Monday, December 31, 2012

Fit Your Game - Internal Consistency & Identity

During the development of Wrath of the Lich King, I was brought on early to help develop the vehicle system when our lead designer became embroiled in a lot of bigger issues. The opportunity to work on such an important system made me ecstatic.

The initial plan was to develop a system for making tanks and airplanes in the new outdoor battlefield known as Wintergrasp. 

In Wintergrasp, vehicles were a limited resource, meaning only a few players could use them while other fought on the ground. 

Being a PvP zone, with vehicle-only objectives, this mixture of players and vehicles worked naturally. Despite the relatively limited amount of time I put into the vehicle mechanics (mostly flavors of cone ram damage and catapult style delayed explosions) - it worked and modeled other PvP games such as Battlefield and Halo. 

Being a conscientious, passionate and utterly reckless game designer, I pushed the programmers really hard to make sure that the vehicle technology was very modular, extremely flexible and could be reused for anything. 

Too Much Rope

As the Wintergrasp technology proved fruitful, curiosity rose among the other designers on the team about the technology. Eager to learn the new systems and give new legs to the game, they began adapting it for other purposes - something I completely encouraged.

If Jeff Kaplan were here, he'd probably say something like, "we took vehicle tech too far, without enough constraints and consistency."  While there's a lot of truth in that, I feel like the problem was even more basic: 

Often we made vehicles without carefully taking into consideration the audiences involved.  

After Lich King shipped, I personally modified almost every vehicle quest in the game, tuning numbers, changing mechanics and adding consistency such as global cooldowns to vehicle abilities. 


Due to this massive variation in not only quality, but difficulty, frequency and complexity, that it meant that each vehicle quest requires a significant learning curve and triggered anxiety from players who didn't know what to expect. This established a baseline negative response which manifested itself in player feedback many times over the course of Lich King.

I don't want to single out any designer on the WoW team by picking on a particular vehicle quest. In fact, if anything, much of this is my own fault for not establishing examples and guidelines for quest designers to start from easily. So I'll pick a specific example that I worked on personally.  

Yes, that's right. I'm the madman behind the source of much hatred and anguish known as the Flame Leviathan. 

Not to take credit from any of the very talented designers who contributed to the planning and execution of the encounter - but many of the issues players had with the experience stemmed from decisions I accepted as immutable early on.  

Riding on the success of Wintergrasp, there was a surge of excitement to see vehicles make their debut in a raid setting.  


Wintergrasp vehicles worked well, as players could destroy and protect them, choose to pilot or not pilot them. Thus a player could easily slot themselves into the role of "vehicle pilot" or "hero".  Likewise, while no single player could destroy a vehicle, no vehicle could kill *all* of the players.  

Role switches at work.
In Ulduar, players faced with handling the Flame Leviathan had no such choice. You're a healer? No. You're a gunner now.  You were a tank? Well, now you're a lowly ammunition loader.  

Sometimes, when you switch things up, they turn out awesome. Other times... awkward. 

When players actively choose to buy-in on that decision, they have a great time - modern encounters such as Amber Shaper Un'sok and Alysrazor show that exotic changes in gameplay can be super fun - for the players that opt into that challenge.

Flame Leviathan had a lot of time, love and complex vehicle design built into it. It was a fun encounter  - but it didn't do enough to reinforce your core identity apart from the vehicle. In fact, I probably worked a little too hard to avoid your class mechanics entirely.

Reinforcing Your Place in the World

Late in the game, to help relieve some of the complaints, I added the "toss you onto the tank" mechanic where players needed to defeat an NPC using their player abilities under time pressure. This helped a bit, but left tanks and healers somewhat out in the cold.

An alternative might have been for one of the vehicle types to have been an SC2 style meditruck, accompanying a smaller fleet of vehicles through the gauntlet - protected by a small squad of normal players on the ground.  

Ultimately, a lot of the complaints stemmed from one real issue: I was trying to put a game that was not World of Warcraft inside of the World of Warcraft.  

If there's a lesson here let it be this: It's perfectly fine to add things like that. It's not perfectly fine to expect everyone in the game to do it and enjoy it. 

Know your audience. Know yourself. 


  1. So now I know who to blame for the vehicle quests and for Flame Leviathan!

    In all seriousness, it makes absolute sense why it would be an exciting idea and Blizzard designers would want to leverage it as something of a defining feature for the expansion. I think you nailed why it did work in Wintergrasp and (by and large) didn't work in PvE generally.

    Friends of mine and I often remarked (especially when doing Oculus) "I want to play my character, not be removed from it." I'm guessing that was a common sentiment.

    I think Blizzard has generally gotten the right idea with vehicles since then...making them more like interactive cut scenes. You may have to point the mouse and press a button, but you don't really have to *learn* anything. They're more interesting plot-moving mechanics rather than crucial game mechanics (with the exception of the raid vehicle instances like Amber-Shaper).

    1. I think that the Oculus would have worked just fine if it had been the first phase of the fight.

      People would have been like "oh, okay, let's beat the dragon minigame" followed by the intense part of the fight being in control of your character.

      Much of the problem with Malygos was that the final, climactic part of the fight was done with foreign tools and required high levels of performance to unlock the "minigame" execution phase.

      Simply reversing the encounter would have let the high-learning curve portion happen first, followed by a high-intensity, high performance phase.

      That said, it *was* cooler to be suddenly dropped onto a dragon's back.

      If it had been a very easy, almost cinematic phase 2, it would have worked better as well.

    2. I agree that being dropped onto a dragon's back was cool and sometimes (in my opinion) it definitely worked, like the defending Wyrmrest Temple daily, where you kill enemy drakes by piloting a friendly one, which was opt-in and pretty short.

      I think you're right that if the dragon-back "minigames" had been front-loaded rather back-loaded, it would have helped. I also think (though this may have been just me and not at all universal) that vehicles just didn't feel intuitive. They moved ponderously and control felt imprecise which is somewhat of the opposite of how WoW generally feels. This is a game that is predicated on quick, precise movement (witness the mockery of people who keyboard turn rather than mouse turn) and vehicles took people away from that. While I understand that the experience should be different, it felt more frustrating-different in that sense than novel-different.

      I guess if I were to boil it down, I didn't feel faster and stronger in Wrath vehicles...I felt more awkward and limited, which is probably not the intended experience to have when sitting inside a purportedly powerful vehicle that does super-damage.