Investigating Holding Cards

Back in April 2019 I decided to go to my first Vault Tour at UKGE. Making that decision I knew I wanted to improve my gameplay, and the very first thing I looked for is someone to help me understand holding or discarding cards in KeyForge.

The article I found by Jake Frydman’s (Co-host of Sanctumonius podcast) KeyForge Tips: Discarding was exactly what I was looking for, and my first step into understanding the game better. it is the best resource on the subject that I’m aware of.

It is not needed to understand this article, but I highly recommend you read it if you haven’t done so.

Test Your Knowledge

This is generally a good tip for life. The first time I heard about this was from a friend that routinely has to hire people. He kept going with his gut feeling and prior experience about who to hire and who not to. Until one day he asked himself: “What if my gut feeling is wrong?” and hired someone, qualified, that he would not normally hire. This decision affirmed his gut feeling, as it was definitely the wrong hire.

Hiring the wrong person can be a costly mistake, but holding a card in KeyForge in a test game and seeing how it pans out doesn’t cost much. It is important, however, to look at how that specific action panned out, and not if you won or lost the game.

I’ve been holding cards from time to time, and let me tell you, I don’t think it is as bad as everyone says it is. Yes, sometimes it backfires horrendously, usually with situational cards, but sometimes it feels very good indeed.

Hypothetical Card

Let’s imagine a card that read: “If your opponent has 6 Æmber, win the game.” Obviously you would hold this card until your opponent has 6 Æmber, as that is sure to happen. Fortunately for us, there is no card that powerful in KeyForge, as that would make for some very boring games. However, there are cards in some decks that come close.

Martian Generosity, in a deck with Key Abduction and a few supporting cards could very well read: If you have 4 or more Æmber, forge a key for free, establish a massive board, and set yourself up to do it again in a turn or two. As with the Hypothetical Card, you’re very likely to want to hold onto it. In fact, playing 2-3 Mars creatures and holding it sets you up to do just that the following turn.

That is a combo card, and many people already know that a game winning combo can be played towards in order to win. Jake said as much back in March 2019, which is when Library Access Nepenthe Seed combo was all the rage.

So what about other cards, could they be worth holding on to? One of the best cards in isolation is Relentless Whispers. With no target to kill, it’s strictly worse than Punch, but with a target to kill it has immense value. Kill a creature along with 3 Æmberswing in your favor.

Opportunity Cost

Many smart players have equated holding a card to chaining yourself, and while that is sort of true, it doesn’t really paint the entire picture. This is because once you run out of those hypothetical chains and play the house again, you are also holding a high value card.

I wrote a little bit of code (github link) to try and get more data on this. The code builds a deck of 36 cards from 3 houses, A, B and C. It replaces one of the cards in house C with card X, which is our Relentless Whispers, the card we may want to hold. It then shuffles the deck, and picks which house to play from the top 6 cards (the hand) based on the house with the most cards.

After playing house C and holding X, it is replaced by C as we assume you’ll play it next time you play C. It stops on shuffle.

This is by no means perfect, and my code may have some issues, but here is one random output:

C A B A A C | A C B B X B A A B C C C B B B A B C A A C C C C A B B A B A
Turn 1 – Playing A
C B C A C B | B X B A A B C C C B B B A B C A A C C C C A B B A B A
Turn 2 – Playing C
B A B B X B | A A B C C C B B B A B C A A C C C C A B B A B A
Turn 3 – Playing B
A X A A B C | C C B B B A B C A A C C C C A B B A B A
Turn 4 – Playing A
X B C C C B | B B A B C A A C C C C A B B A B A
Turn 5 – Playing C
C B B B B A | B C A A C C C C A B B A B A
Turn 6 – Playing B
C A B C A A | C C C C A B B A B A
Turn 7 – Playing A
C B C C C C | C A B B A B A
Turn 8 – Playing C
B C A B B A | B A
Turn 9 – Playing B

In the above example, we hold X and the next turn we play B. Our top card in the deck is B, so we miss out on playing it along with the other four B cards. The turn after that we play A, but the next two cards in our deck are C, so there is no loss there. If we didn’t hold X, it would look like this:

X B C C C B | B B A B C A A C C C C A B B A B A
Turn 5 – Playing C
B B B B A B | C A A C C C C A B B A B A
Turn 6 – Playing B
A C A A C C | C C A B B A B A
Turn 7 – Playing A
C C C C C A | B B A B A
Turn 8 – Playing C
C C B B A B| A
Turn 9 -Playing B

Here we still play five C on turn 7, but we play a different C card to the one held. So the opportunity cost of holding X is one B card, and one C card. So the question you need to ask yourself, is playing those 2 cards better or worse than getting Relentless whispers off.

Say those two cards are Punch and a Creature. The Punch will have a comparable effect to the Relentless Whispers killing something off and you gaining another Æmber (in addition to the Relentless Whispers we played with no target), but not your opponent losing 1. If the Creature is an Urchin then you’d definitely be ahead playing it, Gaining a total of 3 and them losing one. If it’s a creature without an immediate effect, things aren’t so clear cut anymore.

And what if those two cards are Cooperative Hunting and a Skeleton Key, suddenly that Relentless Whispers looks pretty sweet in comparison.

Spread

One of the things I wanted to see is the difference in the number of turns between playing a house and playing it again while holding or not holding the card. As such, I made my program run 10,000 times and count that spread. Here is the result:

Spread when holding: 1.4507
Spread when not holding: 1.9249

As you can see, it takes on average half a turn more to come back to the same house if you’re not holding a card. So If you’re holding X it is much more likely that you’ll have 3 total cards in house C in 1-2 turns than if you didn’t hold it. This leads me to believe that holding cards in a main house, the house you want to get back to in order to use your creatures, has a much lower Opportunity Cost, since you’ll be happy to get back to that house.

It may also mean that Æmber control cards are worth holding, especially in houses that don’t have much. Because if your opponent goes to check, then it would allow you to clear your hand of that house later on, when you might not be able to do so if you lacked something to get them off check.

Too Situational

Playing a huge Brobnar board and holding a Warsong is almost always a bad idea. Your opponent is likely to have many ways to circumvent your plan between playing a board clear, stun your creatures or just play a Foggify, rendering the card you held useless. Cards held need to be high impact, and need to have fairly common situations for them to be effective.

For example a Doorstep to Heaven is not always good, but the situation in which your opponent has more than 5 Æmber is so common it is all but guaranteed.

Knowing which cards you might want to hold, and looking at your opponent’s list at the beginning of the game to assess if you should can help a lot in this decision. For example, if you have a creature heavy Sanctum lineup, it might still be worth holding Glorious Few if you your opponent is likely to play Gateway to Dis into Arise! into 7 creatures.

Contact and afterward

My conclusion is you need to know how your deck wants to play.
If you’re in a Sealed environment you’ll need to assess the potential impact of the card against roughly a two card loss from the rest of your deck.
In Archon, you should learn your deck and step out of your comfort zone to see which cards might be held and which should not. Try holding a card you usually don’t and see if it gives good returns. Try discarding a card you usually hold and see what happens.

And report back with your findings, I’m curious to hear if anybody changes their opinion on holding cards.

As always, you can follow me on twitter for updates. And join us at the Sanctumonious discord server if you’d like to chat with me, or join an awesome community of keyforge players.
The Sanctumonius community now has a website (which I made), come have a look.