How Amnesia: The Dark Descent Was Made and Saved The Developer

Amnesia: The Dark Descent reinvigorated the horror genre back in 2010, but it had to go through many hurdles to accomplish that. It was renamed twice, went through multiple concept revisions, one of them was even inspired by Super Mario and what’s worse, it was plagued with numerous budget issues to the point that ever completing the game seemed unlikely. It even almost caused the entire studio to go under. Despite going through a horror experience of their own, the small team at Frictional Games kept believing in the project and delivered a truly sensational horror game that’s still talked about today.

Development on Amnesia began in late 2007 when the four person team was still hard at work on Penumbra: Black Plague, the second installment in the Penumbra series. Back then, the Swedish studio wasn’t doing great financially so they couldn’t afford to take a break between games. In other words, the team needed to come up with a new concept to work on right after Black Plague would hit store shelves. Even though they didn’t immediately have a solid idea, the four developers agreed that work on their next game had to be streamlined, since creating their previous games proved to be too hard of a task for a studio their size.

Additionally, the team had grown a bit tired of making games like Penumbra that were very story and puzzle heavy and instead wanted to focus on an idea that incorporated more gameplay mechanics. Not only would this make development easier, it would also help them pitch the game to publishers, since it’s easier to explain how fun and interesting the game mechanics are than it is to convince them how great a story might be or how complicated a puzzle is. In February 2008, when Black Plague was done and released, Frictional Games unfortunately didn’t have a new fully realized game idea yet. Some of the basic technology was up and running, but there was no concept for the artist and designers to work on. 

Luckily, Black Plague did well enough that publisher Paradox Interactive approached the studio to see if they were interested in making an expansion for the Penumbra sequel. Frictional Games didn’t hesitate. An expansion meant they could keep the entire team busy, while also being able to set aside enough time to brainstorm for their next big project. It was quickly decided to stay within the horror genre, as the developers felt this was a genre they were most comfortable with and would give them the opportunity to apply all the knowledge they collected from working on the Penumbra series.

Interestingly enough, the team’s first instinct wasn’t necessarily to go bigger and better, but rather, to take a few steps back and consider how they could make things simpler. They thought about how their previous titles never had much replay value, because of a lack of gameplay mechanics in each environment. The player would solve a puzzle in an intricately crafted level, perhaps receive some story exposition and two minutes later move on to the next environment. Because of this, the developers never got much use out of their time spent designing and programming various levels. Therefore, the team wanted to try and see if they could ditch most of the puzzles and instead come up with more interesting gameplay that was based on mechanics that could be used in multiple ways.

Inspiration for this new direction came from all sorts of places. Creative director Thomas Grip remembers how frightening it was to play Silent Hill 1 and 2, something the team wanted to recreate. Text adventure games like Zork made the developers wonder if they could create puzzles in a 3D environment that were logical and felt like they belonged to the world, as opposed to puzzles in games like Resident Evil. Early on in development, they also looked at movies like Saw and Hostel, since they were very popular at the time. The slow, minimalistic build up that’s present in the final game is inspired by older movies like The Haunting. Horror games like Call of Cthulhu: Dark Corners of The Earth and lovecraft as a whole influenced a lot of the general atmosphere and designs throughout the entire development.

One of the more surprising inspirations was Super Mario, specifically Super Mario 64. “Mechanically, we started focusing on making the horror game equivalent to Super Mario: short levels, simple gameplay, repeated mechanics, easy to understand goals and an experience well suited for bite-sized chunks of entertainment.”

Early design documents had examples of the player being stuck in a cell, with someone else being chained up against the wall at the far end who might have important information for you to progress the game. In order to get to that person, you have to avoid or disable a bunch of traps. Along the way you might pick up power-ups, secret notes and other rewards. Paradox Interactive liked the initial ideas, but wanted to wait with signing contracts and providing funding until Penumbra: Requiem was released.

Around spring 2008, the studio applied to the Nordic Game Program, a government program that grants money twice a year to game developers. Any additional funds the developers could get their hands on would be a big help to fully realize their vision. They wrote the title “Lux Tenebras” on the application forms, which is rudimentary latin for “Light and Shadow”. The name was strategically chosen to make the game sound more simplistic and less violent and therefore more likely to be approved by the government. What’s more, “Light and Shadow” also refers to a specific mechanic, namely that the darkness itself would be an enemy in the game.

The player would have to find ways to activate light sources such as candles and fire places to stay safe from the darkness. Furthermore, they also had ideas to build puzzles around this specific mechanic. This concept was inspired by an obscure PS1 game called The Note that Thomas never played for himself, but he did read about it in some walkthroughs. The game used similar mechanics which sounded really cool to Thomas.

Despite this connection, Frictional Games was almost immediately dissatisfied with the title for multiple reasons, one of the being that it was nearly identical to Penumbra, which roughly translates to “the area between light and dark”. During summer of 2008, the studio got the news that they were approved for the Nordic Game Program and received 40,000 euros to spend on development. A huge relief for the developers and provided some much needed motivation. That same summer, the team had to go into crunch mode to finish up Penumbra: Requiem, which was slated to release in August. After Requiem’s release, they were now finally able to shift their full attention to the project that would evolve into Amnesia: The Dark Descent.

By the end of the summer the team had finished the core engine features, very basic gameplay mechanics and a simple version of the level editor. While the developers had retained the physics interactions from Penumbra, pretty much everything else couldn’t be more different from their previous titles and more closely resembled Super Mario 64. “The game was divided into several hub areas that connected all of the games’ levels. By solving certain quests, new levels would become accessible and eventually lead the player into a new hub. Each level even saved stats such as quests completed, items found and best completion time. The whole setup was very similar to Mario 64.”

One very important thing was still missing however: a crucial gameplay loop that can be executed in multiple ways. This was by far the hardest task the studio faced throughout development and many of the things they tried ended up not working. The light and darkness gameplay felt very artificial and forced, sucking away all the atmosphere that a good horror game is supposed to have. The stark contrast between light and dark areas just wasn’t working. A good example of a mechanic that sounded great on paper according to Thomas, but it didn’t actually make for a good horror game once they were testing it. 

This setback prompted them to try a more Penumbra-style approach where the levels behave more like mini-games and the player has access to weapons and other tools. Around the same time, the story was taking shape and featured a secret society doing strange things inside a castle in the 18th century. When fall came around, the studio started negotiating with publishers to see if they were interested in funding the concept. Paradox Interactive was still on the top of their list, but the team wanted to keep their options open. They were still trying to figure out if they either wanted a fully financed deal, meaning no IP and lesser royalties, or if they could partially finance it themselves with a bank loan which presented a much greater risk.

By the end of 2008, the prototype originally pitched for the Nordic Game Program was completed and the title “Lux Tenebras” was changed to “Unknown”. Thomas Grip admitted the new name wasn’t great, but thought it would suffice. While the new prototype provided an efficient way to add more content and incorporated most of the basic gameplay, playing it just didn’t feel engaging according to Frictional Games. Despite this, the developers convinced themselves that adding more features would fix the problem down the line.

To add more bad luck to the mix, the studio had trouble finding the right publisher to partner with. In the end they decided to secure a loan in addition to getting funding from Paradox, who were very happy with this outcome considering they had to invest less money and still get the publishing rights. However, nothing was officially signed yet and by the time Frictional Games needed the funds from Paradox, the studio found themselves unable to secure a bank loan. This meant renegotiating with a very upset Paradox Interactive and the developers were lucky to still secure a deal with the publishing company.

In the spring of 2009, the team was experimenting with combat but none of their ideas were working out. The general concept was for combat to be a last resort for the player, something that sounded good in theory but was very difficult to accomplish in execution. Frictional Games also felt many people would see the combat as sub-par compared to other games, which ultimately convinced the devs to scrap it altogether, simplifying the design of the entire game in the process.

The contract between the studio and Paradox was signed in April 2009 but less than a month after, the developers were unfortunately forced to break it. Frictional Games, along with their current project, were at a breaking point. All of the studio’s money was nearly depleted. Enough was left to pay for another month of salary but then that was it. They would either have to start making preparations to absolve the entire company or quickly find another way to gather more funds.

Frictional Games chose the latter, as they weren’t giving up on their project just yet. Their first idea was to make the game a lot simpler so they could release it quicker. This involved changing the control scheme so that it was entirely controlled by mouse and to create more of a point-and-click escape room scenario. However, it didn’t take long for the studio to scrap this idea entirely. The team started feeling desperate. “Finding some kind of financial backing did not go much better. Banks were still skeptical about the project, angel investors did not understand the project and publisher interest was low. I actually started to prepare my CV to look for job opportunities.”

Just as everyone at the studio was about to give up on the project, something unexpected happened with their previous series. In June 2009, Steam offered a 75% discount on the Penumbra Collection and even though Paradox Interactive owned the digital rights to Black Plague and Requiem, Frictional Games owned the complete rights to Penumbra Overture. As a result, the studio would receive over a third of all profits from every sale. Their expectations were low but in a shocking turn of events, the weekend-long deal sold more units than the combined lifetime of all Penumbra games.

It was the stroke of luck the studio needed and by cutting salaries in half for all employees, they were able to keep production going for a while longer. The money from the sale still wasn’t enough to finish the game though so after feeling inspired from the Steam sale, the studio decided to create their own sale of Penumbra, with Linux and Mac versions included. Profits from that sale matched the income from the Steam sale and now, at least financially, the studio had completely turned things around.

After seeing all the renewed interest in the Penumbra series, the developers decided it was best to step away from the Mario 64 hub design and focus more on crafting a much more linear experience that leaned closer to the Penumbra games. Their new motto during development even became “like Penumbra, but better”. To save time and resources, they kept as many assets and level designs from previous development builds as possible. This shift in focus also made the team reconsider the story. 

The previous storyline involved unorthodox scientists but the developers now felt it was more interesting to explore the innate evil that humanity is capable of and have it take place in the 19th century instead of the 18th. The narrative was influenced by the Milgram and Stanford Prison experiments and the acts done by Unit 731 during World War Two. Sadly, the main writer wasn’t interested in this new direction and parted ways with the studio.

This new beginning for the project also came with a whole new take on the gameplay. The developers were still keen on preserving as many of the Mario-like features as they could, but as everything progressed, less and less of these mechanics made sense for the new direction. As mentioned before, one of their previous ideas involved designing puzzles that worked with light and darkness and the team wanted to see if they could perhaps rework that into something else. 

This eventually sparked the idea for the sanity system that’s present in the final game. If the dark couldn’t work as a gameplay mechanic for puzzles, perhaps it could still affect the player somehow. Losing sanity was something that’s been in games before but Frictional Games felt inspired and wanted to test how it would affect their own game. Experimenting with the player’s sanity opened up the imagination of the developers and made it easier for them to build scarier environments.

The team expanded the sanity system to the point it became the most important part of the experience and it was even intertwined with the story. The protagonist has a phobia of the dark and will become mentally unstable if he stays in the dark too long or if he sees anything unsettling like monsters or corpses. Lower sanity causes the player to move slower and obscure the surrounding environment, making it more difficult to escape certain situations. If the player doesn’t take care of its sanity by using potions or staying in the light, they would eventually die and have to reload. 

Unfortunately, it was very difficult to find the right balance for this gameplay mechanic. Either testers were really good at avoiding the dark completely and therefore much of the scary atmosphere was lost on them, or people would have a lot of trouble avoiding the dark and had difficulties just progressing in the game. Working on Penumbra taught the developers a very valuable lesson, which is to always ask if a certain mechanic needs to be in a game at all. 

Stripping out combat in the Penumbra sequel made it better, at least according to the team, so what if they just removed the sanity system? Well, they didn’t cut it entirely, but the fail state was removed and this vastly improved the experience for all testers. Around the same time, the studio felt like changing the title of their game once more, as Unknown proved to be unpopular. The team simply voted on a number of suggestions and eventually landed on Amnesia: The Dark Descent.

Fast forward to February 2010, when Frictional Games completed the first alpha build of Amnesia, one of the most important milestones during development. With this finished alpha build, the studio felt confident enough to set a release date of September 8th. However, they didn’t have enough money to keep development going for that long so the hunt for a publisher had to start again. In the meantime, the developers decided to make a discounted pre-order of Amnesia available to the public in order to raise extra funds. The reception was positive although the pre-orders were sluggish.

Fortunately, the team was able to secure a deal with Russion publisher Snowball and 1C-SoftClub the following month and received the necessary budget to complete their horror experience. According to their calculations though, they would almost be completely broke by the time Amnesia would be released so it wasn’t an ideal situation. However, another sale would save them yet again in May, when a Humble Indie Bundle with Penumbra: Overture packaged in went live and sold extremely well. To keep the momentum going, Frictional offered discounts of Penumbra: Black Plague and Requiem and also lowered the pre-order price for Amnesia. When all was said and done, they had more than enough money to survive until Amnesia’s release and even well beyond that.

As mentioned earlier, the gameplay of Amnesia was going through a big change. Instead of delivering a Mario-esque horror experience, the developers wanted to create a game that was much more emotional and messed with the player’s feelings. Lots of already implemented features were cut because of this. For example, a progress bar for each level and the heavily used coins were both scrapped. The number of items per map were also reduced.

Lots of these cuts and changes were a result of the studio’s very limited budget, which forced the developers to come up with more creative solutions to simplify all the game’s aspects. Throwing out combat, implementing an interactive physics system instead of creating detailed animations, backtracking through the same areas, these were all things that were mostly done because of budget constraints and the limited number of people working at Frictional Games. Many people’s favorite monster, the water lurker, is yet another example of this. The monster was supposed to have its tentacles and other details come out of the water, but it ended up being much easier to just create some water splashes.

It was only afterwards that they realized cutting all those corners actually improved the overall atmosphere and the player’s emotions during gameplay. Not knowing what the water lurker looks like or what it’s capable of for instance, conjures up a much scarier picture in the player’s mind than the developers themselves could ever create. What’s more, by removing more and more features, the devs also noticed that players were starting to focus on the smaller details like scary background noises. Freeing up the player’s attention caused their imagination to run wild just by hearing the slightest sound. The unknown is often much scarier than having a firm grip on the rules and knowing what’s to come next. Once the developers figured this out, they started to double down on this concept and wondered how they could push the imagination of the player even further. 

The sanity system at this stage in development was merely there to improve the atmosphere by triggering certain visual effects and, outside of slowing the player down, didn’t actually impact gameplay whatsoever. It simply needs to give people the illusion that they’re slowly going insane. Of course, the player doesn’t know this and the studio decided to keep it that way. At no point does Amnesia explain the exact consequences of the sanity system and people are left guessing. In fact, Thomas admitted they went the opposite way and lied to the player to increase the tension and make people believe that sanity is crucial to staying alive. “There’s a hint in the game that says: ‘if you go too insane, monsters are going to have an easier way of finding you.’ This is a lie. There is no logic at all like that in the game. But we’re planting an idea into the player.”

Running out of sanity doesn’t trigger a fail state and instead causes the protagonist to pass out for a moment, regain sanity and continue as normal; but because the sanity mechanic is so obscure, the player might think it was just a close call and that they’ll surely die next time it happens. Dying from low sanity is simply not possible, however. Thomas explained that it’s not very common in game design to hide the gameplay rules from the player, let alone give them false information so they think how a certain mechanic works. In the case of Amnesia though, it couldn’t have elevated the tension and horror aspect more.

A few months prior to the game’s release, Frictional Games was working hard to get a preview build ready for gaming outlets to try out and write preview articles. Since they had no budget for marketing, the team was solely relying on the strength of the game to generate good press. This led to mostly positive opinions, but also revealed an issue with the sanity system. Critics agreed that the sanity potions that regenerated the sanity meter were ruining the atmosphere of the game. After taking another close look at how it affected the overall experience, the developers removed the potions entirely just weeks before launch.

Amnesia: The Dark Descent was released on September 8th and both critics and gamers alike loved Frictional Games’ unique take on the horror genre. It’s safe to say the team couldn’t have been happier. “Personally, one of my own favorite moments occurred five days before release, when the IGN review went out. The game had scored 85% and was also given an Editor’s Choice award. For the first time I felt that these three years of hard labor might actually amount to something.”

Now, just because the game was receiving critical acclaim, that doesn’t necessarily mean it would sell in great numbers. Before release, the developers had decided that if Amnesia didn’t sell 24,000 units during the first two months, they would close down the studio as they otherwise wouldn’t have enough money to keep everything going. Of course, Frictional Games is still around today and that’s because they sold more than 34,000 units after only a month. By July 2011, that number increased to 350,000 units and later in 2012, Amnesia had sold an estimated 1.4 million copies. Those numbers far exceeded the team’s wildest expectations and it meant they could now, for the first time ever, give everyone at the studio a proper salary.

Their success led to an expansion, more installments in the series, console releases and solidified Frictional Games’ future in the gaming industry. Amnesia’s development history is a perfect example of how important it is to experiment, always be open-minded and to ask yourself if a certain gameplay aspect is really necessary for the overall experience that you’re trying to deliver. Cutting certain elements might lead to some surprising results and inspire better systems. Furthermore, it shows the power of people’s own imagination and how you can use it to your own advantage as a developer. Sometimes less is more.