In Magic The Gathering there are 3 main different deck archetypes: control, aggro, and midrange. Each of these archetypes represent a different play pattern and each excel at a different point in the game. An aggro deck wants to win as fast as possible, usually sacrificing its own resources to deal those final points of damage. A control deck wants to stall the game out as long as possible, winning through a battle of attrition and just by locking out your opponent or grinding them to a pulp. A midrange deck wants to control the early game and then go for a strong midgame, carrying that lead through to the end.

These archetypes are also present in Keyforge, but instead of deck archetypes they represent different stages of the game or playstyles shown in the different houses.

Call of the Archons Untamed, for example, is a great example of an aggro house in Keyforge. The best Untamed decks wanted to generate aember as quickly as possible with cards like Hunting Witch, accelerating your keys with Key Charge or Chota Hazri. With this house, it was often the correct play to reap with your creatures in the lategame regardless of boardstate because with a good Untamed deck you were always able to generate a ton of aember and then key charge your way to a win.

Sanctum is a perfect example of a midrange house. This house has strong, resilient creatures without much in the way of aember control or speedy aember generation, rather relying on their difficult to remove boardstates and ability to slow down their opponent just enough with cards like Barrister Joya or Raiding Knight to get the reap engine rolling away to victory.

Finally, Dis is a quintessential control house. This house makes your opponent want to flip the table while you cackle behind your double Tezmal with glee, activating your Lash of Broken Dreams every turn. Cards like Hand of Dis, Gateway to Dis, Poltergeist, Succubus and Restringuntus are all great control cards that pay you off for using them at the right moment and, in the case of the creatures, can single-handedly steal a game when gone unanswered.

Just because you have a good aggro Untamed house does not mean you should always play the house in the same way, however. If you land an early Duskwitch, it may be more valuable to try for a more midrange playsyle in order to dominate the board and get rid of any threats to your Duskwitch. If you try to play Duskwitch in a very aggresive way you will get less value from your Duskwitch overall because while you are busy reaping instead of clearing your opponents board, they are simply throwing two creatures at your Duskwitch to kill it, whereas if you kept it alive you could have gotten much more value from the card in most cases. This would make Duskwitch a great midgame card. If you are ahead on board your opponent must scramble to kill it immediately, costing them resources that usually could have been more efficiently spent elsewhere. If you are behind while playing Duskwitch your opponent may waste time fighting it, potentially allowing you to claw back board control or just distracting them from generating aember. Unfortunately Duskwitch is easily removed due to its measly 1 power, balancing its extremely strong effect.

Early game it is most important to establish a board presence, especially in AoA. If you can drop 2 or 3 beefy Brobnar creatures it is much easier to use those units to fight later on, knocking down your opponent while leaving you with the most powerful board. The faster you get out your units the more options you have later in the game. If you have 5 units and your opponent has 3 you have a wider range of options to fight. You can assign damage more accurately and more efficiently, deciding which creatures live and die, snowballing your lead on the board. Raw aember does not matter nearly as much at this stage in the game because the expected value of creatures just goes up and up the longer they stay in play. If you are able to reap with a creature 5 times over the course of a game, then that one play has netted you more than almost any other single card in the game. If you manage to use your Lollop the Titanic to kill 3 enemy creatures you effectively played a card that read “Destroy 3 enemy creatures over the course of a few turns”. Comparing that value to that of, say, Hand of Dis (which can be considered a relatively strong card) seems almost laughable. During the early stages of the game it is most important to play as many cards as possible, and choosing to play the most possible cards in every turn is usually a stronger choice.

Midgame opens up a couple of new options. I like to think of the midgame as the point in time where it is apparent which player is controlling the board. Once one player has clear unit control, the other player must choose between racing aember or fighting to reclaim board from a disadvantageous position. Knowing both your deck and your opponents is extremely important for this stage of the game because this is where you have to figure out your win conditions. You want to be able to balance your pure acceleration (raw aember and aember generation) with expected aember over the course of the game (reaping and playing for more value over the immediate gain) so that when you get to the late game you aren’t so far behind your opponent in raw aember that you can never win, but you also have enough actualized resources that you can make that final burst in the late game. The midgame can be summed up in one phrase: playing to win versus playing not to lose.

I classify late game as the point in the game where a player is probably going to win the game in the next 3 turns. Late game is where you must extract as much value as possible from your available resources in a very short time. You have to be able to use all the resources you have saved up through the course of the midgame to do two simple things:

  1. Win

  2. Stop your opponent from winning

It looks really obvious but it can be difficult to play disciplined lines lategame. You may have to call a house and only play 1 of the 3 cards in your hand from that house. Your evaluation of win condition from the midgame should come in to play here. What cards explicitly stop your opponent from winning and help you win? Swindle becomes an absolute powerhouse at this stage of the game because the difference between steraling aember to go from 3 to 6 on your last key is way more important than on your first key. Whenever you get to enough aember to forge, your opponent MUST answer you. If you can get to enough aember to forge, you should. Nothing else matters as much as that aember at this point in the game.

This Monday I was watching a player at my LGS’s chainbound event. This was a heavy approximation of his board and hand below.

Both players were at 4 aember and 2 keys.


This player is one of the strongest players at our LGS events and is heavily involved with the community, and here he opted to call Sanctum and establish his board. If this was the board position at the first key, calling Sanctum is 100% the correct play, but at this stage in the game he definitely should have called Shadows. Calling Shadows checks both of the two boxes for endgame play: it stops your opponent from winning (with the steal from the Umbras) and would put him at enough aember to forge a key. Even good players make these kinds of mistakes, and identifying these weaknesses in your play can give you that extra 5% edge that you need to become more competitive.

From this position the Lash of Broken Dreams may deter you from going for the aember-heavy play with Shadows, but there are still a couple of factors that make Sanctum much weaker. Firstly, the Sanctum creatures have almost zero expected value. The game most likely wont last long enough to reap with them once, and even if you did, you are only playing 4 Sanctum creatures, giving you a 4 aember reap once you get around to it. Calling Shadows gives you a 5 aember swing immediately if you factor in the steal. Secondly, calling Shadows forces your opponent to call Dis just to use their Lash of Broken Dreams. If they have have any cards in Logos or Untamed or any other big plays from either of those houses they are probably out of luck, unless they can either win on the spot with a Key Charge or somehow get you off your 7 aember. By calling Shadows you also get a free Restringuntus effect. Attacking with the Umbras also allows you to kill their Yurk, the only Shadows creature they could get value out of, essentially robbing them of another aember given that they are extremely likely to call Dis.

This play is also based off knowledge of the decks that are being played. I know that both decks were AoA sealed decks. As a set, AoA has much worse steal and aember control that CotA, especially in Logos and Untamed. Maybe if aember counts were lower at this point in the game it would have been better to call Sanctum. If both players were at 2 keys 0 aember then it would have actually probably been better to call Sanctum and use the house to establish a stronger board presence.

Knowing when to transition from the different stages of the game is extremely difficult and takes lots of practice, but once it is mastered it becomes much easier to win games that seemed unwinnable or to completely crush your opponents from seemingly even positions because you have managed your resources better and maximized the value from different effects and plays through different parts of the game.