This is a reprint of an article that first appeared in the October 2012 issue of Game Developer Magazine.  It has since also reappeared on Gamasutra at this link.

Ultima Online and EverQuest represent two very different game philosophies. Ultima Online‘s creators tried very hard to create a virtual world with physics and interactions that mimicked the real world, so players could interact with each other in ways meant to model reality: You can chop down trees, dye clothes, build houses, attack almost anyone anywhere, and steal anything that isn’t nailed down.

By comparison, EverQuest is a simple game, not much more than a combat simulator designed to mimic the basics of combat found in tabletop board games and old online Multiuser Dungeons (MUDs). Combat in EverQuest is very deep and intricate compared to that in Ultima Online, with far more ways for players to attack and manipulate their enemies. However, combat aside, EverQuest was perceived to not be a very feature-rich game. Most of the world interactions in Ultima Online aren’t in EverQuest, and when they are, they aren’t particularly deep or fleshed out—to the extent that many observers felt thatEverQuest would be too simple for the newly invented massively multiplayer genre. As it turned out,EverQuest easily beat Ultima Online‘s numbers, and a few years later, a rematch of the two MMO design philosophies paired Star Wars Galaxies against World of Warcraft — with a repeat of the same end result.

As it turned out, Ultima Online has a lot of features, but many of those features don’t have a lot of depth to them; it is broad, rather than deep. EverQuest has fewer features, but a combat model that is very deep (and became deeper as new boss mechanics were added to respond to an increasingly savvy audience).EverQuest is a game about depth.

The Pitfall of Breadth

Most junior designers come into the industry favoring breadth. They want to design the perfect game, and they want to do so by throwing every possible feature under the sun into the design soup. This is especially true in massively multiplayer game design, where the possibilities of what you can do in a game is effectively unbounded — a virtual world can already incorporate almost any feature of the real world. Even worse, most game genre devotees imagine that the perfect game in their genre is one that combines all of the best elements of other games, because they don’t recognize the underlying costs of all of those systems.

Most triple-A games that actually ship (and most experienced designers and producers), however, favor depth over breadth for a few reasons. One is simply a matter of resources — it’s hard enough to even do one game system extremely well. If you’re trying to make a first-person shooter, for example, you are going to have a hard enough time getting the basics of making a deep and engaging FPS that can be on the same playing field as Call of Duty without losing all the resources and focus on building other systems.

This is especially important when you consider how your multiple game systems are supposed to interact with each other. For a game that is about breadth, the multitude of game systems can interact with each other in many often-unexpected ways. In some cases, this can be a good thing — something that designers like to call emergent behavior. Emergent behavior can be wonderful to behold, as the fans will undoubtedly surprise you with their ingenuity. However, the more systems you have, the more time you’ll need to spend on QA and balance, especially if you’re making competitive online games, where that ingenuity can be used to win or to harass other players. The same game freedom that allows players to build amazing things like pianos by stacking items on top of each other in Ultima Online also allows them to build staircases of spoons up to the top of other players’ castles, enabling them to waltz in and rob them blind.

Breadth Can Be Good

This is not to say that all games that are breadth — first are doomed to fail. Ultima Online and Star Wars Galaxies both had sizable fan bases, largely because the two games were wild and unpredictable places where it felt like anything could happen. The idea that you can go anywhere and do anything is an attractive sales pitch, and it is one that two companies in particular have been remarkably effective at building toward — Rockstar (Grand Theft Auto) and Bethesda (creator of Oblivion, among others).

Grand Theft Auto is the poster child of the breadth game, and it illustrates the difference clearly. It has auto racing that fans of Need for Speed would sneer at, car crashes that are snoozers compared to Burnout, a story without the choice or depth of a BioWare game, shooting that pales next to Call of Duty, fighting that has nowhere near the tightness of control of Soul Calibur, and yet it is still widely considered one of the finest games ever made. Why? Because all these things blend into each other — you can crash your car, then get into a gunfight that ends in an epic fistfight. Grand Theft Auto is, at its core, broad and open-ended wish fulfillment of the idea of a world without rules; you can go anywhere and do anything you want. The possibilities are endless, and it captures the imagination like few other games can.

Of course, Grand Theft Auto also shows how expensive that kind of broad world interaction can be; development costs of Grand Theft Auto IV exceeded $100 million, according to Wikipedia. That said, the same source says Take Two earned five times that much in income in the first week the product was on sale. But few companies can afford that initial price tag.

Simplicity is Key

In the context of breadth vs. depth, the concept of simplicity is vastly misunderstood. Wizard 101 and Free Realms are two lighter MMOs designed for a younger audience. Sony Online’s Free Realms includes a series of minigames; today, the web site lists postman, kart driver, and soccer star among more classic combat roles like ninja and warrior. Each of these games has smaller minigames associated with them, with their own respective advancement tracks.

Wizard 101, by contrast, was made by a smaller, scrappier studio (King’s Isle). They didn’t have the resources to make a broad game, and instead focused on making an intricate combat model reminiscent ofPokemon and Final Fantasy. While the world itself is simple, the combat model is not — the game designers went out of their way to design a game with a long life, and with angles of expansion so they could continue to put out content that would be in demand on the microtransaction store. Their tactic appears to have been successful — Wizard 101 got its 20 millionth user in 2011, two years and two months after the game was launched, while Free Realms took four months longer to hit the same milestone.

Plan to Expand

One of the interesting angles of the depth vs. breadth debate, especially in terms of online games with ongoing support and development, is the appetite for expansion. If you have one central game activity, then it is very easy to focus your development on that game activity — adding more creatures to summon inWizard 101, for example, or adding new raid mechanics to World of Warcraft.

By contrast, broad games have a broad player base as well, and all of them want more and more interesting things to do when they consume their favorite kind of gameplay. In Ultima Online, crafters and tamers demanded as much design attention as combatants, and I’m sure the same can be said in Free Realms for fans of the soccer and mail delivery games. It is harder to improve the game on multiple tracks, and keep all fan bases happy while still maintaining the game’s core balance integrity. Adding new features and game systems to increase the breadth of the game is always an option, and is typically popular, but it also risks increasing the complexity of the game — increasing the number of unexpected interactions that need to be considered, balanced for, and tested.

The Broad-Game Bullet List

Making a broad game isn’t impossible, and some of the finest video games ever made are broad games. But here are some things to consider.

1. Don’t think of your features in a vacuum. Think about how they interact with each other, how they balance against each other. Be sure they support each other (at best) and don’t compete or invalidate each other. In Oblivion, the desire to learn by doing to complete a sandbox experience resulted in players hopping through fields picking flowers in order to become master assassins (by increasing their jumping and poisoning skills).

2. Think of the audience. Any feature adds complexity to your game and takes development time. Yes, you could add washing laundry to your GTA clone, but would it hold anyone’s interest for any length of time? Does it support the core fantasy of the game?

3. Accept that your features will be simpler. Complexity in multiple systems is hard to support, hard to test, hard to balance, hard to expand, but perhaps most of all, hard for most players to actually follow and understand. In a broad game, the game systems need to be simple — the complexity will come from how those systems interact.

4. Be willing to lose control. Embracing emergent behavior means you are accepting on some level that players will surprise you. Your design team and management needs to decide up front if this is a good thing or not.

Broad games are hard to do, and really hard to do well. There is a reason that most broad games that get started never ship, and that many established developers flee them, philosophically. But still, I would sorely like to see more broad games on the market. Games like Ultima Online and Grand Theft Auto IV were exciting because they challenged their players to imagine the possibilities. If done well, this promise speaks right to the core of what makes video games amazing.