Card games: good and bad card designs, and how to identify them
As an extension of my previous article about OTK decks, I thought it would be good to write another article about specific card designs.
The same principle applies to card designs as applies to decks in general: cards should be strategically playable and counterable, and need to avoid certain things, such as uncontrollable randomness or fatigue nullification.
I will list some categories of bad designs below, along with some context and example cards.
Bad Design, category 1: Unpredictable Randomness
In a card game, luck (sometimes known as RNG, the abbreviation for Random Number Generator) is frequently an element of the game; it adds a level of excitement and variation to the game that prevents the game from feeling overly predictable or boring. However, that luck should be ‘influencable’ by the player, or else the game can become a dicerolling exercise.
We can define whether a card has ‘unpredictable’ or ‘uncontrollable’ randomness by understanding whether we can know all its possible outcomes from the state of play (i.e. whether we have complete information about those outcomes, or must rely on an uninformed guess).
Dirty Rat is a bad design because as a player, you do not in the vast majority of circumstances in most card games have any idea what cards are in your opponent’s hand (and therefore, you cannot know the possible outcomes of playing it at any given moment). Of course, if you know their deck, you know what *might* be in it — but you cannot, in any reasonable way, know if a particular minion or minions is currently in their hand. Even in the few circumstances where you do know, the outcome is extremely swingy: either you win the game outright, or the Dirty Rat has no effect whatsoever on the state of the game.
For context: Dirty Rat was originally added to Hearthstone as a way to address the prevalence of OTK (one turn kill) decks: by forcing a key minion out of the opposing player’s hand, you could prevent them from executing their combo and deprive them of their only win condition, often winning the game on the spot. The problem with this was that it was very much a total luck of the draw play (and was a lazy solution to a problem it did not fix).
To give an example: an opponent might have in their hand Archmage Antonidas, two Sorcerer’s Apprentices of which one is expendable, a Doomsayer and a Mad Scientist. You do not know which minions are in their hand, nor how many; if you wait until they have depleted their deck or almost depleted it, you will lose before you have a chance to play the Dirty Rat, and so you must play it earlier in hopes of getting lucky. You have a 75% chance to pull out a minion that will do nothing to the state of the game, and a 25% chance to pull out a minion their combo requires and destroy their game plan.
Here is an example of a well-designed RNG card:
This design is good, because the player can use the card strategically at the moment when its RNG is most likely to favour the player — because we know the possible outcomes from the state of play. If you play this card when your opponent has many minions on the board and you have few or none, then it becomes statistically more likely that the card’s effect will hit enemies; while it could still backfire, the chances are lower. Unlike Dirty Rat, the information you need to determine when to play this card is always available to you (the state of the board), allowing you to make a strategic decision to maximise its effect.
Here is another example of a good RNG design:
This card consists of two effects, one fixed and one random. Unlike Mad Bomber, there is no way to try and influence the RNG in terms of the 5-cost minion: you could get a great card like Pit Fighter, or a terrible one like Furbolg Mossbinder. However, this doesn’t matter design-wise: the player who uses this card can be sure of the potential results and their likelihoods (given the pool of available 5 cost minions), and can therefore judge the risk of getting a bad minion. This allows the player to strategically play this card when the risk of getting a poor minion from the card is acceptable (for example, when a high-value enemy minion is present on the board such as Emperor Thaurissan, or when tempo is needed). This also allows other decision making: when considering whether to use this card as a last resort versus an aggro deck, or when in need of lethal face damage, a player can calculate the possibility of getting a charge or taunt minion from this card and decide accordingly.
As such, “unpredictable” randomness is always unhealthy. Uncontrollable randomness, such as the minion-summoning effect of Firelands Portal, is generally fine. While it can occasionally result in overly strong or weak cards, that’s not a game design problem but a fine-tuning balance problem that’s easily fixed.
Bad Design, category 2: Fatigue Nullification
This is a relatively niche problem, but has existed in many archetypes. It is important to note that this doesn’t necessarily mean infinite value, as shown below with an example card.
While this card generates infinite value — it is quite weak. You need to use your entire mana pool each turn to play this card and the generated Yeti — a poor investment for 10 mana — and it does nothing to prevent you losing from fatigue damage in the late game.
In contrast, a slightly different version of this card:
Although the end result of this card is often the same, it has one gigantic and major difference: by playing this card, your deck will never run out of cards. As a result, even though it is poor value, you can now use this card to win against any deck that has run out of steam late game, as you will not take any fatigue damage (unlike the previous card, in which you would lose from fatigue before overwhelming the opponent with your slow but infinite Yetis).
You can therefore include this card in any control deck when you expect to be fighting against other control decks (i.e. where the game will end when a player runs out of resources), and effectively guarantee victory when this happens. This has a far bigger impact than it might seem; because you now know this infinite deck-extension is available, you can afford to be less careful with your other cards, knowing that a difference in resources late game will not be a problem for you.
Of course — as those familiar with past Hearthstone metas will likely have seen coming, the above is a prelude to the explanation of why this card is exceptionally bad:
For context, for those who do not play Hearthstone: Jade Golems are a mechanic, in which your first Jade Golem is a 1/1 minion, the second is a 2/2, and so on with no limit. This card therefore allows a Druid player to both avoid fatigue damage forever, and summon an infinite number of progressively larger minions for the cost of 1 mana, making almost every other class unable to contest them late game and shutting out other control decks. Infinite value is not in itself a bad thing in a card game, but it requires two things:
- The value must be poor (since you can use it infinitely, it should be far inferior to other cards in terms of value for mana cost). This gives players piloting tempo-oriented decks the chance to overwhelm the infinite value deck before it can out-value them.
- It must not shuffle infinite cards into your deck.
Here is an example of a well-designed card that extends the player’s deck:
While Prince Malchezaar was a weak card in play, it was a good design: it extended the player’s deck, gave a large amount of extra value, but came at the cost of diluting the deck at the start of the match (rather than allowing you to add cards to your deck when you wanted to), and only gave finite and somewhat unpredictable value. This meant it couldn’t be abused to shut out other decks in long games, and was not a “free value generator” as you had to dilute your deck in order to use it.
In the current more powerful meta, a version of this card that added the 5 Legendary minions to your deck as a battlecry would be a good Control design (though control decks have, for the most part, died out entirely).
Bad Design, category 3: Cards that ignore card advantage
Card advantage is a common term in card games, referring to the fact that a player who holds more cards in their hand is usually at an advantage, as they have more possible options to counter the opponent’s board state or bolster their own. As a result, cards which ignore card advantage are unhealthy by definition, as they give the player a free advantage without any disadvantage.
Aggressive low cost cards are generally good value for their mana cost; this offsets the fact that when you play them, you have one less card in your hand (as such, a hand full of low-cost cards will quickly be depleted, forcing you to rely on drawing from the top of your deck each turn). You can think of this as every card costing “a card” as well as just mana; cards themselves are resources.
Patches the Pirate specifically sidesteps this: because he is summoned from your deck when you play a pirate, you gain an additional aggro minion without losing any cards from your hand, allowing the aggro player to gain board control without losing card advantage. Cards that do this should not be printed (even in his post-nerf state, Patches is frequently used in the wild meta, as the ability to summon a minion from your deck for free is very useful for many aggro decks). For aggro decks, running out of cards is the least of their worries; most games that aggro decks lose are lost due to not being able to draw enough cards, not because their deck is empty.
The design of Patches gives it only one possible weak point: if you happen to draw him before you can summon him, the free effect is lost. This, however, is statistically unlikely to happen (around a 1/7 chance each game).
Before his nerf, Patches was particularly polarising in the community, as he was almost single-handedly the reason for Pirate Warrior decks dominating at the time; it is hard to contest an aggro player who can both fill the board with threats and still have a hand full of threats to play very early into the game.
Here is another example of how this works:
Baku the Mooneater (and odd Paladin) were extremely dominant in their time, due to the fact that having the ability to flood the board from turn 1 — without expending cards — made it impossible for other decks to stop the Paladin from achieving board control and amplifying it with cards such as Quartermaster and Level Up. It was not uncommon for an enemy to clear the board six or seven times simply to watch the Paladin flood it again next turn; there was no way to have enough board clearing cards to deal with that.
This was in fact so oppressive that Blizzard uniquely nerfed this card (and Genn Greymane) in the middle of a season, moving them to Wild mode far earlier than they normally would have, even though doing so messed with a great many decks and archetypes that used their odd-even mechanics.
As such, cards that allow a player to continually do things without expending card advantage are bad designs, especially when they do so early into a match.
This is especially true because in any card game, non-card abilities that expend no permanent resources (such as Hero Powers) are generally much weaker than cards themselves. They are like an expensive mana sink; because they can be used every turn, unlike cards which can be used once only, to avoid being of unfairly high value they give poor effects for the mana spent. Baku the Mooneater and Genn Greymane’s effects stopped this, allowing decks to ‘double dip’ into resources and have both cards and high-tempo hero powers at their disposal.
Bad Design, category 4: Cards that give an effect to both players
In theory, cards that do this wouldn’t be a problem: in practice, however, they frequently are.
The theoretical logic of printing such cards is that it provides an “equal” benefit for both players, and therefore they aren’t unbalanced. While this is occasionally true, the vast majority of the time only one player benefits: in the case of cards that give card draw, the other player will frequently be forced to burn cards or will be unable to play the cards they have drawn fast enough for them to matter.
The only way you can prevent these cards being problems is to not have any return-to-hand or copy effects in the game, which would severely restrict design space for all future cards in your game. Without this restriction, however, cards like this are bound to go on to cause serious polarisation problems, as shown below.
This card is only ever seen in one kind of deck: OTK decks, in which the player doesn’t care what the other player does and simply wishes to draw cards as fast as possible. The downside is never an issue for the person playing the card; either you force your opponent to overdraw, or you don’t care what they are drawing anyway.
Here is a non-draw example of a “both players” card. It largely had a similar intent: warriors would play this without a care in the world for the other player. Against an aggro deck, the armor will absorb the weapon’s damage, making this a 2 mana 2/3 weapon (better stats than Fiery War Axe against many aggro decks), and against enemy tempo or control decks the main use of this card is to destroy an enemy weapon by replacing it. It could thus be used to delay enemies using such weapons as Doomhammer, a buffed Spectral Cutlass or Kingsbane.
Examples of well-designed “both players” cards would be Tanglefur Mystic and Prize Vendor, whose effects are small and consistent enough to not be abusable for only one player’s benefit.
