Was really looking forward to your game as you used neural network as well !
Unfortunately, I am not fond of card games, so it is hard for me to get into your entry :/
In short, the AI seemed to behave quite nicely, so this is very nice.
I can really imagine the amount of time you spent just training the model !
What indicators did you use to see if they were indeed improving in the game ?
Did you used advsersarial stratefy when they had to compete against each other ?
And for the cheater part, did you used a different model for each cheat ability, or did you have one big model for all opponents and just switched off some outputs ?
Play game
Cheater's Inn's itch.io pageResults
Criteria | Rank | Score* | Raw Score |
Presentation | #11 | 4.000 | 4.000 |
Originality | #12 | 4.000 | 4.000 |
Overall | #13 | 3.711 | 3.711 |
Fun | #32 | 3.133 | 3.133 |
Ranked from 15 ratings. Score is adjusted from raw score by the median number of ratings per game in the jam.
Comments
Yeah, I settled on a card game early on so that I could spend most of my time on training the AIs instead of on game design. (Which isn't normally a good idea but that's what I wanted to get out of this jam.) I had ten groups of neural networks play thousands of games against eachother each round, replacing the worst scoring ones with mutations and combinations of the better ones, and repeating that for thousands of rounds. Each cheat ability has its own pool of neural network because I wanted them to learn to use it to their advantage. For example the Spy can always see all players' cards and I wanted its neural network to be trained to use that information. But structurally all NNs have the same shape, the main difference lies in their starting hand; so one cheater always starts with a Twelve of Hearts, but other NNs can learn to pick it up if it's on the table.
The code I wrote for it is open source, if your interested: https://github.com/SLiV9/MegaNaiadsTrainer
Looks very climatic, almost like a Gwent ;p And Idea behind is also quite interesting, congrats ;)
Happy to hear it works now.
Whenever you have a winning or a losing hand, you can click on "Accuse" below any stranger's name and you'll get a hand of six accusation cards. If you click on those cards, it gives a description of how that type of cheater cheats. This should help knowing what behaviour to look out for, and you can click "Do Not Accuse" and then "No Accusations" to go back to playing. Once you think you've identified a stranger, you can accuse them (the next time you have a winning or losing hand) by clicking on the accusation card and then clicking "Accuse" next to the cards. It you're right, they reveal themselves. If you're wrong, you lose a life. Hope that helps!
I would like to rate it, but as soon as I get to actually playing the game crashes for no reason.
Really cool concept!! The dialogue and game play is great! I loooove the art and color palette too -excellent work!! :D
Not that I saw, the application would just force close itself so unless it spits an error log into its fits somewhere I can check I'm not sure what happened.
All I know is that it only happened at the start or end of a round, maybe 40-ish percent of the time. And after 3 or 5 times trying the tutorial it seemed to always crash at the end of explaining the rules. If there's anything about my specs that you think would be helpful to know, lemme know!
Hey! This one wound up working, so I was able to play!
I'm glad you were able to track down the crashing bug, you've definitely made a really interesting game. With some additional polish I could honestly see this concept turn into a cool indie game.
As is, it definitely stands high on my list of jam entries. The aesthetic, along with even the little satisfying details like the flicking and shuffling of cards, is also really enjoyable.
Really loved the extra mile put into setting up the intro. The game doesn't feel too stable though (getting a lot of Node:1209 instead of revealing the numbers). It also crashed/closed a few times for me. It does seem to have a really cool idea though!
I really liked the idea of your entry thematic wise, and the art is superb! The only thing I didn't quite understand the purpose of discovering the people originally when they get up to leave. I didn't leave a rating earlier, apologies, but your entry was one of the first I clicked on as really close to ours and love the idea!
Superb. My favourite entry so far. I found it really tricky to figure out who was cheating (except the innkeep) but it was really fun to try and keep track of the different styles of cheating.
Writing and introduction were also really well done!
Overall I love it, I think it's worth spending time on it if there's anything you wanted to add, it has real potential.
Wow, it is quite impressive techwise. It looks like you have used PyTorch to create the competitors.
My only complaint would be the lack of feedback. Sound, animation, something to indicate whether I have done the right thing and help me understand what has happened.
Great job nonetheless!
Leave a comment
Log in with itch.io to leave a comment.