20 years of Fallout: Lessons learned shipping games in the wasteland

Perhaps war never changes, but game development sure does.

As Bethesda Softworks braces for the launch of a new Fallout game, it’s worth taking a moment to look back at how the high-profile franchise got its start -- with one game developer, in a room at a California game company, coding an engine in his spare time.

“It was just me working on an engine,” recalled Fallout lead Timothy Cain during his GDC 2012 postmortem of the project, which is absolutely worth going back to watch. “I just kind of wanted to make my own engine, and nobody said no. That was just kind of the way Interplay worked in the ‘90s.”

Dig into the history of Fallout, from a developer’s perspective, and you get a sense of how both the series itself and the industry it grew up in have changed since Cain first began working on the game at Interplay in 1994, just over two decades ago.

It didn’t start out as Fallout, of course. During his postmortem, Cain told the story of how the game, known internally as Vault13, came to be branded Fallout at the suggestion of Interplay's then-CEO Brian Fargo after he took a build home to play.

“Fallout had a double entendre of the radiation from the bombs and then the alternative definition, which is a lingering effect or set of consequences,” Fargo, now chief of Wasteland 2 developer InXile, tells Gamasutra. “Perfect for a game that stakes its rep on choice and consequence.“

Looking back, Fargo suggests modern game developers appreciate the boon that crowdfunding and open development can be, from a playtesting and bug-stomping standpoint, compared to the QA departments and services of yore.

“When we worked on Fallout we had a QA department, but that doesn't give you a true indication of how players will react,” says Fargo, who contrasts the Kickstarted/Early Access development of last year's Wasteland 2 as more of a "spectator sport" than game development ever was in his days at Interplay. 

He also advises that developers lay out a clear mission statement and vision for a project early in the production process, as in hindsight it proved a key turning point in Fallout's development.

“I remember us dissecting [spiritual predecessor] Wasteland 1 before V13/Fallout began and breaking the key sensibilities into a vision document -- things like moral dilemmas and providing a diverse pathway for players,” notes Fargo, echoing Cain's comments that the game only came together once the team wrote themselves a mission statement.

You can read an archived copy of that statement here, though it's worth noting it was written while Fallout was still being designed to use the GURPS tabletop game license -- something that changed late in development.

The game would go on to outstrip Interplay’s sales expectations when it launched in the fall of ‘97, though a lot of that has to do with the fact that Interplay doesn’t seem to have had very high expectations to begin with.

Cain noted that the project was “not a typical Interplay game,” because it was built on its own custom engine (rather than say, BioWare’s Infinity engine, which Interplay had the option to use) and didn’t bear a well-known license.

(Incidentally, the question of whether to use an existing engine or roll your own is one many developers still struggle with today, even as Unity and Epic have done their best to make their engines easily available and approachable.)

What’s more, Cain recalls that some folks at Interplay tried to get Fallout cancelled multiple times because they were afraid it would compete directly with the company’s other projects, role-playing games based on the Forgotten Realms and Planescape licenses that had larger teams.

Todd Howard once estimated that roughly 80 people worked on the team that made Fallout 3, and studio follow-up Skyrim boasted a team size of more than 90. By comparison, the original Fallout was developed by a team of one for months -- at its height, the game had a total team size of roughly 30 people, according to Cain, who recalls the game costing “about $3 million” to develop -- nearly $4.5 million in 2015 if you account for inflation.

That's a significant amount of money, and I think it's important to talk about game budgets (then and now) at a time when some developers are undercutting themselves and the industry by asking for too little on Kickstarter, making the cost of game development seem cheaper than it is. 

In the face of that initial outlay, Interplay started production on a new Fallout game using the same tech and assets while the first one was still being finished, and set a strict ship date of holiday '98. 

Project lead Feargus Urquhart, then chief of Interplay's Black Isle Studios and now CEO of Fallout: New Vegas developer (and Black Isle spiritual successor) Obsidian Entertainment, remembers hard lessons learned during that period about pushing yourself and your team to hit a ship date.

"The biggest challenge of Fallout 2 was that we had set its launch date based upon having started it in the middle of 1997. That meant we would have about 18 months to make the game in order to get it out for Christmas of 1998," says Urquhart. "I pushed everyone incredibly hard to get the game done.  It was pretty early in my career, so I'll admit that we (mostly me) pushed too hard to get Fallout 2 done that year...we ended up making most of the game in about 8 months."  

The team hit their ship date, an achievement Urquhart feels is weighed down by how buggy the game was at launch -- a recurring issue in the Fallout franchise.

"One of the biggest, and most visual bugs, was the car trunk bug," says Urquhart, relating a Pratchett-esque tale of a trunk run amok. "We came up with idea about midway through development for the player to have this car that they could store stuff in. We could easily store stuff in containers on a map, but we didn't have a system that would have that same container available on another map. To make the inventory of the container, the car's trunk, persistent across maps, we decided to make it a companion."

"However, it was a special companion that didn't follow you and only showed up on certain maps.  And, when it showed up, it showed up in a specific space and didn't follow you. Unfortunately, we didn't find all the bugs before we shipped, so a disembodied (disem-chassied) trunk would follow the player in areas from time to time."

Beyond being vigilant about bug-squashing and setting reasonable ship dates (or delaying them rather than crunching,) Urquhart recommends that game developers today learn from his mistakes on Fallout 2 by not being afraid to ask for the tools necessary to work efficiently -- even if it means bothering programmers.

"At the start of Fallout 2, I was watching one of our area designers, John Deiley, layout a level," recalls Urquhart. "I asked him why he didn't just copy one of the buildings he'd already created as a starting point for the building right next door He looked up at me and said that the Fallout Editor didn't have Copy and Paste support."

"I just looked at him blankly until he said it again. I left his office and went over to talk to Scott Everts, the designer who created the maps for every single level in Fallout 1. I asked Scott about not having Copy and Paste, which meant every single tile you see in Fallout 1 was uniquely placed. He said it made making the levels hard, but he didn’t want to bother the programmers. Nowadays, I always ask our level designers, and artists, if they have all the tools they need - all the way down to copy and paste."

Urquhart also remembered the royal pain that was the Fallout and Fallout 2 save system, which were inadvertently designed such that it was "practically impossible" to make game saves compatible with patched versions of the games. 

"So every time we updated the game, people would have to start over," recalls Urquhart. "Not something I ever want to repeat again."

Urquhart has also previously lamented the over-abundance of slapstick humor in the game, something longtime collaborator and fellow ex-Interplay developer Chris Avellone remembers as a learning experience.

"I look back on Fallout as probably being one of the most exciting and juvenile times of my career...Fallout 2 taught me a lot of what not to do with design aesthetics, and I don't think I showed the franchise the respect it deserved," Avellone tells Gamasutra. "It included too much profanity (without purpose - I am fine with profanity with purpose), too many movie/book references...and even too many internal developer references, including my developer nemesis, T-Ray, as one of the chop shop NPCs in New Reno."

The game did become something of a target for developers who wanted to rail against in-jokes in game writing, and making humor work in games remains a challenge for many designers.

An early build of the cancelled Van Buren project

Yet Fallout 2 did well enough, critically and commercially, that Interplay greenlit development of a sequel under the project name “Van Buren.” Hard times at the company saw the project cancelled at the end of 2003 and the Fallout intellectual property rights licensed to Bethesda in 2004.

Years later, Bethesda would buy those rights from Interplay outright for nearly $6 million and license the right to make a Fallout MMO game back to Interplay -- a tricky agreement that the two companies legally feuded over until reaching a settlement in 2012.

Before that, Bethesda saw great success finally shipping Fallout 3 in 2008 -- success game director Todd Howard later attributed to a well-oiled studio culture that was flexible enough to shift the course of a game's development in response to team feedback, even if it led to develoment delays.

"You can have the greatest design document ever made, and you're going to change 90 percent of it as soon as you play the game," Howard said onstage at DICE in 2009. "Your plan is not as important as your culture."

Following the success of Fallout 3 Bethesda struck a deal with Obsidian (which counts a number of former Fallout developers among its staff) to develop follow-up Fallout: New Vegas, an opportunity Obsidian cofounder Avellone saw to apply the lessons he'd learned from working on Fallout 2 back at Interplay.

The metro train in Fallout 3 DLC pack Broken Steel was made to run by attaching it to a (very) fast-moving actor

"When going for humor and comedy in game design now, I tend to err on the side of references within the game world and try to make sure it could be explained within the context of the universe," says Avellone. "When working on Fallout: Van Buren and again on Fallout: New Vegas (especially the DLCs) I made a conscious effort not to make any such references, as I felt they damaged the heart of the franchise."

Avellone and the rest of the Obsidian team shipped New Vegas just over five years ago, and since then they've gone on to work on other projects (including the recently-released crowdfunded RPG Pillars of Eternity) as Bethesda closed ranks to work on Fallout 4. 

Some longtime fans of the franchise bemoan how Fallout's philosophy of design has shifted away from focusing on player choice and consequence under the influence of Bethesda's history of open-world game development and the rising popularity of robust in-game crafting systems.

But in talking to developers who worked on various titles in the Fallout franchise, one of the common themes to emerge is how the size and scope of the series has grown in the past twenty years. 

"Game development has changed quite a bit in some ways," reflects Josh Sawyer, who worked at Interplay in the '90s and later served as game director at Obsidian on Fallout: New Vegas. "Team sizes and budgets were relatively small back in the late 90s/early 2000s.  We made Icewind Dale with a team of about 20 developers.  We didn't even have any leads!  For comparison, the Fallout: New Vegas team was three to four times that size over the course of the project."

Sawyer says he prefers to work in smaller teams, and the scope of a modern Fallout game is outside his comfort zone as a developer. He also prefers to work with a diverse team, which is far more common now than it was when Fallout got its start.

"When I started at Interplay in 1999, there were zero women in development," says Sawyer. "The number of female game developers has gradually gone up from year to year.  Fallout: New Vegas had about half a dozen women on the project, which isn't really that much, but it was the most I had ever worked with. The game industry doesn't have a history of being particularly welcoming to women, so I'm glad that is changing, even if it's not happening at a rapid pace."