawn gazelle
Q&A
When will cgbot be finished/operable?
Probably not anytime soon.
cgbot, as a project, has just started. I've talked about it in #🕹・game-discussion a tad. I don't plan on working on it full-time, and this is a massive project. Progress will be slow, and if I ever do finish it'll likely be in around a year. This thread exists for me to post about progress, and to perhaps ask for help solving some issues that may arise.
Will there be mod support?
If the mod creator implements it themselves.
There will be a documented API for mod creators to add their own custom handlers for their items. This goes for pretty much everything addable into the game, like Jokers, consumable types, etc.
What is the goal of cgbot?
To win, endless will not be a primary goal.
The idea is to be as consistent as possible at getting to Ante 8 and beating the showdown blind. Of course, the bot will be able to go into endless, but it won't be designed for endless runs and will likely never get past around Ante 12. cgbot will be able to play on any deck, on any stake, and each will be equally prioritized (most development will likely happen on Blue Deck White Stake, just to be simple)
Will cgbot support challenges?
Probably not.
It might be possible for cgbot to beat some challenges, but it won't be specifically designed to beat challenges.
On a technical level, what will cgbot be like?
I don't know.
This isn't set in stone yet. I don't plan on following any existing reserach, nor using any sort of neural network and similar self-learning models. I plan on developing all my methods by myself. I'll have to build up a little more before I decide how I want to do this. I'm going in partially blind, only guided by an earlier (yet scrapped) attempt at a Balatro bot.