Deckscalibur uses the concept of evolution to build decks. It starts off with a (mostly) random deck, and makes a bunch of slightly different versions of it (mutant freaks). Then it scores them using a carefully crafted deck quality test, kills off the worst ones, gives the best ones slightly mutated children, and repeats the process over and over again with the best children living on, and the worst children getting put to sleep. D:

I'll be improving the way the algorithm determines card interactions and throwing out the outdated and incomplete information it uses now.

Meta analysys is coming. This is important because it allows deckscalibur to make strong decks that are strong on their own, but especially strong against the meta. This will also improve the algorithm's understanding of how people use cards, and what cards have a high skill cap.

I'll also likely be adding win rate projections, which will allow deckscalibur to make even better decks using artificial intelligence. This will be a fun project which should improve the algorithm a lot.

Decksperiments and the Skill tree are also planned!

Pros (so far):
• TJ Trav - @TJ_Trav01
• Jules - @julesy2307

Earliest beta:
• Chad Potro
• Spagett!
• Zeppelinmaniac
• Robroy

My clan, BSG!

Thanks to everyone who helped!