You're in charge of the second most popular videogames franchise in the western world in recent years. Your studio produces a game so big - both in terms of sales and in terms of cachet - that it's actually become a label in its own right within the world's biggest third-party publisher. In total, it's sold 95 million pieces of software in twenty-two languages since its inception, and come to think of it, it's the only PC gaming property that's actually bigger than Blizzard's Warcraft juggernaut.

Nobody is denying that Rod Humble, the man in charge of EA's Sims studio in sunny Redwood Shores, is in a rather enviable position. The veteran developer, who moved into EA - and the Sims unit - after a long stint working on the EverQuest franchise for Sony Online Entertainment, is at the helm of a franchise whose success runs so deep as to defy description.

No game whose audience is 62 per cent female and only 21 per cent under 17 - with a heavy bias for the over-45s - could comfortably be called "hardcore". Yet equally, can a game for which hundreds of thousands of pieces of user-created content have been uploaded, with four million active online users on the community site every month, really be called "casual"?

The Sims, then, is a phenomenon - one which defies conventional analysis as a videogame. Which means that for all that Humble's position is enviable, it's also challenging - because when it came to creating the next instalment in the series, he and the team at Redwood Shores were facing questions about game design that no other game had ever posed.

Clean Slate

Two different approaches were taken to feed into the final solution. On one hand, the Sims business unit - under the watchful eye of label boss Nancy Smith - set about discovering everything it could about its user base. Who are they? Why do they play The Sims? What do they get out of it? What frustrates them? What excites them?

All of this information was fed back to Humble and his team - who used it as fodder for an enormously extensive brainstorming session. Over the course of months, the design team would tear apart the very essence of The Sims and put it back together again. Along the way, they would discover what makes the game tick, where the elusive element of "fun" lies - and how it can be improved.

It all started, as these things do, with a single meeting - the very first design meeting for The Sims 3. It was, according to Humble, "a very quick meeting, actually."

"The one thing that I walked into the room with as a mandate was 'No More Hamster Cage'," he says rather cryptically. "That came from a phrase Matt [Brown, design lead] used - during The Sims 2, they referred to the Sims as hamsters with jobs. There's a little hamster cage, and they have a little job..."

"I always thought that was incredibly limiting," Humble continues. "We needed an open world. You're not going to get where this game can go unless you can literally walk across the street and see the neighbour's kids playing in the opposite yard. You're not going to get to the next place you want to be unless you can buy a house at the top of a mountain and look down on a town. That's incredibly important, and that was job one."

So, at least one aspect of The Sims 3's design was firm from the very first meeting. As for everything else... Well, not so much. "The next step was to ask what was wrong with the Sims themselves, and how that could be improved," Humble says. "That was the initial thought - and then, frankly, our thoughts just wandered. We deliberately went blue-sky."

Over the course of an hour or so, Humble and Sims 3 lead designer Matt Brown show off some of the weird and wonderful prototypes and proof-of-concept pieces that resulted from that blue sky thinking. With an extremely broad brief to explore the principles that underpin The Sims - everything from the sociological concepts of simulating a town through to the engineering challenges of how the player interacts with game characters - the team allowed their imaginations to run free.

Under Pressure

In the bizarre menagerie of prototypes, several different types stand out. There are prototypes which seek solutions to specific problems in the world of The Sims - like a screen full of blobs which represent groups pedestrians walking along a pavement, and changing formation to continue their conversations while passing other groups walking the other direction. Another similar experiment sought to create realistic conversation dynamics, exploring the way that groups form and dissipate at a party - and even how some people have different "talking distances", which can make others uncomfortable.

Another group of prototypes is rather more abstract - and in some cases, it can be hard to see how they relate to the game at all. Humble explains that for some time, the designers worked on "one-day prototypes" - coming up with a concept to explore on Friday afternoon, thinking about it over the weekend, and then building a demonstration to explore their thinking on Monday.

"Every couple of weeks we'd do this, almost like an improv warm-up," explains Brown. "It was designed to get people thinking about the different topics that are relevant to The Sims."

Some of the resulting prototypes are simply weird, like a procedural tree that grows new shoots and branches as you watch - but allows you to prune various branches, which "stains" further branches down that offshoot with the colour of your shears. "I'm not really sure how it applies to the game," Brown admits - but that applies to many of the applications developed in the "one-day" brainstorming. The idea wasn't always to be directly relevant; many of the weirdest prototypes, Humble insists, provoked really good discussions about fun interactions or interesting systems, and how to apply new concepts to the game.

Another prototype that resulted from taking a rather askew glance at game design was one which Humble describes as being like Pandora's box - "taking all of the evils in the world, putting them in a box and telling everybody not to open it." With the team already set upon removing large swathes of micro-management from the game, what better way to crystallise the design than to distil that very micro-management into a prototype of its own?

This was, the designers say, tedium distilled. A simple 2D representation of a house and stick-figure Sim, it called on you to click on the various rooms in order to keep your Sim's associated meters in the green, dividing up your time as best you could to keep his mood high before he went off to work. Later stages added objects that reduced the need to spend quite so much time in each room, new rooms for learning skills, and even passers-by to befriend - but remained, essentially, a frantic exercise in clicking furiously to try and keep all the plates spinning.

Perversely, it turned out to be great fun - so much so that the team spent some time building a multiplayer version. Despite this, the prototype reinforced their decisions on the overall design. "We realised from distilling it down that this sort of play is really frenetic," says Humble, "and that was explicitly what we were trying to get out of. We didn't want you to feel this crazy kind of stress when you were playing. We wanted people to feel that they were growing and nurturing their characters, rather than constantly meeting this kind of pressure."

Genesis Zoo

Such distillations of the Sims' core gameplay concepts provided the team with superb insight into the real appeal of the game - and allowed them to make some slightly surprising design decisions for The Sims 3.

"One of the [things we learned] that was very powerful was that it felt right to be able to go into the world and tidy up for your Sims," says Humble. "This was a huge deal. Before the prototypes, I don't think that anybody would have agreed that that was a good idea - but after doing a prototype, we realised that when your Sim is at work, and there's a book lying on the floor, it bugs players. They just want to move it. That felt right."

"We would never have put it in the game because it sounded tedious and awful. Imagine it as a feature - now, in Sims 3, you get to clean up your Sims' trash yourself! But what it means is that if your Sim is sloppy and leaves stuff lying around, you can fix it. People did it reflexively. We didn't even mean for it to be in the prototype, we just forgot to turn it off - and then we watched people do it. If you don't want to, it doesn't matter, but there's a weird side-band nurturing motive that makes people want to do that."

Yet another category of prototype experiments was designed not just to pick apart The Sims' core concepts - but also to experiment, in an engineering-led way, with some of the sociological ideas that informed the game design. Some were simple, like a program that created procedural "paintings" using a random set of parameters - but could create similar-looking paintings consistently, as a way for a Sim to express their artistic side. Others were more complex.

One of the team's favourite prototypes certainly falls into that category. Designed to explore ideas about socialisation, it created a sea of 2D faces with distinctive features - and showed you both your own face in the game, and that of your nemesis. By singling out other people with similar features to your nemesis, but different features to yourself, and declaring them to be a witch, you could gradually influence the crowd into declaring your nemesis a witch and burning him - complete with mandatory Monty Python sound effects.

Silly? Yes, but also an intriguing exploration of communal social knowledge. "It led us to have interesting and meaningful conversations about things like your town's social trends," explains Brown. "There might be one person in your town that everyone hates, or one person that everyone likes - we looked into how to find value in those kinds of macro trends, and in whether players pick up on them or not."

Even emotional response was explored with simple prototypes. "Bring out the one I hate," Humble says with a groan, prompting grins from the other designers. This prototype explored how to make players feel genuinely bad for a Sim, and how to create emotional displays in the game. With a simple stick figure, Brown shows how it can develop an affinity for a ball if you play with it - or can develop a fear of the same ball if you throw it repeatedly at the Sim's head. Moving the ball in and out of the scene then changes the Sim's emotions; which can be mixed up by moving around other environments, even drowning the poor chap by depriving him of oxygen underwater.

Your World In 2D

The studio's focus on prototyping didn't stop once The Sims 3 got off the design drawing board, however. All of those simple "exploration" prototypes pale by comparison to the complexity of later prototypes - which the team used to implement and try out the entire design of the game before committing to developing it in a complex, expensive, 3D form.

In fact, by the time the 3D engine was created, the studio already had The Sims 3 up and running in its entirety - as a fully playable 2D game. Every piece of the design was implemented in this prototype, from character personalities to object interactions, careers and so on. It may only have been made of labelled squares and circles, but the 2D version includes the entire town - and was so advanced that the team even used it to try out some user interface ideas.

The advantages of building this prototype rather than going straight to the 3D version are obvious; by the time the world was being implemented in 3D, the team already knew that they had a working game on their hands. Moreover, they knew that they had something fun - and The Sims 3 was already capable of surprising its creators with the emergent behaviour that came out of its complex, trait-based character creation system. Some of those traits rapidly turned out to be both anti-social - and hugely entertaining.

Humble, who played the 2D version of the game for some time, spent much of his time trying to collect every book possible for his Sim's bookshelves - "I'm easily entertained," he says with a grin, although Brown rather sheepishly interjects that the team had never actually told Humble that books didn't do anything useful in that version of the prototype.

Brown picks up the story. "Suddenly, one day, from Rod's office we hear a more obscene version of 'where the hell are all my books!' - and someone else asks him if he's met Bob in the game. Well, it turns out that Bob steals stuff... And in his living room, he's got all of Rod's books. Getting to know individuals in the community has turned into a really fun part of the game - finding out that your neighbour is really flirty, but she's 80 years old, or that the two kids across the street are total opposites and hate each other's guts. It's great fun to explore the characters around you."

Those characters are kept consistent by another system we see in a more complex simulation. Here, an entire neighbourhood is represented by 60 boxes on screen, each a house, each each holding a Sim - or a household of Sims. In this high-level simulation, rather than running around going to the toilet or making spaghetti, they do the big life things - meeting people, falling in love, getting married, moving to new homes, getting jobs, getting promoted, having kids, and even dying.

As we watch, the whole town moves around in a fluid way. People get married and have children, the children grow up, the parents die. Icons indicate objects and furniture people buy; dollar amounts show how wealthy they are. They start dating, move in together, and new families move into the vacant lots. It's all consistent, and it runs all the time. In the game, of course, you see only glimpses of this; most of it happens in houses with their roof firmly on. In the prototype, however, you can see the life stories of every Sim, and understand how the team has ensured that stories in the game match together - helping to make the whole town feel real.

That, in the end, is the whole objective of the exercise. Described variously as a game, a creative tool and simply an escapist outlet, The Sims 3 is - from an engineering and design standpoint - an attempt to create a community of people that's believable, consistent, and realistic - without ever sacrificing the principles of being fun and entertaining.

The blue-sky approach used by Humble and his team may seem decadent to other developers working to shorter schedules and tighter budgets, but the principles at work here remain fascinating. It's not just about exploring concepts for the design, after all; the studio's prototypes also allowed mistakes to be made in simple, cheap 2D models, rather than in an expensive, artwork-filled 3D world. So while some aspects of the process may sound more pie-in-the-sky than blue-sky - in some regards, taking lessons from how The Sims 3 has been built could be a decision that's as sound financially as it is creatively.