Ben Brode on Defining Complexity, Depth, and 'Design Space'
Ben Brode came to reddit today to clear up some confusion about his recent comments! He also followed up on some of the reactions people had.
Quote from Ben BrodeHey all!
I rarely start new threads here, but there was a bit of confusion regarding recent comments I made about complexity in card design, and since my comments had low visibility, and I thought the larger audience would find it interesting, here I am!
Defining Complexity and Depth
Complexity is different than Strategic Depth. For example, 'Whirlwind' is very simple. So is 'Acolyte of Pain'. So is 'Frothing Berserker'. Together, these cards were part of one of the most strategically difficult decks to play in our history. Hearthstone, and it's individual cards, are at their best when we have plenty of strategic depth, but low complexity.
You can sometimes get more depth by adding more complexity, but I actually think that cards with the highest ratio of depth to complexity are the best designs. That doesn't mean we won't explore complex designs, but it does mean that they have a burden to add a lot of strategic depth, to help maximize that ratio.
My least favorite card designs are those that are very complex, but not very strategically deep. "Deal damage to a minion equal to it's Attack minus it's Health divided by the number of Mana Crystals your opponent has. If an adjacent minion has Divine Shield or Taunt, double the damage. If your opponent controls at least 3 minions with Spell Damage, then you can't deal more damage than that minion has Health." BLECH.
At any rate, making cards more complicated is easy. Making them Strategically Deep is more difficult. Making them simple and deep is the most challenging, and where I think we should be shooting. It's important to note that an individual design doesn't necessarily need to be 'deep' on it's own. Hearthstone has a lot of baked in complexity and depth: 'Do I Hero Power or play this card?' 'Do go for board control or pressure their hero?' And often (as in the case of Whirlwind) a card's depth exists because of how it is used in combination with other cards. Creating simple blocks that players can combine for greater strategic depth is one of the ways we try and get that high ratio of depth to complexity.
Defining 'Design Space'
Sometimes we talk about 'design space'. Here's a good way to think of it: Imagine all vanilla (no-text) minions. Like literally, every possible one we could make. Everything from Wisp to Faceless Behemoth. Even accounting for balance variation (i.e. 5-mana 6/6 (good) and 5-mana 4/4 (bad)), there are a limited number of minions in that list. Once we've made every combination of them - that's it! We couldn't make any more without reprinting old ones. That list is the complete list of 'design space' for vanilla minions.
The next level of design space would be minions with just keywords on them (Windfury, Stealth, Divine Shield, etc). There are many cards to be made with just keywords, and some are quite interesting. Wickerflame Burnbristle is fascinating, especially because of how he interacts with the Goons mechanic. But eventually (without adding more keywords), this space will be fully explored as well.
When you plan for a game to exist forever, or even just when it's time to invent new cards, thinking about what 'design space' you have remaining to explore is important.
Some day (far in the future), it's conceivable that all the 'simple but strategically deep' designs have been fully explored, and new Hearthstone cards will need to have 6-10 lines of text to begin exploring new space. I believe that day is very, very far off. I believe we can make very interesting cards and still make them simple enough to grasp without consulting a lawyer.
Some design space is technically explorable, but isn't fun. "Your opponent discards their hand." "When you mouse-over this card, you lose." "Minions can't be played the rest of the game." "Whenever your opponent plays a card, they automatically emote 'I am a big loser.'" "Charge"
Sometimes design space could be really fun, but because other cards exist, we can't explore it. Dreadsteed is an example of a card that couldn't exist in Warrior or Neutral, due to the old Warsong Commander design. (in this case we made Dreadsteed a Warlock card) The Grimy Goons mechanic is an example that couldn't exist in the same world as the Warrior Charge Spell and Raging Worgen. (in this case we changed the 'Charge' spell)
In a sense, every card both explores and limits 'design space'. The fact that Magma Rager exists means we can't make this: "Give Charge to a minion with 5 Attack and 1 Health, then sixtuple it's Attack." That's not very useful (or fun) design space, and so that tradeoff is acceptable. However, not being able to make neutral minions with game-changing static effects (like Animated Armor or Mal'Ganis) because of Master of Disguise... that felt like we were missing out on lots of very fun designs. We ended up changing Master of Disguise for exactly that reason.
Cards that severely limit design space can sometimes be fine in rotating sets, because we only have to design around them while they are in the Standard Format, as long as they aren't broken in Wild. Because Wild will eventually have so many more cards than Standard, the power level there will be much higher. Most of that power level will come from synergies between the huge number of cards available, so sometimes being 'Tier 1' in Standard means that similar strategies are a couple tiers lower in Wild. We're still navigating what Wild balance should be like. It's allowed to be more powerful, but how much more powerful?
I think defining these kinds of terms helps us have more meaningful discussions about where we are doing things right, and where we have room to improve. Looking forward to reading your comments!
Quote from Ben Brode
I don't know where or even if I read this, but I remember something along the lines, that Wild should not even be balanced, just a clown fiesta of all the old cards to have fun with. Am I just misinformed or was this a thing and got scrapped later?
While "clown fiesta" is the driving goal for most of our decision-making, we would actually like Wild to be balanced. The problem is defining what 'balanced' means in that mode, because the power-level will always be much higher than in Standard. Does it mean all classes are at 50% and see exact equal play? Does it mean Freeze Mage is the best class but there are 4 other archetypes at tier 2? How frequently should we be balancing cards there?
There is a lot to learn about the right way to approach Wild, and what the expectations are from players who like to play there. It's something we'll need to figure out over time.
Great post Ben. Out of curiosity, how do you view linear or parasitic cards like Jade? They tend to open some amount of design space but at the tradeoff of interacting with fewer existing cards. Have you been happy with them so far?
I tend to like some parasitic designs, but I don't think we should do them exclusively. C'Thun was also a fun parasitic design.