Design and development of Hearthstone



This page aims to document Hearthstone's design and development, from the cornerstones of its initial conception to present-day philosophies and design paradigms.

For a quick overview of Hearthstone's milestone events, see the Hearthstone timeline.

Initial development

 * Team 5 are the team behind the development of Hearthstone. Until near the end of the closed beta, the team comprised only 15 members, the smallest team at Blizzard. This allowed them to work much more quickly and in different ways.


 * Team 5 was created specifically for the development of Hearthstone, with the intention of taking a different approach to game creation than that previously taken by in developing its games, with 50+ person teams and multi-year development cycles. Team 5 was created with the intention of working on a smaller scale but at the same level of quality. The mandate for the team was to keep the team very small, and to "think of ways to develop a game that might be non-traditional within Blizzard's walls". The smaller team required its members to be "old-school" "garage programmers" and able to "wear a lot of different hats", with far less specialization than that typically found in larger teams.


 * Hearthstone development began in 2008, along with the assembling of Team 5. However, for a long time Team 5 was a very small group mostly focusing on prototyping. Ben Brode states that his first records are from June 2008, but that "full-on development didn't start for a while after that". Development by the whole team appears to have started in earnest in spring 2012. Hearthstone was first announced at PAX East in March 2013, reached beta in August 2013, and was fully released in March 2014, roughly 6 years after development began.
 * The developers and Blizzard in general "have always loved CCGs [collectible card games]", with some such as Blizzard president having played them for more than 20 years.


 * It was decided early in the game's development that Hearthstone would be based in Warcraft, featuring characters, powers and lore from that universe. Many of the team, including Ben Brode, Eric Dodds and Derek Sakamoto, had previously worked on World of Warcraft, and many others were "big WoW fans". Hearthstone was inspired by WoW.


 * The development team wanted to bring the "great World of Warcraft flavor" to the game, but also to create a unique Hearthstone flavor that was a little different to World of Warcraft. Lead Designer Eric Dodds explained that this was achieved by taking the 'basic recipe' of World of Warcraft, "2 parts epic and maybe 1 part ... whimsical fun", and adding "a few more dashes of charm and whimsy … and maybe even a dash of irreverence," to produce a unique blend that was more comical and light-hearted than World of Warcraft while still drawing from its deep lore roots and epic feeling. The developers also drew on experience playing games like Peggle to imbue a whimsical fun and excitement into the recipe for Hearthstone.
 * Hearthstone was initially known as Warcraft Legends. It was a Warcraft Legends build named Fire and Ice that got the game greenlit for development. At the time at which it was greenlit, the game had the rogue and mage classes playable, with warlock a work in progress.


 * Early in the game's development, prior to the beginning of prototyping, deadlines for StarCraft II caused all of Team 5 except Eric Dodds and Ben Brode to be reassigned to that game for around a year. This isolation allowed the two designers to work in an even more focused way, with only the two of them to settle prototyping decisions, and were able to rapidly progress through myriad design iterations. For a long time Dodds and Brode used only pen and paper to create their prototypes, cutting pieces of paper to create test cards.
 * When the rest of the team returned, they discovered that Dodds and Brode had created a working Flash version of the game in their absence. The basic game was complete; according to Dodds, "We pretty much pointed at the computer and said — 'the game is done' ... Just remake that game over there." This Flash prototype established the core game for Hearthstone and according to Executive Producer Hamilton Chu survived to a surprising degree into the finished product.
 * The rest of the team's "tour of duty" on Starcraft II also had some impact on the development of Hearthstone, with the balanced yet asymmetrical nature of the game's different races "definitely [carrying] over into Hearthstone", according to Dodds.


 * Finding a seed
 * Initial work on the game lacked a central theme or "seed". The developers experimented with many settings and approaches, but eventually decided that none of these were what they wanted for the game. Derek Sakamoto explains that at that point, "the path to Hearthstone as it is today was not obvious or clear to [the team]".


 * One discarded prototype for the game included a World Map, with the player travelling Azeroth, engaging in single-player quests (similar to what would become adventures) to unlock new zones. A fully working Flash model of the game existed at this point, including functioning quests created by Ben Brode.


 * Other ideas included "Hearthstone in the forest", "adventure books", and 3D or even holographic cards. Reasons for rejecting these early models included too much complexity and a lack of aliveness in the display.


 * When it came to matchmaking, the developers knew they didn't want to simply have a counter or a 'searching for players' message. One prototype involved a representation of, with the two challengers appearing on either side of the arena, while the names of other matches appeared overhead.


 * Eventually, following a "super long meeting" between Sakamoto and several of the team's artists seeking a central vision or "seed" for the game, the idea of a game box emerged. Within the box, a World Map would show other battles taking place across Azeroth, with the player eventually zooming in on their own match in a specific area of the world. Once the player's battle was found, the game would shift into a gameboard view, taking place within the game box.
 * Once the idea of a box was established, this "seed" "grew roots into the team's minds". The box gave the developers a framework for what they wanted to do with the user interface (UI), with wooden panelling and trays that would fit inside the "jewelry box" or "music box" design, and would go on to influence elements such as the shop and even announcement banners. The box also eventually gave rise to the concept of arena keys, with this fitting the box concept. The box theme would later be extended to the rest of the UI.
 * The box also provided a solution to concerns about the UI distracting players from the game and breaking immersion. With the game conceptualised as one being played by the denizens of Azeroth themselves, the developers considered it critical that players felt that they were playing the game from within Azeroth, and saw a standard frame UI as "a sure way" to break the immersion of the experience.
 * With the central setting of a game box in place, the developers turned to thinking about the wider setting for the game, and where exactly the box was supposed to be situated. It was at this point that the idea of a tavern or inn was struck upon, stimulating another "burst of creativity". The idea of a tavern would go on to become central to the game, with the location even reproduced in the stage design of the 2014 Hearthstone World Championship.
 * Prior the development of the box idea, work on the game's setting had begun with a "really awful, flat parchment garbage" design. The developers soon decided that the game should be 3D, as well as feeling more physical. This included making the game feel valuable, both as a way to make players feel rewarded for putting time into the game, and as a way of recreating the experience of opening card packs in real-life CCGs. This included constructing the cards themselves using materials such as gems, leather and gold, something which would be impractical and excessively costly to do in real life.
 * Once established, this physicality served to unify the interface, through use of common elements such as bouncing and over-rotation. For example, the shop rocks and turns as the player selects packs to buy, with the chosen number of packs falling from the sky and impacting upon the shop board; while announcement banners would go on to be implemented as wooden signs, swinging back and forth. Later elements would expand the range of impressions while maintaining the sense of physicality, such as Goblins vs Gnomes 's electrical theme, which saw sparks and electric current added to the game's arena button. This not only provided a central theme for the designers, but helps the game's presentation feel cohesive, whether players realise it or not.
 * The title for Hearthstone itself is taken from World of Warcraft, where the is an item which allows the player to instantly return to a designated inn or tavern. Conceptually, by setting their Hearthstone to a new location, the player can establish that inn as their new base of operations, or home away from home. Inns also provide an experience buff in the form of the  bonus, rewarding players for ending their gaming sessions with their characters relaxing at an inn, again reinforcing the concept of inns as homely havens for adventurers. This fits well with the game's warm and welcoming theme, and also provided the game with its signature icon: the Hearthstone swirl, as seen in the World of Warcraft item.
 * For all these reasons, when the name Hearthstone was suggested during one of many "horrendous meetings" trying to find a name for the game, the designers found the name to suggest a desirable sense of warmth, safety, home and friendship, as well as "a lot of symbolism" and art potential.


 * Progression
 * The initial announcement of the project to the rest of Blizzard was met with "apprehension", "confusion" and uncertainty. According to Production Director Jason Chayes, "selling the game internally was one of our biggest challenges", due to the project's marked difference to other Blizzard games to date. Eric Dodds states, "When we announced the game a lot of people went, 'wow, that's a little weird that you're doing that'", while  describes the reaction as "Why are we making this?" The "evangelizing" process was ultimately achieved through rolling out playable versions of the game to members of the company, layer by layer - first the development team itself, then other development teams, followed by the wider company. Ben Brode describes how as the team were playing test matches, other Blizzard personnel walking past would frequently stop to watch the game play out. When the team held the game's first internal tournament, there were too many people who wanted to watch, forcing the company to set up "overflow areas" screening the tournament in other conference rooms. By December 2012 there was broad exposure to Hearthstone across Blizzard, with employees "really excited for the game".


 * Accessibility was a key focus for Hearthstone from the start. Making a game which was quick to learn and fast-paced to play was an early design goal, as was making a game which appealed to a wide audience. This design was conceived in contrast to many well-known CCGs, with their complex rules, long game times and inaccessibility to those unfamiliar with the genre. The designers envisioned a game you could play a quick session or two of within a 20 minute period, such as during a lunch break.


 * The designers also wanted individual cards to be easy to understand even for newer players. This was important so that players gaining new cards could immediately comprehend their function and strategise as to how they might be put into a deck, but also so that players encountering enemy effects during play could immediately understand their abilities and consequences, even if they had never seen the card before. Card text was heavily iterated in order to make cards as easy as possible for players to understand, and in some cases card complexity itself was also toned down, with cards that required too much thought simplified in order to expedite and ease play, both in terms of learning a card's function and in calculating its outcome.


 * With the desire for accessibility and speed came the eventual removal of concepts and mechanics from other CCGs, such as replacing more elaborate and cumbersome resource generation methods such as land cards with the automatically generated mana crystals. Making the players automatically gain a mana crystal each turn allowed the players to focus on more interesting gameplay decisions, but also allowed the game to more quickly reach its "climax" with powerful minions and cards entering play. This served to reduce each game's ramping-up period, and also to reduce the overall duration of each match.
 * This same desire also led to the simplification of card representations, especially minions on the field, and it was found that this worked well. This eventually resulted in the modern form of minion portraits, featuring only a picture and Attack and Health numbers.


 * While the developers wanted to make the game accessible, they also wanted make sure it still had depth, giving players a reason to keep playing the game for a long time. With individual cards kept relatively simple, one way they chose to add depth to the game was through the interactions of different cards and effects; though each card was easily understood, the outcome of combining multiple effects could lead to deep and strategic gameplay. The developers worked to build in interesting and strategic interactions between cards.


 * Another design intention was for it to be possible for sequences or combinations of cards to produce spectacular and often unforeseen outcomes. The developers wanted a lot of cards to support "player stories", tales of remarkable events which the Blizzard staff themselves began to trade early in development. The team considered these stories to be a strong and positive part of playing the game, even when the player ended up losing the match, and it was a design goal to include cards capable of generating interesting and memorable stories. The designers created many "story cards" with the potential for spectacular stories, but without specific outcomes in mind, and were often surprised to hear reports of unexpected synergies which they had not foreseen.


 * The developers wanted to make the game fun to play regardless of whether the player won or lost. In addition to making games faster and shorter, some more frustrating elements from other CCGs were also discarded in favour of a lighter and more fun game. For example, mechanics which the player was unable to actively counter, such as the ability to destroy the opponent's resources, or force them to discard their cards, were removed due to being non-participatory and not fun to lose to. The developers also wanted players to enjoy a number of "little victories" each game regardless of the ultimate outcome, such as successfully countering a cunning play by the opponent, and to enjoy the 'puzzle' of tackling the state of the board each turn.
 * Another feature that was removed early in development was "Combat Tricks". Similar to secrets, these allowed players with active Combat Tricks who attacked their opponent to play a card on their opponent's turn, disrupting their play. Ultimately it was found that removing Combat Tricks actually made the game more fun, as well as further improving its speed. This dynamic appears to have been partly reworked into secrets.


 * Early prototyping allowed for experimentation with and ultimately rejection of numerous gameplay styles, including the use of "fortresses". In addition to equipping the hero with weapons and armor, players were able to shelter their hero in a fortress with 20 Health. Players had to destroy the fortress before the enemy hero would "pop out" and they could attack them. However, the element was eventually discarded due to strategic issues. The developers even tried a version of the game without minions, but this was quickly discarded.


 * effects were also considered during prototyping. "Many" of these effects were tried, but ultimately it "just didn't feel right in Hearthstone.


 * The game was designed from the start to be an online game. This allowed the developers to do a lot of things they couldn't have done with a physical card game, including card effects like those of Nozdormu, Thoughtsteal and King Mukla. A lot of key gameplay and card design decisions were made in order to make the game work well as an online game.


 * Despite the online focus of the game, making the game feel physical was an early goal, later implemented mostly through art, animation and sound design. The developers wanted to ensure that the tangible feeling of collecting cards and opening new packs found in CCGs was also present in Hearthstone, despite the game being digital.
 * The game was initially experimented with in the form of a physical game. However, after this initial phase the game was moved onto a web interface. This would go on to define much of the Hearthstone playing style, with simplicity of display and interactions key to the design.


 * The team tried various conceptualisations of what the game itself represented, before hitting upon the idea of the players controlling pieces on a board, with the game taking place in a tavern within the world of Warcraft itself. This went on to affect game board and interface design.


 * Hearthstone was initially conceived as a PC game, due to Blizzard being primarily focused on this platform. However, fairly early in development, in response to the growing popularity of games on mobile devices, the team realised that the game could work well on a mobile platform, and later committed to this as a specific goal. For information on the eventual release and adaptation of Hearthstone for mobile devices, see below.


 * Once the team had decided that the game would be released for mobile devices such as iPad, the developers were forced to assess aspects such as interface design with this eventual goal in mind. For example, features intended to be discovered through mouse-over had to made with an alternate mobile-friendly interaction in mind, in order to ensure that it was possible to later make a mobile version of the game that was intuitive and easy to use, while still matching the game's PC version. Because this goal was kept in mind through the development process, relatively little redesign was necessary for mobile release.


 * The Arena - originally titled 'The Forge' - was first conceived as a way to incorporate 'draft mode' style play into the game. Drafting with a physical CCG involved players passing round packs of cards, drawing individual cards until they had each built a deck - something many of the developers enjoyed, but which would be difficult to implement within Hearthstone. To solve this problem, the developers implemented asynchronous drafting, allowing each player to separately - yet randomly - build, or 'forge' a unique deck. Early versions of the Forge had players keeping all the cards they drew for their deck. For more background on the Arena's development process, see Arena.


 * Hearthstone originally featured a "graveyard", a location where played cards and destroyed minions would be placed when they were removed from the game. This feature was removed during the game's alpha, due to "making the game unnecessarily complex without adding much", when the game was already "deep" enough. The finished game's battlefields still feature skull icons next to the players' decks, indicating the previous location of the graveyards. Earlier versions of the graveyards can also be seen in some images of the game's alpha battlefield designs.


 * The developers explored various ways of speeding up the game, including players drawing a card from their deck at the end of their turn, instead of at the beginning. This meant player could spend their opponent's turn planning how to act next turn, allowing them to make their play more quickly when their turn came around. However, in tough situations this system often led to a feeling of powerlessness, since the player would know in advance that they were not going to draw next turn a card that was capable of dealing with the state of the board. Alternatively, the player might know in advance that they would have access to the card they needed, thus taking "the drama and excitement" out of the game.


 * Alpha iterations of the game featured 4 or 5 keywords which were later removed from the game. Several minion types - Blood Elf, Draenei, Dwarf, Gnome, Goblin, Human, Night Elf, Orc, Tauren, Troll, Undead, Worgen, Elemental, Mechanical, Nerubian, Ogre, and Scourge - were also experimented with but eventually removed. 'Mechanical' was notably reinstated with Goblins vs Gnomes as the Mech type, while the Dragon type was kept in the game despite lacking even a single instance of synergy, due to related cards either being removed during alpha or never implemented. The first Dragon synergies would not be introduced until Blackrock Mountain, more than a year after full release.


 * Heroes


 * The game's starting class was originally hunter, but it was changed to mage because the developers wanted to use a non-weapon class, presumably to reduce complexity.


 * At one point the base game had two heroes for each class, one representing the and one representing the . The Alliance warlock was Wilfred Fizzlebang, whose presence helped to inspire the creation of the Lord Jaraxxus minion. This was later reduced to one for each class, although alternate heroes have since added expanded this number.


 * In most cases, several possible characters were considered before the final hero for each class was decided. This can be seen in development screenshots (see sections above), with heroes such as as priest,  as mage, Tirion Fordring as paladin, and Edwin VanCleef as the original rogue.   On selecting the right character to represent each class, Ben Brode states "In general it’s hard to find a perfect fit."


 * Faction, race and gender representation were also factors in the selection of each class' hero. The original selection of heroes features 7 males and 2 females; 3 humans, 3 1/2 orcs (Rexxar being half-orc, half-ogre), 1 night elf and 1 blood elf; and ties roughly evenly for faction balance. Valeera was chosen for rogue partly due to being a blood elf female, as opposed to options such as or Edwin VanCleef (human male, already represented by both Anduin and Uther), or  (a female half-orc, making the fifth orc hero), specifically in order to add more diversity to the cast of characters.

Finding a style

 * Ben Thompson was the first artist to be brought onto Team 5, later becoming Lead Artist when other artists joined the group. As such, it fell to Thompson to define the artistic style for what would go on to become ''Hearthstone.
 * Thompson used his training as a traditional painter to produce the game's style, a primarily 2D form reminiscent of early Blizzard art for games such as the original Warcraft and Diablo. As a strictly 2D artist, Thompson turned his lack of capability in 3D image manipulation into a strength by focusing on creating an earlier style of art that would "celebrate the feel that I know and loved and ... wanted to bring back into a video game setting". Thompson was also central to the work to create a digital card game that still evoked the feelings of a physical card game, while expanding from the extremely 2D imitations of such games into the additional capabilities of a purely digital experience.

Collectibles

 * From early in the game's design process, there was a focus on making the game's 'collectibles' - cards, and what would become termed 'minions' - attractive and compelling. Making the game "look as good and as epic as it played" was considered very important.
 * The designers wanted the collectibles to feel like physical objects, but also to feel like they belonged in the world of.
 * Thompson was not initially convinced that the game should retain the focus on actual cards, considering that the digital medium gave them the opportunity to move beyond cards and explore other forms of collectible. The designers experimented with many visual styles for Hearthstone, including "voodoo dolls, gear-driven contraptions, magical jars, moonwells and even papercraft minions."
 * However, as Thompson and the team experimented with various forms and styles, eventually they realised that by abandoning cards they were throwing away a powerful foundation for the game, the "common language" of familiarity with cards; instead of instinctively understanding their use, form and value, players would have to be educated about the game's collectibles, making matters unnecessarily complex. The designers eventually found a more traditional card-based visual design "hard to beat" due to its simplicity and players' familiarity with this design.
 * As a result, Thompson and the team decided to return to the card as the basic form for the game's collectibles, but to do it their own way, combining the familiar strengths of the traditional card design with the exciting new possibilities of the digital realm. For example, minions cards were designed to evoke the basic feeling of collecting and holding a card in the player's hand, but once played to transform into a less traditional and more dynamic form of collectible, moving of its own accord, crashing into other minions, and crumbling into pieces when destroyed.
 * Earlier card-based designs featured lavish styles, expanding upon the standard forms of CCG cards, but without the physical limitations. However, these designs turned out to be too ornate and distracting from the base stats and casting costs of the minions, as well as taking up too much space on the battlefield. At this point the developers struck upon the idea of having the minions change state when they were played from the hand, something else that would not have been possible in a physical card game.
 * An early visual design idea for minions involved having minions' frames reflect their 'race' or type. However, this was found to be too overwhelming with so many different types of frame.
 * Ultimately the developers decided to focus on the three most important features of a card: its numeric values, its title and its art. The developers then worked-out a design which conveyed all of these elements as simply as possible. The resulting design would become the foundation for the eventual form of the collectibles in the game.

Battlefields



 * "I get the board from the 2d artist as a finished painting. Then I throw it in Maya and start to lay down basic shapes and project UV's from the games camera angle and then bring that into Unity.
 * Then, corner by corner, I flush out the details and export again to Unity checking perspective issues and clipping and such. It usually takes me about 4-5 days just for the board itself. Then I get a week or so to do all the clickables, effects and animations." - John Zwicker


 * The game's battlefields are created by using projected textures on 3D models. Paintings created by Ben Thompson and Jomaro Kindred are given to John Zwicker, who then maps them onto 3D models created using the Maya image editing software.


 * The models themselves are slanted and 'deformed', in order to accept lighting and cast shadows, as well as to create the angled viewpoint from which the gameboard is seen. However, the top of the board is far more deformed than the bottom.


 * The boards' bonus features or "clickables" are created by John Zwicker, with assistance from technical artists Kyle Harrison and Becca Abel.

Cohesive design

 * In order to present a cohesive artistic experience, there are certain standards that are maintained for variants of objects such as card packs and card backs. The most central of these is the presence of the Hearthstone swirl. Found on both card packs and card backs, this is originally taken from the game's World of Warcraft namesake, the . For card packs, a horizontal band is also required, while for card backs, spikes from around the central swirl are almost always used in some form or another.

Music

 * From the start, the music for Hearthstone was intended to feel familiar to fans of the Warcraft franchise, but also unique and distinctive. One way of achieving this was to add key themes from World of Warcraft, such as the  theme heard briefly in the game's trailer. This helped to achieve continuity and place the game "in the Warcraft universe."


 * The other key focus for the creation of Hearthstone music was to make it light-hearted and "whimsical". The music was also intended to be fast-paced and energetic to complement the game's design for being played in short bursts.


 * The game's trailer provided a key opportunity to convey the game's tone to players. The music for the trailer was composed by Jason Hayes, who had previously composed music for Blizzard games including World of Warcraft. It was decided that the music should combine a sense of myth and magic with a sudden twist into light-hearted fun, with a notably lighter tone than some of Hayes' other work, conveying the two key elements of Hearthstone. Contrasting the 'epic' sounds of the Warcraft themes, Hayes used folk instruments like the accordion and penny whistle to immediately convey a Celtic tone and to evoke a warmer, less dramatic mood, at home in the setting of a bustling tavern.


 * Jason Hayes also worked alongside fellow World of Warcraft composer Glenn Stafford to produce many of the musical "stingers" in-game, played when a legendary minion is summoned. Set against the backdrop of the game's background music, these were intended to "add a little extra epic punch" to the musical experience, again combining an overall light-hearted tone with brief bursts of a more serious and epic tone.


 * The game's background music was composed by Peter McConnell, best known for composing music for LucasArts'  games. It was decided to have the musical style suggest a small ensemble, like a group of musicians playing within the tavern. McConnell decided that the music should have "a swagger, and an attitude, and a certain groove", to convey a sense of "tough guys and bar-room brawls". Combining Celtic with blues and rock influences, McConnell came up with what would become Bad Reputation, one of the game's core tracks. The piece uses wind instruments like the contrabassoon to "give an ancient flavor", while folk instruments like guitar, fiddle and harp combine Celtic tones with a "steady blues groove". Bad Reputation also incorporated Warcraft themes, intended to "work like little cheers, as if they are from the player's faction, or taunts from the faction the player is facing." The piece would go on to set the precedent for the entire score, including the game's title music.


 * The Finding Opponent screen features music from the RTS game  while the player is waiting for a match to be made. Prior to Hearthstone 's first announcement, this music had been put to use in a similar fashion as background music for World of Warcraft 's s.


 * The game's first major content addition, the Curse of Naxxramas adventure, saw the creation of several pieces of new music available when playing that specific content, as well as one new piece added for the main game. Director of Audio Russell Brower stated that this will be their approach to the generation of new music going forward, described as "in [the] World of Warcraft tradition", with the game's playlist slowly growing over time without removing earlier pieces.

Sound

 * Sound and animation come near the end of the card development process. All sound and animation is done in-house, leading to "a ton of collaboration" between the designers, sound and art teams. Ben Brode in particular works "a lot" with the dialogue and voice actors for the game's soundbites and sound design.


 * Soundbites
 * Hearthstone features a substantial amount of quotes and emotes. Another example of something that is not possible in a physical card game, these soundbites add extra depth to the game's characterisation, illustrating the personalities of minions and heroes. Many of the minion and especially hero emotes have become memes in their own right, such as "Well met!", "Mistakes were made." and "This is our town, scrub./Yeah, beat it!", while many draw upon pre-existing phrases or memes, such as "LEEROOOOOOOOOOOOY JENKIIIIIIIIIIIIINS!", "You are not prepared!" and "YOU FACE JARAXXUS, EREDAR LORD OF THE BURNING LEGION!".


 * Many players are known to refer to minions by their summoning quotes, such as "Taz'dingo".

AI

 * As of March 2016, the developers are "always" working on improving the AI.
 * The game's basic AI (as seen in Practice mode) was developed by Brian Schwab. The game's designers asked Schwab to create an AI which would "imitate [an] intermediate player". It was also intended to keep the AI lightweight in order to reduce player waiting time when playing against the AI as much as possible, contrasted by Schwab to games such as Magic Online which sometimes saw players waiting 10 seconds for the AI to make a move. In order to ensure this, the designers asked Schwab to create an AI capable of running 1,000 games simultaneously on a single server core. Schwab ultimately exceeded this goal by a power of 10, with between 9,000 and 11,000 games running concurrently on a single server core by the end of the project.
 * In order to discover how exactly an intermediate player played the game, Schwab spoke to many players around Blizzard and watched how they played, including long-time card players and those with little or no experience. Schwab observed that while the higher-level players played around the game's meta - identifying the opponent's strategy and deck construction, and planning around complex combos for both the opponent and themselves - intermediate players determined their actions in a far simpler way, essentially seeking to gain the most 'resources' or strengths within the game each turn.
 * To model an intermediate player, Schwab created an AI which measured each card and action's potential value using a "very simple scoring system", based on stats, effects and the current state of the board. By imitating an intermediate player, the AI therefore does not attempt to anticipate the opponent's actions, but simply chooses the most valuable or 'resource'-granting card to play or action to take at any given point. In determining combat actions, the AI compares how many 'resources' they will lose, versus how many the opponent will lose.
 * Another consequence of this is that the AI does not use any preplanned combinations, instead assessing the best individual action at any given time without thought of the next. The only exception is when the combination of two actions will allow the AI to directly win the match; in this case the AI may chain those two actions in order to defeat the opponent. Similarly, the AI features very few "holdbacks", where a card is held until it can be combined with others to better effect.
 * The AI does not include any cheating, "omniscience" or fixing of otherwise random processes.
 * Creating the separate Basic and Expert AI difficulty levels was achieved in a few ways. Firstly, the Basic AI will sometimes intentionally choose an option that is not the highest-rated option available. Schwab states that while in some games this would directly reveal the AI's lack of intelligence, due to the player being uncertain of the content of the opponent's hand, in Hearthstone less than optimal choices are usually not apparent to the player. Secondly, the Basic AI intentionally chooses less desirable options when mulliganing. Schwab initially tried having the AI discard all low cost cards during the mulligan, in order to weaken the AI during the earlier rounds of the game. However, by increasing the amount of mid and late-game cards in the AI's hand, this ultimately had the effect of making the AI more difficult to defeat, especially as beginner players typically failed to achieve much during the early rounds, as once the mid-game was reached the AI has far more "ammunition" to make use of. The biggest factor in differentiating between the Basic and Expert AIs was in adjusting the decks themselves; if the AI was too powerful, the designers intentionally weakened the deck in order to lower the AI's success rate.

Connectivity

 * Hearthstone communications take place over two types of network connection. The first is an encrypted connection to the Battle.net servers. It handles all Battle.net related activities such as account verification, friends, messaging and notifications. The Hearthstone servers (known as Pegasus) are responsible for handling all non-match related activity, including card collection, decks, hero levels and the shop. The secure Battle.net connection transmits all these activities back and forth with the Hearthstone servers. There is no direct connection between the client and the Hearthstone servers, all traffic goes through the Battle.net connection. The second type of connection we call the session connection and it is used when playing a match. When matchmaking completes, the Hearthstone servers allocate a session server for that individual match, both player's clients receive the address of this session server and attempt to connect to it. The session server deals with all match related communications such as game logic, board state and the sending of emotes. Any spectators of the match are also connected to the same session server. When the match is over, all players and spectators are disconnected from the allocated session server.


 * It is possible that a player can disconnect, meaning they lose their connection to the session server. When this happens, it is not possible for the disconnected player to make any more actions in the match unless they are reconnected, because they are not able to communicate with the session server. Due to this lack of communication, you will not be able see any actions your opponent may make while you are disconnected, such as mouse-over visual effects or emotes.
 * Do not confuse disconnection with an animation hang. Animation hang is a client-side bug which causes the game client to halt for one or both players with the inability to proceed or even concede the match. In this case, you can see your opponent's mouse-over visual effects and emotes (due to these being routed independent of game turn processing ) but are unable to take most actions. Your connection to the session server is still intact (this can be observed by looking at the game log) and the only way to fix this is to restart your client and reconnect, because reconnect will clear any animations that are stuck. One of the most known animation hangs is the Floating Card bug.

Release on mobile devices

 * Due to players on all devices being placed in the same matchmaking pools, it is not viable for the team to release a patch for one platform without also making that patch available for all other platforms. As a result, it will be necessary to have patch versions available for each platform before the patch can be released.

iPad

 * Making Hearthstone available on iPad required a thorough review of the mechanisms for interface interaction, with optimisations for touch screen rather than mouse control. It was a design goal during the adaptation process to make it possible for players familiar with the PC version of the game to transition very quickly to the adjusted interactions of the game's iPad version.


 * One specific difference between the PC and iPad versions is the interaction for targeting minion Battlecries. The original PC version of the interaction was initially implemented, but this was found to confusing and non-intuitive and had to be adjusted. The original single "fluid motion" was eventually broken into multiple steps. Targeting visuals for such mechanisms were also adjusted, with viable targets for effects highlighted by growing larger. Battlecry targeting was the most challenging part of adapting the game's interactions for the iPad, and had the hardest transition to that platform.


 * iPad also lacked the facility for right-clicking and hovering, requiring the team to find new ways to display tooltip minion information, eventually implemented through holding down on the relevant interface element.

Phones

 * Development for smaller mobile devices such as iPhone may require the developers to more substantially overhaul the interface, due to the extremely limited screen space on these devices. This was the reason for the game not being launched for phones around the same time as iPad, due to the far larger amount of work necessary to successfully achieve the game's translation to this platform.

Pricing

 * Making Hearthstone viable as a free-to-play game was a design goal, with "making sure you could actually earn all the things you needed to be competitive" being "something [the developers] put a lot of time into."
 * With the game's beta, the success of prominent streamers in using cheap or 'free' decks (formed without directly purchasing any cards) to reach high ranks in Ranked play, the developers felt their approach to the game's pricing model was "validated", stating in April 2014 "Seeing that happen and seeing people construct the decks they need to play at the very top levels of play and seeing that happen through the community was awesome. So overall we felt very good about how that evolved and how in some ways helped validate our approach, and we feel pretty good about the direction that's been heading."
 * Hearthstone is Blizzard's first free-to-play game.

Card changes

 * ''For a list of all card changes made to date, see Card changes


 * For most of Hearthstone 's initial development, there were only two designers: Ben Brode and Eric Dodds. The two took personal charge of every design and card balance decision. The two used internal testing to experiment with ideas. However, in January 2014, with the game due to be released to a far wider audience with its open beta, Mike Donais was hired specifically as balance designer. The designers also continue to work with a wider balance team to "really hone in on the specific balance changes we want to make before we go to a live audience."
 * While the alpha saw numerous changes to the game, and a number of significant adjustments were made during the beta, Eric Dodds has stated that the developers plan on making "very few card changes, unless they are absolutely necessary", with the intention of not changing cards except "in emergencies." The developers feel it's important that the cards "feel solid", and that changing cards which players have spent time and resources to obtain would undermine that feeling.
 * When a certain card in the game starts to appear too strong, the developers try to find a way to address the card without changing the card itself, such as through the introduction of new cards which will allow players to counter that card. In response to certain decks dominating the meta-game, the overall strategy is to provide players with new "tools" to shift the meta themselves, rather than changing existing cards. Introducing new cards is the developers' "preferred method of changing the current state of the game".
 * The developers prefer for players to create their own counters to dominant decks, and also try maintain stability within the game. For this reason changes are often made minimally, before giving the meta time to reestablish itself, the changes hopefully having allowed the problem to be righted without further intervention. Brode states, "there's 'problems that might fix themselves with the next set', 'problems players might fix themselves' & 'problems we must nerf'."
 * The developers are strongly resistant to simply 'buffing' or improving 'bad' cards, preferring not to change existing cards "unless absolutely necessary".   As with overpowered cards, the addition of other cards may change the value of a 'bad' card without direct modification. Developments in the meta also frequently bring undervalued cards back into the spotlight, with cards that a month earlier were considered objectively terrible choices suddenly included in top-level decks.
 * Critically, Brode explains that by changing cards to counter dominant decks, the developers would "place the onus on Blizzard" to solve challenges within the meta. Players may then be tempted to simply sit back and wait for the developers to make the changes they consider necessary, since a nerf seems likely to be made. By being consistently conservative in making changes to cards, the onus is instead placed on players themselves to develop counters to those decks. Players are therefore driven to innovate and find their own solutions using the tools provided, exploring underdeveloped archetypes and experimenting with panned cards, since they know that the problems will not be solved for them.
 * In June 2015, Brode spoke at length about the necessity and value of 'bad' cards in Hearthstone, as well as arguments against buffing cards, in a shared video with Kripparrian. Brode explained that while seemingly simpler than creating new cards, the balance testing necessary when changing an existing card is still extensive, given the impact on the value of all other cards. In addition, changing old cards "corrupts the experience" for returning players, who would come back to find the value of familiar cards to have changed in their absence. Brode also discussed how changing existing cards would "interrupt the cycle of tuning" in the overall cycle of the game. Immediately after the release of new cards, the meta is in flux, with players experimenting with entirely new archetypes and re-assessing the value of old cards as well as new ones. Following this relatively chaotic period the meta tends to settle, with less extensive experimentation with basic design, and more focus on refining existing decks. Brode explains that this "polishing" stage is important for the game, as players get the chance to focus on honing specific decks and finding the absolutely optimum formulation of that archetype, in contrast to the rough-hewn designs seen earlier in the cycle. By changing existing cards later in the cycle, this polishing stage would be disrupted, with players instead sent back to the drawing board in experimenting with new possibilities following the changes. Such changes would therefore prevent decks from reaching their fullest potential, and remove from the game this more refined stage of play.
 * Brode has also stated that "it's ok to have some bad cards in the game." and that "bad cards exist for good reasons, aside from being relevant at high rank play." While some players may consider poor cards to serve no purpose in the game, features such as the Arena pick system and random summoning and transformation effects such as Recombobulator, Webspinner and the various Shredder minions can act to bring them into play even when not directly chosen in constructed play. Bad cards also serve the purpose of expanding the card pool, and provide alternate options for players as their collections develop. Most critically, while many cards may never be featured in top-level professional decks, most are chosen at some point as part of a variety of more diverse and inventive decks, and by different players due to personal preference. The perception of a card as being 'just bad' is also highly subjective.
 * In his June 2015 video, Brode explained that it is "impossible" to have no 'bad' cards, due to the nature of the meta and the relative value of other cards, which would inevitably see some cards dubbed 'bad'. He also discussed how 'bad' cards can provide a touchstone for the evaluation of other cards; be helpful for beginners, by teaching them how to assess the strategic value of cards; and can provide community value, such as the variety of efforts by players "trying to win with Magma Rager".
 * Rarely, new cards may provide strictly superior versions of pre-existing cards. While it is not the developers' intention to cause new cards to overshadow existing cards, Ben Brode has stated "if I print a card, then realize it is so bad that nobody plays it, I should be able to print a better one." This statement was made in reference to cards that "might as well not exist", with the emphasis being that the developers should not be restricted from creating new, playable cards simply because they eclipse previous, extremely poor cards.
 * Brode has stated that making cards which are strictly better than other cards does not necessarily represent power creep, for several reasons. For example, Ice Rager is a strictly superior version of Magma Rager. However, Magma Rager is in the Basic set, and thus available to all players, while Ice Rager belongs to the Grand Tournament set, and thus requires collection through cards packs or other rewards. As such, Brode explains, this does not constitute power creep, but rather a path of progression for players to obtain upgraded versions of the cards initially available to them. Dr. Boom vs War Golem is another example, with the former card being far superior than the latter. However, the former is an extremely costly/difficult card to obtain, while the latter is available to all players in the Basic set; as such it represents a path of progression rather than simple power creep.
 * Eric Dodds has stated that the main reasons for cards being changed are:
 * causing non-interactive games;
 * being frustrating to play against;
 * causing confusion or not being intuitive enough;
 * being too strong compared to other cards of the same cost; causing a specific build or style of play to be too strong;
 * being too weak
 * For a full exploration of these principles, see Eric Dodds' Hearthstone's Card Balance Philosophy blog.


 * Decisions regarding card changes are strongly determined by actual performance in-game and the "back-end stats" available only to the developers, but are also affected by player feedback.
 * The developers may be more likely to make changes to existing cards to coincide with the release of new cards, especially large sets such as expansions. This is because while the developers prefer not to directly adjust the meta, the introduction of new cards is already expected to result in "upheaval", and with "everything changing" it feels more acceptable to also alter existing cards. Another perspective on this is that the introduction of new cards can be expected to substantially change the value and functionality of specific pre-existing cards; since change is therefore inevitable, directly adjusting these cards' text or stats is less significant, and could even result in less overall change to the role of the card, such as by increasing a card's mana cost in order to prevent new synergies from making it overpowered.

Future trends

 * By the release of Goblins vs Gnomes, card text had already become more complicated than at release. The developers expect this trend to continue as the game progresses.

Accessibility

 * The developers are aware that the increasing number of cards is making the game less and less accessible for new players. While the original game contained only 382 collectible cards, less than two years after its release it had already nearly doubled in size to 743 collectible cards. As of November 2015, the developers have no immediate plans on slowing down the rate of card release. However, they are aware that the rate is ultimately unsustainable.
 * As of November 2015, the designers are discussing possible solutions to this problem, and intend to announce plans "in the coming months".
 * There are some processes already in place to help ease new players into the ever-growing game. The team has a member specifically dedicated to the new player experience. Tavern Brawls also offer a way for new players to play with premade decks, and gain additional card packs quickly, accelerating the growth of their collections. The prevalence of Classic card packs is another way the designers aim to help new players, as Classic cards are considered to be more essential on average than those from more recent card packs.

Card creation

 * The process of designing new cards can take place in several different ways. The card's starting point may be its function, flavor, or occasionally even a specific piece of artwork. According to Senior Producer Yong Woo:
 * ''"First design team comes up with the theme for the given expansion. In the case of GvG, the theme was going to be Goblins, Gnomes, and their wonderful mechanical inventions. The backbone of the set is designed around that theme: e.g. we definitely need to have a Goblin Shredder! So those cards start from the flavor and then game mechanics is built around it. "Piloted" mechanic arose from the hilarious image of random denizens of Azeroth ejecting out of these things. Other cards start from interesting and fun game mechanics we want to highlight in the expansion and then we come up with the best name/art work that would fit that card mechanic: e.g. Bouncing Blade."


 * On occasion the designers will create a new card on the basis of an "awesome" piece of art. In other cases the designers will create a function and mechanic for the card, before then choosing a title, character and overall concept from a list of possible options, seeking one that is a good fit for the card's design.


 * Other factors initiating the creation of new cards include wanting to create more cards of a certain mana cost, function or type.


 * Pre-existing Warcraft lore is also a factor. Iconic class abilities may be priorities for inclusion, with their concept serving as a basis to develop the card. Hamilton Chu states "There’s the lore direction, where we go ‘Rogue has got to have Eviscerate, right? That’s just part of being a Rogue. So how does that work?’"


 * Each new card must be balanced around all existing cards, making the creation process an extremely delicate one. At the same the designers try to create interesting new cards, which will interact with existing cards in a variety of ways, creating new combinations and synergies without completely unbalancing existing ones.


 * The main two approaches to card and boss encounter design are known as "top-down" and "bottom-up".
 * A top-down approach is based on lore, or a strong central concept, with this providing a specific idea of what the card should be about, elements the encounter should incorporate, or even very specific mechanics that should be included. A top-down approach usually aims to evoke a specific feeling or fantasy, often from established lore, and should satisfy the conceptual 'dream' and feel right to the player.
 * A bottom-up approach takes its lead from gameplay or other practical concerns. A bottom-up design might be invented from scratch, with no underlying lore or conceptual framework, or simply be focused primarily around what is fun to play with, or what will introduce constructive elements into the game. Bottom-up designs may even lead to the invention or adaptation of lore in order to explain or fit the card's design.


 * In practice, the designers usually aim to combine both approaches. A design that is purely top-down might feel magical but lack satisfying gameplay, while an excessively bottom-up design might be fun but not fit with players' ideas of the character or lore depicted. By combining the two, the designers aim to produce a high-quality gameplay experience that also feels authentic to the fantasy and concept of the game.


 * Readability is very important not only when deciding on card text, but also when designing the cards themselves. According to Eric Dodds, "We have a rule when we’re building a card that if someone reads the text for the first time and they don’t immediately understand what the card does, we need to change that power or re-word it – that is critical.”


 * According to Ben Brode, the team prefers to focus on card depth over complexity, explaining that in his opinion "cards with the highest ratio of depth to complexity are the best designs". Conversely, he states that his least favourite cards are "those that are very complex, but not very strategically deep." He explains that while making cards complex is easy, making them deep is more difficulty, while making them both deep and simple is the most challenging, but also what he believes the team should be aiming to achieve.


 * Cards are created in a tool called "HearthEdit", using a simplified programming language which can be used directly by designers. This allows designers to experiment and iterate on card design and function without technical assistance. Once created, cards are plugged directly into the game, allowing new cards to potentially crash the game servers.


 * Card art
 * For more background on the creation and selection of card art, see Card art.

Process

 * The card design process is split into various phases. In addition the design team is split into various sub-teams, with each focusing on a specific area or phase of design. For more, see Team 5.
 * Initial design looks primarily at concept and the "thematic direction of [the] set", including flavor and story. During the creation of a new set, the initial design team "lays out the vision for the set", determining the setting, theme, and mechanics, and "sketch out ideas" for new deck types that "support the set vision". Initial design "handle the names, art, and flavor text for each card", as well as "a solid pass on what every card does and costs".
 * Final design looks at mechanics, cost, stats and what would generally be considered "balance", trying to predict the new meta and ensure it remains varied and diverse. The last step before release, final design is responsible for "[making sure] that the cards and decks are fun, intuitive, clear, and balanced", as well as ensuring the set is of an appropriate level of power and complexity, removing or polishing overpowered and "needlessly complicated" cards. Final design involves a significant amount of playtesting, to test for balance against a wide range of decks - "strong enough to compete with current decks but not so strong as to obsolete them" - and to ensure that the new cards and deck types are fun to play. As of April 2016 there are four people specifically working on final design, including Mike Donais and Dean Ayala. However Yong Woo confirms that the number of people working on game balance is far larger, stating "Adding cards and balancing the gameplay involve a lot of people and a complicated process. We have more than a hundred people working on those at a time, so it's definitely not just four people."
 * In a nutshell, while the job of initial design is to "figure out what can be made to be fun", it is the job of final design to "[take] the set and [make] it as fun as possible".


 * Set design begins with initial design, who will spend around four months on the set. Once initial design have finished, the set is passed to the final design team. Final design spends around another four months working on each set, finishing around two months prior to release. Most cards consequently stop changing about two months before release, but the developers do at times change cards as close as one or two weeks before release.
 * While each set passes through these (and other) phases over the course of its production, multiple sets are in production simultaneously, with designers who mostly work on card design working on multiple sets at any given time.


 * Almost every card is reworked at some point during development, with some seeing several revisions. Ben Brode describes iteration as "a big theme for us."


 * Cards typically become "cooler" as the designers move through the development process. As a result, earlier iterations of card designers often seem boring in comparison to the finished versions.


 * Sound and animation come near the end of the card development process. All sound and animation is done in-house, leading to "a ton of collaboration" between the designers, sound and art teams.


 * The developers work very closely when developing new cards, with the physical proximity between the team makes collaboration "super easy". "Brainstorming" often begins in email or excel, then progresses to more focused discussion once the designers have found "a direction we like".

Class design

 * One rule for the developers in class design is to make sure that each of the classes "feels very strongly like its World of Warcraft class." This encourages the maintenance of constraints such as which classes have access to weapons.


 * As of Goblins vs Gnomes, adding specific minion type "themes" to classes is one "fun" way in which the designers aim to "make each class feel distinct and have its own flavor".


 * Along with Hero Powers, class-specific cards are the main way of creating distinct playing styles and experiences for the different classes. However, in some cases effects allow players to use cards belonging to another class, such as Unstable Portal, Piloted Shredder or Grand Crusader. When a class starts imitating a behaviour specific to another class, this is known as "class bleed". In many cases, a small amount of class bleed can be fun and interesting, but too much can cause the classes to lose their sense of uniqueness or even their identity. As a result the developers try to maintain "the right amount" of class bleed in the game. For example, when designing the Discover keyword, the developers intentionally restricted the effect to class and neutral cards, as the keyword's initial ability to draw cards of any class added too much class bleed to the game.

Lore

 * The majority of characters and abilities in Hearthstone, as well as many of its weapons, are taken specifically from Warcraft lore, and in most cases are found in World of Warcraft. Additionally, adventures such as Curse of Naxxramas faithfully reproduce setting, bosses and even abilities from World of Warcraft.


 * While many are based on very specific sources, others are more loosely set. For example Coldlight Oracle is a specific type of mob found in, while Coldlight Seer is obviously a murloc of the same tribe, but does not actually exist in World of Warcraft; meanwhile Old Murk-Eye is a specific named murloc found in , the slaying of whom is the subject of its own quest. Likewise, while most class cards are based upon specific abilities from World of Warcraft, others are simply created to fit the feel of the class. For example Preparation is a rogue ability from World of Warcraft, while Cold Blood is a former passive rogue talent which was removed prior to Hearthstone 's launch, and Sabotage sounds like an appropriate ability for a rogue, but does not have any actual basis in game lore.


 * With each new addition to the game introducing more cards and characters, the developers expect to be "exploring the WoW universe" for "a long time". Many characters which have not yet featured in Hearthstone are likely to be added in future content, and the developers have even stated that existing heroes such as Jaina may be added later as minions.


 * As the game continues to expand, a larger portion of new cards are being created based on new ideas, rather than existing lore. This can be considered an inevitability due to the finite resources of World of Warcraft, but it also represents Hearthstone beginning to create its own lore, distinct from that of the general Warcraft universe. This includes spells and abilities as well as minions such as Annoy-o-Tron. The game may see more original content in the future, although the developers prefer to focus on existing lore.


 * Hearthstone also produces new lore in the form of card mechanics and concepts. For example, in Hearthstone ogres are 'forgetful', with a chance to attack the wrong enemy, while gnomes have a propensity for transforming minions into chickens. While these behaviours are based on pre-existing ideas, their implementation in Hearthstone brings them to life in a new way, not necessarily seen in the franchise's other games.

Minion and deck types

 * The designers like to add new cards which allow players to succeed at playing the decks they love, for example adding new Dragon minions in order to make Dragon decks viable. They also like to reinforce specific "fantasies" such as race decks. However, there is no intention to provide a set number of cards of a specific type with each new content addition. Trends such as minion types may come and go with "peaks and valleys", with no new cards of a given type for some time, before new content brings interesting new additions.

New features

 * The developers try to release new features separate from new content. For example, a patch containing a new expansion or adventure will usually not contain new game modes or features. The developers instead aim to introduce new features in the patches between content releases. This is intended to reduce the amount that can go wrong, as well as obviously allowing the team to focus on launching one project at a time.

Expansions and adventures

 * By the time an expansion is released, Team 5 are "already working a lot on future expansions", including "not only the next [expansion] but the one after that as well". According to Senior Producer Yong Woo, the developers are always "trying to get ahead of the curve":


 * ''"Often different parts of the team are working on different stages of an expansion. While card FX and engineering implementation wraps up, design team is often moving ahead and concepting out the next set of content. We have some cool tools that allow our designers to play with different card ideas and missions before we decide to go all the way with implementation."


 * For example, immediately following the release of an expansion, final design are already working on polishing the next expansion, while initial design are working on the one after that, and others are concepting the one after that.


 * Ideas for new content releases are often gained through brainstorming sessions, with the developers "yelling out ideas" which get written up on a white board.

Systems design

 * The developers are "hyper protective" of "options bloat", both in terms of the user interface and overall complexity. This is often the reply when asked why additional configuration options are not already in the game, the developers preferring to keep the game's options and menus as simple as possible.

Testing

 * Unlike World of Warcraft, Hearthstone does not use a Public Test Realm (PTR) or other closed server to allow for testing of new content with a limited pool of regular players. Rather, the developers have chosen to focus on internal testing - both among the development team and among the wider pool of other Blizzard staff - to explore and refine their ideas. One reason for this is that a PTR would require making each new set of cards available to players for at least a month prior to their actual release.
 * Prior to Hearthstone 's initial release, the developers did not have a large team of balance designers to thoroughly test cards before adding them. This added value to the game's relatively large pool of beta players, serving to stress-test the balance of the game's cards and classes. With the expansion of the team into 2014, the developers are able to fairly thoroughly test cards without the need to release them to players outside the company itself.
 * The developers spend "several months" testing the cards in each new card set.
 * For large user interface changes such as the introduction of Deck Recipes or game formats, the developers test the new designs with users new to the game. The tester is left alone in a room to attempt to use the interface or play the game, while a camera records their experiences. The designers progressively iterate upon the successes and failures of each design, until they arrive at a version that performs well with the testers.
 * While initial testing is done internally, the designers do observe trends in the player population to inform balance choices as well as ideas for upcoming designs. The developers have access to full records and statistics for every deck and every match played, including which cards were played in each match, and can check overall records for individual players as well as individual cards.
 * The developers hold ongoing testing to improve the tutorial as part of the new player experience. During the testing hundreds of new players are filmed trying the game; the developers then watch the sessions and ask the testers about their experiences. This type of feedback has been used to "tweak" the tutorial since the game's alpha.

Initial development

 * Polygon.com - The three lives of Blizzard Entertainment (2014-10-03)
 * Kotaku.com - Pens, Paper And Envelopes: The Making Of Hearthstone (2014-05-04)
 * HearthPwn.com: Fireside Chat - BlizzCon Panel (2013-11-08)
 * MMO-Champion: Hearthstone - WoW TCG Game Announced (2013-03-22)


 * Videos
 * Hearthstone: 10 Bits of Design Wisdom with Eric Dodds - GDC Vault (2014-03-21)
 * The Art of Hearthstone: Playing the Cards You're Dealt with Ben Thompson - GDC Vault (2014-03-20)
 * BlizzCon 2013 Hearthstone panel with Eric Dodds and Ben Brode - YouTube (2013-11-10)

Card balance and design

 * Designer Insights with Ben Brode: Basic cards and power creep (2015-12-02)
 * Hearthstone: The Grand Tournament: Special Event (Video/Podcast: Hearthstone production director Jason Chayes talks about the game and offer a behind-the-scenes look into the development of the latest expansion, The Grand Tournament. Duration: 23 mins. 2015-09-25)
 * Ben Brode & Kripp On Bad Cards (2015-06-14)
 * How a Hearthstone: Heroes of Warcraft Card is Made (2015-04-30)
 * Hearthstone's Card Balance Philosophy (2014-01-16)

AI and gameplay engineering

 * Building the AI for Hearthstone - GAMASUTRA (2014-08-25)

Art

 * #73 - The Angry Chicken: "ARTstone" [Part 1 w/ Ben Thompson] (2015-03-24)
 * Hearthside Chat - Golden Cards with Jon Briggs (2015-03-05)
 * Hearthside Chat - Art with Ben Thompson (2014-02-25)

Crafting and card trading

 * Hearthstone Crafting: In Dust We Trust (2013-06-20)

Music

 * Destructoid: Exclusive tunes along with commentary from the sound team (2013-08-20)