A real gem of gibberish interpretation by the thread: https://i.imgur.com/AWTQ0bM.jpg
Generating Magic cards using deep, recursive neural networks
31–40 of 69 posts
Re: Generating Magic cards using deep, recursive neural networks
#32Magic is an interesting game just from an text analytics perspective. Using a source like mtgjson you get a reasonably sized corpus of text that looks natural enough for a human to comfortably grok but sufficiently templated to apply even naive methods and acquire awesome results. I built a tool that uses n-grams to find cards similar to each other. The result is comparable to manual suggestions done by users but tak…
I am extremely interested to see anything you have public on this project - both from a technical and deckbuilder's perspective!
Re: Generating Magic cards using deep, recursive neural networks
#33> Shring the Artist > 2BB > Legendary Creature - Cat > Flying > Whenever you cast a spell, you may return target creature card from your graveyard to your hand. > 2/2 That's actually a fairly plausible card, in terms of mechanics. It's thematically appropriate for black, it's about the right cost, it's powerful but specialized (returns creature cards only, and only to hand rather than to play), and it's legendary to…
Not going to lie, I would totally play that in my junk reanimator EDH deck...
Re: Generating Magic cards using deep, recursive neural networks
#34The author remarks that most of these are nonsensical.
As a non-magic player who lives in a house that sometimes hosts magic games (and thus clicks links re:mtg), I found every line to be entirely plausible overheard banter.
Re: Generating Magic cards using deep, recursive neural networks
#35> Shring the Artist > 2BB > Legendary Creature - Cat > Flying > Whenever you cast a spell, you may return target creature card from your graveyard to your hand. > 2/2 That's actually a fairly plausible card, in terms of mechanics. It's thematically appropriate for black, it's about the right cost, it's powerful but specialized (returns creature cards only, and only to hand rather than to play), and it's legendary to…
How long until someone draws up artwork for Shring the Artist?
Re: Generating Magic cards using deep, recursive neural networks
#36This is a very interesting idea. I'd like to see it played with 100% of the cards being generated through a system like this. 4 players, 4 automated decks of random (possibly nonsensical cards), that would be _so much fun_!
Re: Generating Magic cards using deep, recursive neural networks
#37This is a very interesting idea. I'd like to see it played with 100% of the cards being generated through a system like this. 4 players, 4 automated decks of random (possibly nonsensical cards), that would be _so much fun_!
You probably would draft instead of just automate the deck.
I think the mechanic is hard to generate a deck that utilize these mechanic and build synergy and combo out of these decks.
Re: Generating Magic cards using deep, recursive neural networks
#38Earlier quoted context omitted.
Matlab has a neural network toolbox that you can use to train small recurrent networks. There is a book that I can link if you are interested.
I am interested :)
If you are comfortable in MATLAB this tutorial is great: https://theclevermachine.wordpress.com/2014/09/11/a-gentle-i...
Re: Generating Magic cards using deep, recursive neural networks
#39This is a very interesting idea. I'd like to see it played with 100% of the cards being generated through a system like this. 4 players, 4 automated decks of random (possibly nonsensical cards), that would be _so much fun_!
You probably would draft instead of just automate the deck. I think the mechanic is hard to generate a deck that utilize these mechanic and build synergy and combo out of these decks.
Oh totally, draft would be fine, but even the idea of playing something randomly generated like this would be pretty fun.
Re: Generating Magic cards using deep, recursive neural networks
#40Reminds me: it feels to me like MTG card text (excluding flavor text) obeys a formal grammar of some sort. Does it?
If it does, the NN's could possibly be made a lot more powerful if you put a "parser" in front of its input end (so it was being fed ASTs instead of text) and a "code generator" on the output end (to convert confabulated ASTs back into text).