Making a Hearthstone Card: The Madness of Yogg-Saron

Blizzard's Mike Donais and Peter Whalen give us an exclusive inside-look at the design process that went into creating that most wild of Old Gods, Yogg-Saron.

3

Now we've got a tale to tell you of the ones from long ago.

The Hearthstone: Heroes of Warcraft expansion, Whispers of the Old Gods has 134 new cards to collect, but none of them have quite gotten attention and fan chatter like Yogg-Saron. The Legendary card is one of only four Old Gods, and his random effects make him by far the most unpredictable. With special access from Blizzard's Team 5, we can reveal unused Yogg-Saron card concepts and design insights straight from designers Mike Donais and Peter Whalen. 


The story of Yogg-Saron's development actually begins as a broader picture, with the development of the Old Gods as a family of cards. The initial concept was actually to have each card impact one of the core pillars of the game, and to take effect immediately when the game started. C'Thun would replace your entire deck, N'Zoth had an impact over your minions, Y'Shaarj replaced your Hero Power, and Yogg-Saron gave you extra Mana crystals. It worked well thematically, but it was limiting from a design perspective.

"We hit on the cultist ritual mechanic that we wanted to do for C'Thun, and at that point we started moving away from that symmetry where all of the Old Gods affected you at the start of the game," Whalen said. "And once we broke with that symmetry a little bit it made more sense to find the most fun designs for each of the Old Gods rather than try to tie them together into that mold."

Instead, the focus of the Old Gods designs shifted from having a singular theme, to a broader idea of what Donais called "blow-your-mind" effects. And some of the rejected designs, while they didn't work for a variety of reasons, were truly mind-blowing.

Design Notes: This is the first version of Yogg-Saron that we have a record of. We wanted the old gods to blow your mind when you saw them. We tried some crazy versions that had effects if they started the game in your deck. This one is a mixed blessing, your turns are really short, which can be exciting if you opt into it. As a bonus, once it’s in play, all your spells are 1 mana which can cause some crazy turns.

Despite this effect being limited to only when it's on the board, this effect was simply too powerful. Mike Donais said some combinations were "absolutely ridiculous," and as a result it didn't take long to change this design into something a little more reasonable. In particular, he said, playtesters would use Mage decks with Pyroblast and Arcane Intellect to deal massive damage and keep their hands full for more burn spells. As long as they could keep Yogg-Sarron on the board for a turn, it was essentially a guaranteed win. However, the shortened turn timer was one idea that the team would circle back to before development was finalized.

Design Notes: Playing these games had a weird feeling where one player felt like an adventure boss and the other felt like they were playing normal Hearthstone. It seemed like it was possible to make it fair, but the games didn’t feel particularly good, so we moved in a different direction.

This design would cap you at four mana, but you'd start with an important advantage. The result was a unique game style built around early leads and a "snowball" effect, but it ultimately wasn't very enjoyable for either player, for very different reasons. 

"If your opponent got to 6 crystals and could stabilize, they won, and otherwise you could just kill them with 5/4s," Whalen said. "It really felt like playing an A.I. where it feels good when you win but really bad when you lose. And it felt kind of boring for the guy who had just 4 mana crystals. Just, 'I'm going to do this thing.'" 

There wasn't much flexibility on the side of the player with Yogg-Saron, since the deck would be built with as many four-drops as possible, along with a handful of 1- and 2-Mana cards so you could play multiples on one turn. Your opponent's best hope was to "get lucky and stabilize with AOE." And while it wasn't necessarily unfair to the opponent, it certainly felt that way. That sensation was more important than the statistics.

"More than just the power level, it was the feel," Donais said. "Even if his winrate ended up being 50%, the times you lost to it felt really bad."

The team says they also attempted this with one less Mana crystal, so you would be capped at three, but it still didn't work.

Design Notes: We decided to try the 15 seconds per turn version again because it was exciting to play. People who liked the gameplay would include it in their decks even if it wasn't powerful. Some people didn't like being rushed and would avoid playing it.

This effect, which is essentially a one-sided Nozdormu, had lots of flavor going for it. Whalen even said that if they'd gone with the design, they would have liked to have some special animation to announce its presence. 

"We'd do some cool effects where he popped out of your deck, ate your mind, and popped back in," he said, "to announce to both players that something crazy happened."

As a trade-off for your short turn timer, you would have access to a 5-Mana 10/10, the largest Mana-to-stats ratio in the game to date. It was a Legendary built around calculated risk, and Donais said it was best for the player who tended to take quick turns anyway. However, it ran into technical limitations. 

"We had some problems with that, because the phone interface is a bit slower, and also if you had just a minor lag on your Internet service or if some cards don't interact perfectly," Donais said. "We'd have to really engineer turn timers so they were perfect. We didn't have the resources to do that and weren't sure it would be worth doing the card even if we had that."

Design Notes: We wanted to move the Old Gods to be more build-around, so we sat down and redesigned almost all of them. This one was the crazy spell build-around. People were very excited about it, but it turned out to be very, very dangerous.

This design was very close to the finalized one, but rather than cast a set of random spells, it would recast all of the ones you already had. The exploits were plentiful. The team said they simply found too many one-turn kill decks.

Blizzard cited a Druid deck with Bite and Claw, so you'd immediately build up a lot of armor and attack value at once. A Warrior deck used Charge with Blood Warriors, so you could get extra copies of Yogg-Saron and give him Charge as well. The Rogue could exploit Assassinate and Headcrack, but the bigger risk was to sap all the interactivity from the game.

"Infinite Yogg Saron," Whalen said. "It turns out if you play him and you cast Vanish that game, really, nobody gets to play Hearthstone anymore. You don't always win those games but they're not that fun."

Design Notes: Here is the design we finally arrived at and tentatively all liked. One last change after a lot of testing was removing 2 from his Health. He was pretty strong, we wanted to make him a touch weaker while still feeling large enough to be Yogg-Saron. We couldn't change any numbers in his Battlecry, and we liked all the gods costing 10, so Attack and Health were the only things we wanted to change at this point.

The near-finalized design had him as a 7/7, requiring only a minor tweak. Donais said reduction to five health made him "a little less reliable" so he wasn't simply an auto-include in spell-based decks. Even then, though, the engineering team had a task ahead of them. The randomness, and the need for Yogg-Saron to interact with every spell in the game, required a lot of meticulous testing. Ben Brode recently said on a stream that Yogg-Saron took the longest to program, and Donais and Whalen agreed. 

"Some of the spells on the back-end needed a little massaging to make it work in this crazy context being cast alongside other things," Whalen said. "Also the cards involving choosing had some difficulties because Yogg-Saron made the choice for you, like Tracking for example, or the Druid choice cards. That required a bit of additional work too. Secrets uses a slightly different code path than everything else so it took a little extra as well. The engineers needed to go through basically every card and make sure it worked right."

The result is a random effect that may not make its way into competitive tournament decks, but has made for some of the best replay and highlight videos of the expansion so far. Even in Blizzard's own reveal stream, Yogg-Saron impacted the board in unexpected ways leading to a dramatic finish. 

"One of the important things [for Old Gods] was you had to build your deck a certain way to take advantage of them," Whalen said. "We tried a bunch of different big payoffs for what Yogg-Saron does. The super-consistent Yogg-Saron variants caused you to kill your opponent that turn or create such an overwhelming advantage that it wasn't that much fun. So we ended up tweaking it for this random version, which ended up being a lot of fun and much healthier for the game than creating a bunch of one-turn kills."

Editor-In-Chief
From The Chatty
Hello, Meet Lola