MArcomage

14 wall for 5 recruits on a Common card? That's a LOT of wall gain, especially from a resource that generally isn't used for building. Generally, using an atypical resource for something should cost more, not less.

I assume that discarding a random card with an attack value was meant to be part of the cost, but in many cases, discarding an extra card from what has become a stale hand is a benefit in itself.

This card is far too advantageous as it is now. I'd dread seeing three of it in anyone's hand, a very real possibility for Common cards.

Is it better now?

I feel like this would be perhaps a little too powerful for defensive decks. If your deck is focused on tower and wall-building, any attack cards you have are probably just keyword filler (like Bronze Golem or whatever), and you're not going to miss them. Also, recruits is the hardest resource to find cards for for a defensive deck, so this would be a big boon.

I'm also wondering if there's any elegant way to determine what an "attack" card is. Does a card like Clay Golem that can do damage or raise your wall count as an attack card? Does it only count cards that have "attack: x" where x is some number, or does it also include "enemy wall -x" or "enemy tower -x" cards?

Clay Golem is listed with the Attack cards in the deck building advanced features, so I'd assume so.

I still say that, because Recruits are generally not used for wall building, the cost needs to be higher, probably along the lines of a 1-to-1 ratio. Perhaps higher for Common.

Oh, yeah, I forgot that they added those filters. I wonder if Mojko hand-coded them, or if we has able to automatically classify them.

Advanced filter is based on searching a substring in the card effect text. So, attack cards are defined as cards that have the "Attack:" as a substring in their card effect.

Any filter configuration can bu used to specify for example summoning cards (see game manual -> Card template). We currently can't check if a card is an attack card, but such functionality can be added, since it's not very difficult to implement.