Vaulting ambition: Fallout 3 and the making of an RPG classic
From death threats to Deathclaws, how Bethesda made an open-world classic
One of Joel Burgess's proudest moments as Fallout 3's lead level designer was giving a big "F*ck You" to players. Literally.
It started when the team realized, only six months or so before release, that the map felt too small. They had used The Elder Scrolls 4: Oblivion as a template for how big the map should be and how often players should discover points of interest – but where Oblivion had forests and mountains to hide far-off towns, Fallout 3's flatter world meant that you could see landmarks on the horizon, which made them feel closer.
"The tone of Fallout doesn't work if it's not a bit lonely," Burgess tells me. "The rate of discovery was a bit high, and we wanted to thin that out."
It meant the team had to add around 20% to the map in its north-west corner. Burgess describes it as "one of our last big scrambles", and it mainly involved shuffling existing locations. But it also added fresh structures in these newly-created regions, including explorable satellite towers hiding enemies and loot.
For one tower, the team plonked down its exterior shell – but forgot about the interior. The flaw was only discovered "a few weeks" before development was due to finish. "When you opened the door it just went to nothing," Burgess says.
The team had to pull its finger out. The middle one, specifically.
Burgess: "My claim to fame here is that I said: 'behind that door put a slab of concrete, grab one of a graffiti sprays that says F*ck You, and stick it on'. So this one door out on the edge of the world just says 'F*ck You'. It was a big planning lapse turned into a good joke."
Get daily insight, inspiration and deals in your inbox
Sign up for breaking news, reviews, opinion, top tech deals, and more.
Making the best of any given situation and giving individual developers the freedom to build out their own ideas are two of the themes the come up consistently during my conversations with Fallout 3's creators.
The troubled launch of this year's Fallout 76 is a good a time as any to remember what made Bethesda's first Fallout entry tick. Over the past month I've spoken to six of its developers to find out how Bethesda transformed the series from an isometric RPG to a first-person blockbuster.
The development team told me about death threats and "creepy visits" from angry fans, how the game's first guns were built from The Elder Scrolls 4: Oblivion's bow and arrow, and how the now-iconic V.A.T.S. system began as a nod to earlier Fallout games.
10 years after Fallout 3 was released, here's the story of how it was made.
- Fallout 76 review: our first steps in the wasteland
- Fallout 76 refund policy could land Bethesda in hot water
Redefining the Fallout series
Writer and designer Erik J. Caponi says he felt a “weight of responsibility” when working on Fallout 3. The team wanted to live up to the legacy of the series while acknowledging it would be many players’ first Fallout game.
“Fallout 3 was going to be what Fallout was to them," says Caponi, "and I wanted to make sure it was the same Fallout at its core that I knew and loved.”
The team was divided on which of the previous Fallout games to take most cues from – Burgess says that, on balance, the first Fallout felt closer to what they wanted to create, so they made sure they were clued up on its source material, including Six Strings Samurai and A Boy and His Dog.
But they also wanted to make something that was decidedly theirs. They began to find inspiration in other sources, such as Cormac McCarthy’s novel The Road, Margaret Atwood's The Handmaid’s Tale, and Pat Frank's Alas, Babylon, as well as art from the 1950s – an era they wanted to evoke.
It was the perfect chance to experiment: the Fallout IP was bought cheaply by Bethesda owner Zenimax in 2004, and Bethesda never felt under pressure to make it a financial success.
“Seven years had lapsed between Fallout 2 and Fallout 3’s inception in 2005, the sensation was that not many people had heard of it, and [the previous games] didn’t sell that well,” says Burgess, who describes the IP as a “genius grant”.
Bethesda was so determined to do its own thing that it, or at least Zenimax, turned down “very generous offers” from other groups that were keen to get involved in Fallout 3, Burgess says. “The sense was we had to make our own game.”
He knew taking the series in a fresh direction wouldn’t please all fans – but he wasn’t prepared for quite how aggressive the backlash would be. “We were literally getting death threats the minute anybody knew we had the license,” Burgess tells me.
There were apparently a few “creepy visits” to the office from aggrieved fans – though never any violence. This is backed up by others: Caponi calls the threats an “appalling window into the way some people think and behave”.
More reasonable sceptics existed within the team – and Caponi was one of them. “How do you take something so traditionally turn-based, a previous generation RPG, change it so much and have it still be Fallout?”
His worries turned out to be unfounded – but it was good to have sceptics within Bethesda because it meant senior staff had to “prove it to the team before they proved it to the audience”.
Proving how good Fallout 3 could be didn’t take long, and the ambition of the project leads quickly gripped the team. “I remember talking to Todd Howard very early about combat,” says lead producer Gavin Carter. “I was asking: ‘do you think you want to try turn-based, real-time, or what?’ And he said: ‘I'd rather do something no one has seen before.’ So that's illustrative of the kind of tone we were trying to set on everything.”
For combat, that involved building of one of Fallout 3’s most recognised features: V.A.T.S., or Vault-Tec Assisted Targeting System. This allowed players to freeze time, queue up shots on enemies – targeting specific body parts – and then unleash those shots all at once, severing limbs and heads.
“One day Steve Meister, Fallout 3's lead programmer, stopped by my desk and said: ‘I just figured out how to get turn-based combat into the game’,” says programmer Brendan Anthony. “His vision was to blend real-time and turn-based together, so that players could use either mode. In theory we could please everybody.”
Todd Howard was particularly keen that V.A.T.S. was “cinematic”, Anthony says. It was Jean Simonet, an AI programmer, who came up with the idea for a slow-motion camera that would change depending on the situation.
“It would conditionalize on the weapon, the kind of enemy, the environment,” Simonet tells me.“Basically, a cool way to display what happens when you shoot people.”
Aside from being a smart innovation on real-time gameplay, it was emblematic of the way individual team members could get their voices heard. Everyone I speak to, particularly the junior staff, say they were surprised at how much freedom they were given.
They also talk about how ‘flat’ the leadership structure was, and how small the team was. Simonet recalls there were around 55 people working on the game at the start, and around 75 when it finished – small enough for a young programmer like him to convince the higher-ups to let him completely rewrite Bethesda’s pathfinding system, for example.
Cultural resistance within Bethesda
Before V.A.T.S., though, the team had to figure out the shooting systems. Their starting point, as it was with so much of the project, was Oblivion: specifically, its bow and arrow.
“The first gun in Fallout 3 was a modified bow that shot very fast bullet arrows,” says Anthony. “After that, it was a matter of slowly morphing the code to act more like guns – hitscan bullets, different gun types like lasers, etc.”
Anthony says he was “very happy with the projectile system”, which allowed for multiple ammo types and could show bullets in slow motion for V.A.T.S. “The bullets were instantly-resolved hitscan projectiles, but in slow motion you want to be able to see them flying. But you don't want them to use a different physics model, for balance and consistency. So it was a more complex system than many FPSs might have used.”
While the projectile system worked well, Burgess says that “a lot of stuff about the shooting is not very good” – he calls Far Cry 2, which released a week earlier, a “far superior shooter”. Indeed, Fallout 3’s guns felt floaty, and bullets lacked impact outside of the cinematic V.A.T.S. system.
“We didn’t really have [FPS] experts, we didn’t really know,” Burgess says. “If nothing else, it speaks to some of the ways we were successful that the mediocrity of the shooting didn’t matter.”
Shooting was by no means the only system that borrowed from Oblivion, either. Bethesda’s games are often accused of being same – but some of that is by design. How players discover locations, how fast travel works, what’s marked on the map, inventory priority, and corpse looting are supposed to be similar. “When you work on a consistent tool set, you don’t have to reinvent those wheels,” Burgess says.
The story of Fallout 3, however, was something entirely new.
“Like most things in Fallout 3, it had an official ownership and then an unofficial group ownership,” says Caponi. That official owner was Emil Pagliarulo, who had already written most of the main quest by the time the team came together. It was expanded and “quest-ified” by another group of designers, and the rest of the design team could make suggestions about improving it. Pagliarulo came back in at the end for an “authority pass”.
The same was true for side quests. When development started in earnest, Pagliarulo and other senior staff had written summaries for more than 20 specific side quests, and that number grew with suggestions from the team. Through clear editing, and constant collaboration, the team ensured that the game spoke with one voice – and Pagliarulo again had the final say.
“I’ve gone back and looked at stuff in Fallout 3 and said, ‘this is really good’, and then realised I wrote it,” Caponi says. “I mostly chalk that up to good feedback and good editing.
“Usually if I’m reading something I wrote, I can tell, and so if I don’t immediately recognise it, it means I got out of my own voice enough to actually produce something that fit.”
Caponi’s role was to design the small tales players could find all over the world that weren’t tied to a specific quests. For example, he wrote the sequence that lets players cure Leo Stahl in Megaton of his chem addiction.
“My goal was to be the carrot that led people out into the wilderness,” Caponi says. “To make sure that when you picked up a rock there was something under it – that you were rewarded for your exploration, and wanted to pick up the next rock.”
Bethesda had not employed anybody to directly tackle non-quest stories before, and it was a sign that the developer was recognising that, for some players, exploring the open world was more important than the main quest, or even the side quests.
It shows in the final game. In Oblivion, you could find mini stories tucked away in the world, like finding a troll with a suicide note on its body. But in Fallout 3, it felt like I could walk in any direction and find something worth stumbling across.
Burgess also advocated for these types of stories, as he had done in Oblivion, where most of the dungeons were cut and paste. On that game, Burgess managed to carve out time (initially without telling senior staff) to build a dozen or so custom ones. He proved himself further when he and another level designer built the Mehrune’s Razor DLC in four weeks – off the back of its success, and because of the strength of Oblivion’s custom dungeons, Burgess was told to hire a full level design team for Fallout 3 to make the world feel more alive.
But convincing some parts of the team that designing bespoke dungeons with custom stories was still a “huge fight”, and he came up against lots of “cultural resistance”, Burgess says. “Everyone looked at us as these upstarts that had been given a level of authority that was out of turn. Just making the case that quests weren’t the only important content was difficult.”
He felt like he was “scrounging” for support with Fallout 3’s dungeons, he says – but still wanted to make sure they all had their own story, and hinted at what the world was like before the bombs. “If there were 500 words of backstory, we might give the players 5.”
Fallout 3's big mistake
Fallout 3’s major locations had to be spot-on too, and they didn’t come any bigger than Washington DC, modelled on the real-life city. “There were a number of times where we sent people into DC to gather reference photos,” says lead producer Carter. “I remember one of our artists was taking very detailed, close-up pics of the Jefferson Memorial when a security guard asked him what he was doing. He had to think to himself: “Ok, don’t tell him you're taking pictures for reference for blowing it up.”
Burgess says DC was “really challenging” – and admits the city was “the big mistake I feel I made on Fallout”. He wanted the city to feel much harsher than the Wasteland outside, with tougher mutants and lots of radiation. The power of the game’s engine meant DC couldn’t be an open world, so it was split into neighbourhoods connected by an underground subway system.
“We hit a point where the game is playable, and a lot of the people on the team are giving me negative feedback,” Burgess says. “People are saying: ‘Downtown is really overwhelming, really hard, I can’t trust wayfinding because everything’s disconnected – I can see the Washington monument but I can’t walk to it’.”
On reflection he agreed, but it was “too late” to drastically correct course. “I’d been too stubborn for too long on that particular point,” he says, and while DC captures his original vision of “overwhelming” the player, that vision “was a bit broken”.
If he could go back, he’d properly test whether DC could be open world. Bethesda conducted a “rudimentary test that wasn’t encouraging”, but Burgess says a proper test would’ve let them make a more informed decision.
He also wishes he’d used the team around him more – and predicts the end result would’ve been something like Fallout 4’s Boston, where “everything is a bit harder, a bit more uninviting, a bit more hostile”.
But Carter believes the DC setting “increased the impact of the story” by setting the game around recognisable landmarks, and says the world building is Fallout 3’s biggest strengths. “If you could get a few people interested in going off and doing something cool, you could really just go off and do it without a ton of approvals,” he says. “That kind of culture created a lot of the best content in the game.”
Fallout 3’s creatures were as important as its world – and its artists had plenty of material to draw on thanks to lone concept artist Adam Adamowicz, who passed away in 2012. His early art was detailed enough to evoke a strong sense of the game’s style, but “scribbly” enough to leave room for interpretation, says artist Jonah Lobe, who mainly worked on creature design: “For example, I created the armour for the Super Mutant. There’s lots of different concept art for that, and a lot of its very scribbly. There’s a lot of room for flexibility.”
The freedom he and others had means “you can really feel people’s fingerprints” on the way the world and creatures looks, he says. Speaking to him also highlights just how much thought went into individual creatures, which took a month each to create at the art stage. Lobe crafted the feral ghouls, the yao guai, and the Deathclaw, which he says is the creature he’s most proud of.
But it’s his process in making the feral ghouls that stands out. Lobe didn’t just set out to create a cool, scary-looking monster – he wanted to fill them with personality. “I try to think about what their lives are like when you’re not around,” he says. “The ghouls are not the most vicious predators, and they live in a dangerous world where food is scarce. Above all, they’re lonely.”
To that end, he wanted to make the player “pity” them, giving them large eyes that are “wild and sad and confused”.
“I created my own story in my head, where I pretended it wasn’t the radiation that made them crazy. I imagined that the eyelids of the ghouls had rotted off, and they hadn’t been able to sleep for decades. Who knows what they see out of their orb-like eyes that can never close? I really took pity on them, wandering the wasteland alone, desperate for a rest that would never come.”
The march of Liberty Prime
Translating thoughts like these into something tangible wasn’t always easy: the team tells me they were constantly battling the development tools.
AI programmer Simonet says he spent half his time building systems that would make the game better – such as a dynamic cover system and AI for combat – and the rest of the time fighting the tools. Working with the game’s creation kit required a kind of “arcane knowledge” passed down from long-time Bethesda designers, he says.
“There was frustration across the board at times. I had to spend a long time trying to make the tools better, but they were still pretty broken.” Spending more time putting out fires meant less time on perfecting new systems, which meant some were buggier than he hoped. “When my stuff worked well, you don’t really notice it. You notice when NPCs run into walls. When I saw NPCs warping places when they have to be somewhere, that made me really sad.”
Battling the tools contributed to long hours for Simonet. Most of the developers I speak to, including Simonet, say there was no “crunch culture” during Fallout 3. They tell me the schedule was relatively well-planned, and that the long hours were confined to final. But Simonet tells me that crunch was “self-imposed” by some developers.
“It’s not mandatory, but that’s the perverse part of it – people ended up crunching a lot. Mandatory crunch is easier. It’s imposed on you by the company, it doesn’t take the same mental toll as self-imposed crunch, where you’re always thinking about work. It’s the dangerous time,” he says.
“On Fallout I know people made themselves sick, I made myself sick crunching on Skyrim. You don’t want to be the one letting the team down, or the reason the game doesn’t do well. You want to prove something, or get a shot at a promotion – whatever the reason is. There was definitely a sense of pride in working really hard.”
In the end, Simonet believes the crunch, and battling the tools, was work it. He was the person responsible for programming the giant Liberty Prime robot’s final march at the end of the main storyline – something he says was a “huge nightmare” initially and required months of work to get right.
“It felt like a great thing – I got through the main quest, and got the giant robot to put his hands on those pillars properly. There was a lot of swearing in between, and fudging this way or that way, but the end result was that, most of the time, Liberty Prime worked. When I’d see one Brotherhood of Steel guy taking cover behind Liberty Prime’s foot, I know what’s happening under the hood and I’m like – Yes!”
Simonet says working on Fallout 3 was a “magic time”. The sentiment is echoed by the rest of the developers, who each express a sense of pride at being involved. The game had wonky mechanics, shipped with plenty of bugs and, some players would argue, failed to capture the essence of the earlier Fallout games. But the magic Simonet describes is something I still feel when I boot it up, 10 years later.
- Fallout 5: will we see another single-player Fallout?