Modern game design has been changed in many ways by Portal’s release in 2007. Not only have first-person puzzles games such as Antichamber and The Talos Principle become a genre of their own, the game has influenced many others with its physics-based gameplay. It’s safe to say that these days, we’re thinking with portals.
However, Portal wasn’t the first game to feature its namesake as a problem-solving mechanic. A much bigger game came out a year earlier that made use of similar features. Prey was released in 2006 to a rather positive reception and yet, the game barely seems to have made an impact on the industry as whole. There are a number of reasons why this might be the case, including the famously protracted development that surrounded the game. Prey had been trying to implement portals into gaming since it was first announced in 1995. During that 11 year development cycle, the team behind Prey faced numerous challenges in bringing the game to market, some of which we will begin to uncover here.
Under development at 3D Realms (of Duke Nukem fame), Prey was envisioned to be a showcase for new, cutting edge technology. While the engine was being developed solely for Prey itself, there was a hope that the engine could be used internally for other games in future, similar to how id Software and Epic operated with Doom/Quake and Unreal respectively. This meant that the engine behind Prey would have to be a showpiece that would set 3D Realms apart in the market, essentially selling their future games based on that technology.
Paul Schuytema worked at 3D Realms around the time that the game entered early development, and was bought on to lead the project. Speaking to us, Paul remembers some of the issues that troubled development.
“The ubiquity of 3D tech really messed us up” recalls Paul. “When we were developing Prey, it was one of the first games that would require a 3D accelerated graphics card. That was still pretty new tech back then, and that was a huge risk Scott Miller and George Broussard were taking (which turned out to be almost prophetic, since every computer now has 3D accelerated graphics), but they wanted the visual experience to be stunning.”
In the mid-90’s, this was especially ambitious. To understand why you only have to look at the changes occurring in the games industry around this period. On consoles, 3D games were largely an unknown quantity. 3D platforming was still in its infancy and wouldn’t be fully realised until Mario 64 in 1996, although all the consoles were moving in that direction. Meanwhile, the recent emergence of dedicated graphics processing in home PCs meant that the potential for 3D video games was increasing exponentially. Companies such as id Software led this charge with 1995’s Quake; a game which featured a fully 3D environment and character models. This was a remarkable change from the likes of their previous games such as Doom and Wolfenstein (both also considered impressive in their day), where enemies were rendered as flat 2D sprites rather than fully polygonal models. It was quickly becoming clear that this was the direction in which video games were headed, creating a Wild West environment where developers were still establishing capabilities, best practices and efficient ways of working.
On one hand, this new frontier represented the perfect time for developers to create new tech and stamp their flag in the ground. On the other, there were a lot of risks involved. The technology was still in its infancy, meaning that many PC owners didn’t have capable GPUs. There were a wide range of hardware manufacturers fighting for market share in this space, which only served up more instability and uncertainty.
“With Prey, we were developing it using the 3DFX cards.” States Paul “We even considered making Prey 3DFX-specific, but at that time, other 3D card manufacturers were popping up and 3DFX was losing market share. We knew it couldn’t just run on that one card. At that time, Microsoft was just coming out with their first versions of DirectX – a 3D graphics library to allow your code to work on multiple cards from multiple manufacturers. William Scarboro and I made the call to move Prey to Direct X.”
Moving to Direct X proved to be one of the decisions that protracted Prey’s development. “In hindsight, it was some bad luck in terms of timing. Direct X wasn’t really ready for prime time yet, and we had to basically rewrite the engine from scratch, sort of like taking a Ferrari apart piece by piece and then putting it back together again. It wasn’t really a ‘prime time’ API until DX7”
Even back then, the team considered portals to be essential to the game. However, they weren’t initially intended to be used in the way that you would expect. Portals would actually break up parts of the game world similarly to doorways breaking up individual rooms. “The idea was to load smaller chunks of geometry real-time (so you could do real-time transformations on the geometry), and then connect them with portals.” Paul elaborates “If the engine couldn’t see a portal, then it wouldn’t have to worry about any of the geometry on the other side.”
“Of course, you could totally bog down the engine by allowing a user to see from portal to portal to portal (like looking down a long corridor through multiple rooms), so as a level designer, you had to design the environments so those situations didn’t happen.”
“Then, of course, we started playing with some of the gameplay ramifications, and it was amazing. The Portal games really showcased the cool, mind-bending stuff you could do with those little ‘tears in space!’”
Prey was also to feature several gameplay mechanics that would have been rather unique at the time. The team were working towards the idea of “portal mines” that you could drop around a level in multiplayer, similar to how the portal gun would work in the Portal games. Also, the guns would exist in the game world rather than being part of the UI, meaning that the gun would bob around if you were running, making the gun less accurate. These are features that we take for granted nowadays, but would have been rather exceptional back then.
While the team were still getting to grips with the gameplay details and the emergence of 3D visuals, many of the initial design concepts were also forming. In fact, a large portion of the final game that released under Human Head Studios in 2006 is heavily based on those decisions made back in the mid-to-late nineties. The concept for the character remained mostly complete, with one of the main differences being that the protagonist was originally named Talon Brave. This was later changed to Tommy (with his pet hawk being called Talon in the final release), but the story around his Native American ancestry remains intact. Talon/Tommy also had the ability to spirit walk, which he gained from his grandfather. This was meant to contrast the spiritual side with the science fiction environment that players would find themselves in.
Paul’s team poured plenty of effort into the story at this stage “We actually wrote a novella of the game first, and then used that to guide the level by level design of the game.” Explains Paul. This was certainly not the norm in an industry that didn’t tend to provide rich, interesting narratives at the time. “We came up with a semi-flowchart-like plot scripting language that we could use to script the flow through a level (both in space and within the larger narrative), and I think we were onto something there, maybe a little ahead of our time.”
So it seems that a lot of Prey’s plot survived up until the final release, but were there any major differences? “The alien ship in our version of the game was a modified Dyson Sphere with three diamond shaped land-masses (and not a wholly enclosed sphere), with each land-mass serving as the home world to a different species of alien, while Human Head’s version had the ship be an organic, almost lattice work sphere ruled/controlled by the ‘Mother’.”
While the general story seemed to be coming along, the team were still struggling to rebuild the engine for Direct X. Not only that, but the attentions of the studio were gravitating towards 3D Realms flagship franchise; the Duke Nukem series. Following on from the success of Duke Nukem 3D, the studio was starting to ramp up production on the follow up, Duke Nukem Forever. DNF itself went on to become legendary for its lengthy development cycle, but at the time, it felt like a new Duke Nukem game was just around the corner. 3D Realms must have felt that way inclined, as the production of a follow up affected Prey significantly.
According to Paul “Duke was in full development mode with another engine before we had PreyTech nailed down. At the tail end of our team’s work on Prey at 3D Realms, they were getting some pressure to release Duke Nukem Forever, and they needed more bodies on that team to accelerate the process. With the fact that Prey was in pieces, being refit for Direct X, and then the need to accelerate Duke Nukem Forever, Scott [Miller] and George [Broussard] made the tough decision to pull the plug on our version of the game. It was the right call for the situation and it was in the best interests of 3D Realms.”
Fortunately, Paul seems to be very understanding about that decision. “Though it was hard to see so many years of blood, sweat and tears come to a halt. I’m glad that the game did finally get to see the light of day with Human Head as the developer.”
“Of course, there’s another $64,000 question that we can never answer: would our portal technology, if it survived the move into a mature Direct X API, be able to perform at the speed and frame-rate required for a twitch first-person shooter? Our tests and early levels showed great promise, but with everything we wanted to see in the game, it’s hard to say.”
While Paul’s team wrapped up their work on Prey, the project was never fully cancelled. Instead, it was picked up again several years later by Corrinne Yu who worked on the engine singlehandedly until Human Head Studios took on the project full time in 2001. With Human Head’s appointment, they quickly licensed the Doom 3 engine (id Tech 4) and set about developing the game once more. It finally released on the 11th of July 2006, bringing an end to 11 years of on-and-off development. Given that the PC was primary platform target, it was a surprise that Prey was also released on the Xbox 360 on the same date. However, it seems that a console version had been on the cards for quite some time. “We had just opened up conversations with Sega about a possible Dreamcast version of the game.” Recalls Paul, before adding “Wow. Saying the word ‘Dreamcast’ really reminds me of how long ago this was!”
The happy ending here is that Prey didn’t just end up on the cutting room floor, as so many ambitious projects do. Although the original team didn’t get to see the game through to release, the final product seemed to represent their original vision somewhat. Paul seems pretty pleased that the game saw the light of day, and gave credit to Human Head for their work.
“When Human Head took over the game, they had the core story and concepts from our work, and their shipped game had a look and feel like we imagined for our own iteration. I think they did a great job, and it was so cool to see a world and characters we’d lived with so long finally come to life. They crafted a larger story to fit their idea for the vision and the flow of the game, which is exactly what they should do, and the results were pretty darn strong in my opinion. I’m proud of what Human Head put out, and proud that a bit of our story survived. It really was a great game and I had a ball playing it! Though man, I sure would love another chance to tilt at that particular windmill!”