Delving Deeper – The Evolution of Hearthstone’s Single Player Mode

Considering how many times I’ve brought up Hearthstone in my blog posts it should be no surprise to anyone that I’m a pretty big fan of the game. And while I’ve talked before about balancing cards and deckbuilding, I really haven’t talked about the game’s single player content. Honestly, at launch there wasn’t much to talk about. Initially the only single player content available was playing against the AI Innkeeper. And while I believe this mode serves a purpose (it acts as a good teaching tool for new players that feel intimidated playing against real people), it’s not a very challenging or compelling mode.

Thankfully, Blizzard took a massive step forward with the introduction of Adventures. Starting with Curse of Naxxramas (and continuing in Blackrock Mountain, League of Explorers, One Night in Karazhan, and Knights of the Frozen Throne), players could play through a series of boss encounters each with their own rules, mechanics, and encounter specific cards and hero powers. And these modes were a ton of fun (and are honestly still pretty fun today)! Each one did a fairly good job telling a story and offering unique and often difficult challenges.

…yeah, I’m doing great.

…yeah, I’m doing great.

However, these modes had two major limitations:

Replayability: On average, each adventure had around 10-15 encounters, typically with both a normal and a heroic difficulty. But once you’ve finished them…that’s about it. There really isn’t an incentive to play any of these adventures again after you’ve completed them.

Accessibility: Most of the boss fights in these adventures have players crafting their own decks to use (with a couple exceptions such as the Chess encounter in One Night in Karazhan). While this lets you flex your creativity, if you’re a new player with a relatively small collection some fights are extremely difficult (if not impossible) to complete.

This is where Dungeon Run comes in. With the release of Kobolds and Catacombs, Blizzard completely revamped their single player content into a deckbuilding roguelike mode similar to games like Dream Quest and Slay the Spire. Players would start off a run with a small deck of relatively weak cards, but as they defeated bosses they could add new cards and passive treasures to their decks to make it stronger. This gave an awesome sense of progression since you could feel yourself getting better as you got deeper into your run. Each run would end either when the player loses, or when they defeat 8 bosses in a row. Once a run was over you could then start again with a fresh weak deck of cards.

I absolutely adore Dungeon Run. It took massive strides in addressing both of the major flaws of the earlier adventures:

Replayability: Dungeon Run oozes replay value. While each run only contains 8 bosses, there are 48 different possible encounters you could get. On top of that you’ll be offered different cards and passives each run so your deck will be completely different. Add on the fact that you can play it with 9 different classes and you have countless hours of gameplay.

Accessibility: Since all the deckbuilding in Dungeon Run occurs during the run itself, you card collection is completely irrelevant. You could have a full golden collection or literally nothing but a few common cards and it won’t affect your ability to play the mode.

As great as Dungeon Run was (and still is), it had some room for improvement. Most notably in terms of its storytelling. While the old adventures featured interesting narrative arcs with the occasional twist (especially League of Explorers and Knights of the Frozen Throne), Dungeon Run didn’t really have one. You were just an adventurer looking for treasure. So with the next expansion, The Witchwood, Blizzard incorporated a much stronger narrative in its Monster Hunt. Now rather than taking control of the typical 9 classes you could choose from 4 classes unique to Monster Hunt: Tracker, Cannoneer, Houndmaster, and Time-Tinker. Each of these classes had a unique hero and story, culminating in a 4-on-1 battle against Hagatha the Witch. I really liked this way of incorporating story, plus it came with the added benefit of introducing some new classes!

Get wrecked by doggos, Hagatha!

Get wrecked by doggos, Hagatha!

But this came at a cost. Compared to Dungeon Run, Monster Hunt just didn’t feel like it had the same amount of replay value associated with it. Sure, you still had tons of different bosses to play against and plenty of ways to build your deck, but I think the drop from 9 heroes to 4 heroes hurt a lot. Plus, once you’ve completed the story the incentive to play through it again goes down a little.

This brings us to Rastakhan’s Rumble’s Rumble Run (geez that’s a lot of Rs), which if I’m being honest I was really disappointed by. This time you play as a troll named Rikkar who is competing against 8 other challengers to become the champion of the Gurubashi Arena. The biggest addition to this mode is the shrines, which are semi-permanent minions that appear on your side of the board and give you powerful passive benefits. Your opponent also had a shrine of their own, so you’d need to come up with a plan to play around it. This was a neat idea, and really changed up the way you play your run. However, Rumble Run took a lot of really significant steps back:

Replayability: In both Dungeon Run and Monster Hunt you had dozens of possible encounters. But in Rumble Run there are only 27 (3 possible shrines for each of the 9 classes). While that is a decent amount of variety, after a couple of runs you can’t help but feel like you’ve seen most of what it has to offer.

Agency: To offset the number of bosses, player gained a lot more variety in terms of how to build their deck. After all, you can choose any of those same 27 possible shrines for yourself. However, at the start of each run you are only offered three of them at random. This means you lose the agency you had in Dungeon Run and Monster Hunt to choose the class you want. For example, let’s say you were offered Druid, Warrior, and Hunter shrines. If you hate playing those classes, that’s too bad because you’re stuck with those options until you play one of them. On top of that, if there’s a specific shrine you really want to try you only have a 3/29 chance of it being offered on any given run (technically the odds aren’t quite that clean but you get the idea).

…sigh, still no Paladin…

…sigh, still no Paladin…

Narrative: Rumble Run’s story is about as deep as Dungeon Run’s. There really isn’t much more here than “a troll fights a bunch of other trolls until he becomes the best.” This is a huge step back from Monster Hunt, especially in terms of the finale. Even in Dungeon Run the final encounters felt special compared to the earlier bosses, but in Rumble Run you were just facing off against another normal contestant.

With the latest expansion, the Rise of Shadows, Blizzard took all the lessons they learned and made what is currently in my opinion the definitive “Run” experience. If Monster Hunt and Rumble Run were like versions 1.1 and 1.2 of Dungeon Run, The Dalaran Heist is like Dungeon Run 3.0. It pulled out the best parts of each of the modes that came before, ditched what wasn’t working, and piled on a bunch of extras. In The Dalaran Heist you play as one of 9 different henchmen of the League of E.V.I.L, a nefarious organization made up of Hearthstone’s greatest villains. Over the course of 5 different wings you help them achieve their master plan of literally stealing the entire city of Dalaran.

What a nice city…I THINK I WILL TAKE IT!

What a nice city…I THINK I WILL TAKE IT!

Here are all the elements that The Dalaran Heist improved upon:

Replayability: The amount of replayability in The Dalaran Heist is kinda ridiculous. When you add up all the classes, starting decks, and hero powers there are 108 different ways to start off your run. Add to that the 5 different Twists for each wing, Anomaly mode (which adds a random extra mechanic to your run), and Heroic difficulty there’s a stupid amount of stuff to do.

Agency: In the previous “Runs” the only way you could modify your deck was by adding to it. This meant that while your deck at the end of the run was very powerful, it was often bloated with some bad cards you gathered along the way. In The Dalaran Heist they added in Bartender Bob, a friendly encounter that allows you to modify and/or remove cards in your deck. I adore this addition because it gives the player even more control on how they want to shape their deck.

I’d straight up give up villainy for Bob if he asked me to.

I’d straight up give up villainy for Bob if he asked me to.

Progression: I’m a huge fan of Roguelikes in general. They give a tremendous sense of accomplishment when you successfully complete a run. However, one of the biggest “feels bad” moments of a Roguelike (Hearthstone’s “Runs” included) is the “bad run.” By bad run, I mean the ones where you don’t get access to the items/upgrades/cards you want and lose early on. These runs happen from time to time and they feel awful because they make you feel like you wasted your time. But The Dalaran Heist alleviated this a bit with unlockable content. Initially when you start the mode you only have access to one hero power and one starting deck per class. However, as you defeat bosses and play specific strategies over multiple runs you can unlock additional hero powers and starting decks to use in the future. This really helps give a sense of progression and reduces the negativity of “bad runs” because even if your run isn’t successful you still made progress towards unlocking more content.

Storytelling: Each of the five wings of The Dalaran Heist focuses on one of the leaders of the League of E.V.I.L accomplishing their part of the heist. At the end you get to see it all come together as you successfully kidnap the city. While this is a nice bit of storytelling on its own, Blizzard has stated they plan to have the next two expansions this year build off of The Dalaran Heist’s story. This is a great way of expanding on the worldbuilding and lore of Hearthstone.

So what can we learn from all this? I think the biggest takeaway is that game design is an iterative process. Making something as fun and robust as The Dalaran Heist just wouldn’t have been possible right when Hearthstone was launched. Blizzard needed to first test out their boss battles in the early adventures, take a big chance with Dungeon Run, and then finally experiment and refine the formula with Monster Hunt and Rumble Run. Not every step in this progression was perfect, but that’s kind of the point. It’s okay to make the occasional misstep as long as you keep running forward.


Thanks for reading! If you liked it, please dungeon run your way over to our newsletter! You can also follow us on Twitter and Facebook that’s more your style. Stay lofty!

The Deck Archetype Spectrum

If you’ve played a deckbuilding game before, such as Hearthstone or Magic: The Gathering, you may have heard other players talk about deck archetypes. Usually it’s in the form of an argument about what archetype a given deck falls under, or which one is the best/most fun to play. But the thing is, deck archetypes can be confusing and sometimes the lines between them are fuzzy. So today, I wanted to propose my way of classifying deck archetypes: The Deck Archetype Spectrum!

There are two axes to my spectrum: Speed and Synergy. Speed refers to how quickly your deck intends to win the game. For example, an aggressive deck that aims to defeat their opponent as quickly as possible would be a fast deck, whereas a deck that intends to make every card count and to outlast their opponent would be a slow deck. Synergy refers to how much individual cards rely on the other cards in the deck to be powerful. An independent deck would be a deck where the cards in the deck are powerful in their own right. By contrast, a synergistic deck may use cards that are fairly weak on their own, but that can be used in specific combinations to become extremely powerful.

To give some examples, I have plotted some of the most common deck archetypes onto the spectrum below. I’ve also given some brief descriptions of those archetypes in case you’ve never heard of them. If I missed one of your favourite archetypes, leave a comment and let me know. Or even better, see if you can figure out where it should go on your own!

Spectrum Image.png

A few final notes about this spectrum:

  • Both Speed and Synergy are relativistic terms. This means that where exactly a deck falls depends on what other decks exist in the game. So while one day a deck might be classified as Aggro, if faster decks get created it might shift to being considered Tempo or Midrange by comparison.

  • No deck archetype is inherently more or less fun than any other. Fun is subjective, so play whatever kinds of decks you want. We here at Cloudfall won’t judge you for the decks that you like to play, even if we personally don’t enjoy playing them.

  • I consider this spectrum a work in progress. If you have any suggestions for making it better, let me know in the comments below!

Aggro (Fast, Independent)

Your typical aggressive deck. Its goal is to finish the game quickly by playing lots of cheap, strong cards. It will sometimes interact with what the opponent plays, but will prioritize attacking the opponent directly.

Face/Burn (Very Fast, Independent-Synergistic)

A hyper aggressive deck where it only cares about beating their opponent as quickly as possible. It will always attack the opponent directly unless it is immediately apparent that it will lose otherwise. Cards in the deck tend to be independently powerful so that the deck can be more consistent, but will sometime utilize synergies.

Midrange (Average, Independent-Synergistic)

The middle ground between Aggro and Control, Midrange decks can adjust their speed based on what they’re facing. This allows the deck to be the aggressor or play defensively depending on the matchup. These decks vary a lot in terms of synergy.

Tempo/Zoo (Fast-Average, Independent)

Similar to Midrange, Tempo decks can adjust their plan somewhat based on what they are facing. However, their goal is to control the speed of the game and push incremental advantage over time. This means that in most matchups it tends to be the aggressor. Cards in the deck are independently strong since the goal is to play the strongest thing possible each turn, and synergies can make achieving that goal inconsistent.

Control (Slow, Independent)

Your classic slow deck. The goal of the deck is to clear the opponent’s threats over and over again until you can play a few large threats to end the game.

Fatigue (Very Slow, Independent)

An extreme version of Control that plays for the long game. Unlike typical control, Fatigue doesn’t run a specific win condition, but rather doubles down on survivability to outlast anything their opponent can throw at them.

OTK/Combo (Slow, Synergistic)

Similar to other slow decks, OTK decks also try to stall the game. However, their goal isn’t to outlast but rather to draw a very specific set of cards which, when played together, instantly win the game. This means that they often run a lot of card draw in order to reach their combo quickly.

Tribal (Fast-Slow, Synergistic)

Tribal decks are a difficult type of deck to classify. They can be any speed, though typically they are fast-average. Their defining trait is that all of the cards in the deck are highly synergistic, such as all being the same tribe or card type. Often this means that individual cards are relatively weak, but grow significantly more powerful the more of them are played.

Mill (Average, Synergistic)

Several card games have the rule that if your deck runs out of cards, you lose the game. In those games, some decks are created specifically to force their opponent to draw too many cards too quickly so that they deck out and instantly lose. These decks tend to play similar to OTK decks, except that they can often reach their goal faster.

Thanks for reading! If you’ve made it this far, why not sign up for our newsletter? That way you can stay up to date on all things Cloudfall. You can also like us on Facebook, or you can follow us on Twitter.

Stay lofty!

Design Tips: Power Curves

For a while now I’ve been working on a power curve analysis of the collectable cards in Hearthstone. I started out trying to prove if specific cards/classes/mana costs are overpowered, but it’s evolved into a really valuable design exercise on understanding game balance. Hopefully one day I can share all the lessons I’ve learned, but today I wanted to take a moment to talk about one of the valuable things I’ve been learning about power curves. But before we get into that, let’s first establish what a power curve is.

What is a Power Curve?

Put simply, a power curve is the average value a player gets in exchange for spending a given amount of resources. It’s most often used when discussing card games like Magic the Gathering, Yu-Gi-Oh!, Gwent, and of course Hearthstone, but it is still applicable to most games. As a general rule, the more resources you spend the more value you should get, and this relationship tends to be linear.

Power Curve.jpg

A fairly ideal powercurve.

To give a simple example, take a look at these three cards I made:

Underpowered Bomb.png
Balanced Bomb.png
Overpowered Bomb.png

Thank you Hearthcards for making it so easy to make custom Hearthstone cards.

Let’s say for the sake of argument that in Hearthstone dealing 3 damage is worth about 2 mana (similar to an existing card Dark Bomb). This means that Balanced Bomb would be right in line with the power curve of the game because it’s cost (2 mana) is in line with it’s value (dealing 3 damage). Underpowered Bomb would be below the power curve because you’re paying the same cost while getting less value than you should. Finally, Overpowered Bomb is above the power curve because you’re getting more value than you should for the cost. Seems straightforward right? Here’s where it gets a little bit more complicated.

Independent vs. Contextual Power Curves

So far we’ve been comparing cards to an Independent Power Curve. An independent power curve assesses the power level of cards in a vacuum, and it does not factor in positive or negative synergies between cards. Using what we know, what’s the independent power level of this card?


Probly one of the best cards ever made.

Seems pretty terrible, right? I mean, 10 mana for a 1/1 that can’t attack is awful. However, this is when we need to consider a Contextual Power Curve. A contextual power curve assesses the power level of cards in conjunction with one another, factoring in things like synergies and the metagame. Now imagine we were assessing the power of Useless Ooze if this card also existed in the game:


See? I told you Useless Ooze was good.

Suddenly Useless Ooze seems pretty busted. But remember Useless Ooze is only strong in a world where something like Useful Ooze exists. If that card didn’t exist, it’d go right back to being trash.

Why does this matter?

This matters because all cards have two different power levels and you need to consider both when balancing a game. For example, it’s possible to have a card that is independently overpowered but contextually underpowered. Let’s take a look at Alien Queen as an example:


“Wait…aren’t I supposed to be human now? Make up your mind, Blizzard. Human. Zerg. Human. Zerg. Pick one!”

Independently this card seems extremely strong because it’s a 3/3 for 3 mana AND it gives +3/+3 to all your other Aliens. Assuming all your other minions in play are Aliens, that means she could be giving +3/+3 to 6 other minions for a total of +18/+18 (for context, a 3/4 is roughly worth 3 mana, as shown by the existing card Spider Tank)! Alien Queen seems pretty broken, right? Well, contextually speaking, maybe not. What if Alien Queen was the only Alien in the entire game? Suddenly it’s very difficult to get any value out of her battlecry, to the point where she might as well be a 3/3 for 3 with no additional text. Now she’s not only no longer overpower, but she may in fact be underpowered!

So let’s say we’ve determined a card too weak or too powerful. How can we fix it? Well, based on what we’ve learned today here are two approaches we can use when balancing cards:

  1. Adjust the card’s independent power level (i.e. decrease the value you get per resource spent); or

  2. Adjust the card’s contextual power level (i.e. create new cards that have positive or negative synergy with existing cards)

Both approaches are completely valid and have their own pros and cons. But make sure you’ve compared your card to both your independent power curve and your contextual power curve before making any changes. Otherwise you might make the problem worse than it was before.

TL;DR: Context matters when balancing your game.