Third Helix is now Kickbomb Entertainment! Please visit kickbomb.us for all future updates.


13 Mar 2010

In this session, Blizzard Executive VP of Design Rob Pardo discussed Blizzard's high-level game design philosophies, values which inform every aspect of the development of their games, even across departments. This was unfortunately not the enlightening, inspiring talk I was hoping for.

Gameplay first

Pardo noted that some companies tend to focus on their tech and lose track of their gameplay, and insisted that tech and art need to serve the game design, not the other way around. He did caveat that statement with the strong admonition that "game design is *not *more important than other disciplines." He also warned against falling into the trap where "designers have the fun, instead of the players have the fun", referring to some designers' tendency to over-design systems because it's fun to geek out over a spreadsheet.

While this is certainly a valid principle at the high level, it felt so self-evident that its inclusion in a GDC talk seemed redundant. Also, I thought Pardo was somewhat unfocused on this point: he never addressed the apparent conflict between "tech and art must serve the design" and "design is not more important than other disciplines", two positions which seem to be at odds with one another.

Easy to learn, difficult to master

Pardo noted that "easy to learn, difficult to master" is almost too "textbook". Blizzard tweaked the concept to "easy to learn, almost impossible to master" instead. He defended *World of Warcraft *as "a pretty hardcore game" that "just happens to also be a much more accessible game than some of the other games" (presumably referring to other MMOs). He cited battlegrounds, arenas, and raids as evidence of "an extraordinary amount of depth".

Again, this is a principle that felt too self-evident to warrant inclusion in a GDC talk, and I can't say I agreed with his assessment of *World of Warcraft *as "a pretty hardcore game". Within the Blizzard pantheon I think *Starcraft *would've been a much better example: *World of Warcraft *is, in my estimation, a very casual interpretation of historically-hardcore RPG mechanics, and I would argue that the endgame content Pardo cites as evidence of "an extraordinary amount of depth" is not so much deep as it is broad. There's more gear, more abilities, and more players, but it's largely still the same simple mechanics, just presented in a variety of different contexts.

Depth first, accessibility second

This talk wasn't the first time I'd heard about this particular Blizzard philosophy, and here I can enthusiastically agree with Pardo on his strategy of building depth into your game first, and worrying about accessibility later. In this modern age where "accessibility" is almost as big of a buzzword as "DLC", this seems a counterintuitive and even controversial position, but Pardo's argument is simple and powerful: depth is harder to design, it takes longer to get right, and thus it's a much greater production risk. And without depth, your game will never be anything more than a flash in the pan.

What is the fantasy?

Here, Pardo talked about delivering on player expectations. He indicated that players will generally take the path of least resistance, and that if the "right" way to play is otherwise, you have failed to deliver on the fantasy.

By way of example, he referenced Starcraft's implementation of hero units. The "fantasy" of heroes is that they are among the most powerful units on the field, that they lead fierce armies into glorious battle, and that they can turn the tide almost single-handedly. In that game, however, the reality was that heroes were fragile, got lost in a sea of non-heroic units, and their death resulted in mission failure. As a result, players tended to hide heroes in the back of their base and use them as infrequently as possible to keep them from harm: not exactly "heroic".

For Warcraft 3, Pardo explained, Blizzard learned from that mistake. In that game, heroes were much more powerful, the game was balanced at a high level to tend toward smaller engagements in which heroes could stand out and play a more critical role, and heroes were able to be resurrected if killed. Pardo claimed that here, heroes successfully delivered on the fantasy, and I'm inclined to agree.

Make everything overpowered

This was a curious value. Pardo argued that every class, every unit, every ability should feel "unstoppable", but that behind the scenes everything should still be appropriately balanced. Pardo cast this as a tuning technique, but I think it comes down more on the presentation side of things. He made the strange assertion that "it doesn't cost anything extra to make something epic", and the only example he offered up was scaling up a dragon model in *World of Warcraft *to something like ten times its original size. This was a vague point that doesn't seem like it holds much water in actual practice, in part because players' and developers' expectations of "epic" vary so greatly with context.

Concentrated coolness

Pardo urged designers to "make each feature the coolest, most concentrated expression of gameplay". He talked about distilling Warcraft 3's heroes -- three per race, times four races -- down to just a few classes for World of Warcraft. This was because, he said, "There's a limit to the amount of complexity that a player can process." He indicated that the ability to instantly recognize the capabilities of a foe on the battlefield, for example, compels this kind of concise focus.

Play, don't tell

Pardo argued that players should play as much of the story as possible, rather than designers telling the story via quest text, NPC monologues, cutscenes, etc. "If you look at World of Warcraft," he said, "'it was not an accident that we only gave our quest designers 512 characters to do quest text with."

He cited the "Green Hills of Stranglethorn" quest as an example of a failure of this principle. In that quest, players must collect 15 pages -- from random monster drops, in standard *World of Warcraft *fashion -- each of which expands to a chapter of a book containing a fairly deep story. But the act of gathering the pages itself had nothing to do with that story, so in practice, players simply purchased all 15 pages on the Auction House and turned them in for their quest reward without reading anything.

By contrast, he cited the Death Knight's introductory quest line in *Wrath of the Lich King *as an example of a success of this principle. The quests in that line tasked players with doing things that directly affected their character and the world around them, and where their actions were tied closely to the lore of the Death Knight class and the continent of Northrend.

Pardo said that his goal is for players to have a basic understanding of the story even in the event that they don't read a single word of quest text. In his opinion, quest text is there to enrich the story, not to lay its foundation.

Make it a bonus

It was for this principle that Pardo gave his most awkward example. In the *World of Warcraft *beta, players would incur a 50% penalty to rate of XP gain after having played beyond a certain number of consecutive hours. Players complained, loudly and rightly. Pardo proudly described inverting the system such that players instead incurred a 100% bonus to XP gain, for a limited time, after having been offline for a number of consecutive hours (e.g. the "rest" system the game uses today), and that the effect of this inversion, despite the macro mathematics coming out identical, was that players now loved the system.

What made this example awkward was that a 50% penalty to rate of XP gain even made it past the drawing board in the first place. That seems like such a self-evidently bad idea that I can't imagine how it ever got implemented, much less made it through internal playtesting all the way into beta. I don't even think this is a case of "hindsight bias", either: the original approach was, simply and objectively, punishing players for playing the game.

Control is king

Pardo argued that "responsiveness trumps cool animation". He cited the process of "mounting up" in World or Warcraft: at one point, Blizzard tried having the player's mount run in from beyond the horizon when summoned, but the delay between player interaction and result was too great, thus leading them to the current implementation of magically summoning the mount into existence directly beneath the character.

He also noted the importance of what I might call "micro-time" response: in this case it was a three-frame lag in the mouse cursor motion during Warcraft 3's development that, once removed, made the game feel exponentially better.

Done well, Pardo said, control responsiveness becomes the key differentiating factor for player skill. He pegged the average *Starcraft *player's actions-per-minute at about 100, while the pros are at 400+. He suggested that that kind of range absolutely requires precise, responsive control.

Tune it up

Pardo spoke briefly about the importance of tuning, but didn't dig into the topic in any particular detail. He did argue that developers should plan and integrate tuning hooks into their games from the very beginning. Every variable should be exposed to design as early as possible. He also noted that designers should know who they're tuning for, and why. Unfocused tuning, he said, just leads to an ambiguous mess.

Avoid the grand reveal

Though simple, this was one of Pardo's better points. He warned designers -- and indeed, all creative types -- to suppress their desire to hang on to their creations until they're "perfect" before putting them in front of the team. He encouraged designers to instead embrace a mindset of, "I know there's something wrong with what I'm showing you: can you help me make it better?"

Culture of polish

This is certainly the most obvious of Blizzard's core values, with polish being perhaps the most famous among the studio's virtues. Pardo argued that polish should happen throughout the entire development cycle, not just at the end as many other studios have done.

Then he closed the talk by saying, "Don't ship it until it's ready." Which is great, but the ability to actually take (and hold) that position is somewhat unique to studios like Blizzard. That doesn't really help the rest of the industry even a little bit, and it makes me wonder about Pardo's -- and indeed, Blizzard's -- perspective on the industry beyond their own walls.

Posted In:

game-design gdc video-games