Jean-François Dugas, the game director of Deus Ex: Human Revolution and Deus Ex: Mankind Divided reveals how the team’s journey began ten years ago, detailing how they brought Deus Ex back to life, and how the game has evolved since then.

Seeing a cult classic like Deus Ex turn into the franchise it now is, with Human Revolution and its sequel Mankind Divided standing beside it, is incredible. Back in 2007, legend had it that making a new Deus Ex game wasn’t possible, especially since a few sequel attempts had either disappointed or never seen the light of day. Yet, it was in exactly this context that a handful of seasoned but virtually unknown developers like me agreed to take on the challenge when the offer came on the table. In Spring 2007 there was only a producer, David Anfossi; an art director, Jonathan Jacques-Belletête; a game designer, François Lapikas; and myself as the game director. I remember those days vividly because we had »carte blanche« for the project, something we don’t get very often, especially in big AAA studios. We had to start somewhere so to launch the project, I wrote on the plastic window glass surrounding our creative room a question that we would spend the next 4 years answering: »How do we turn a developer’s game into a gamer’s game?« The blunt truth is that, at that time Deus Ex (DX) was mainly recognized by critics and game developers as a genre-defining experience, and by a small but very devoted fan base. We needed to make it something more and it was clear to us that something had to change. We didn’t want to change things just for the sake of changing them, though. We wanted to modernize the experience, make it a bit flashier and more accessible, without sacrificing the game’s roots.

The Journey Begins

Our first step was to go back to the source material, Deus Ex and Invisible War, to really understand where the franchise was shining and where it was struggling. I’m a strong believer in the importance of understanding the material you want to play with before thinking about transforming it. It’s just fundamental logic, but often needs to be reinforced because it’s very easy to be tempted by shortcuts. Especially if you’re a fan of something and think you understand or know it well.

Reviving Deus Ex turned out to be a very challenging exercise because we were all big fans of the series to begin with. We all had fond memories of playing the original game when it came out in 2000, and we all had different aspects in mind that we loved. To get beyond that, we had to learn to divorce ourselves from our memories and examine the source material with a very critical eye. We began by identifying strengths and weaknesses of the franchise which could commonly be agreed upon and were not just based on personal tastes. We then moved into defining core values of the experience: what makes Deus Ex a Deus Ex game? We ended up with a list of key words and sentences that defined what it was to us.

I think this step turned out to be very crucial, allowing us to succeed in the journey ahead. Indeed, by really distilling down the experience to primal values like »choices and consequences«, »a mix of several gameplays (i.e., combat/stealth/hacking/social)«, »worldwide conspiracies« etc., we were able to move ahead with a clear view of what the game needed to have. It made us understand both the flavor and the global recipe. It also made us walk with confidence when it came time to play with systems. As long as we kept those high-level goals alive, we knew we’d be heading in the right direction. Understanding the flavor and recipe was key to making the old fan base feel like they were at home even as we explored various ideas to make the game more accessible. Or at least, more dynamic for a new audience unfamiliar with such deep game experiences.

It was during those early days that I decided to stray away from the classic RPG-esque management of the Player Character. In the original game, you had to heal each of your wounded body parts individually to become fully functional. Even though I liked doing this back in 2000, I felt it was detrimental to the experience we now had in mind. I wanted players to focus on their tools, their augmentations, the environment, and the challenge at hand. I felt these were more than enough things to think about, so if I could liberate their minds from the health management aspect to focus on the core experience, I’d do it. Surprisingly, even with all the core value exercises we’d done internally, I met resistance from some of my teammates – and later on, from the original fans when I went public with this design. Nevertheless, we decided to stick with the controversial direction and it turned out to be a good decision that kept the experience focus where we wanted it to be.

»We wanted to modernize the experience, make it a bit flashier and more accessible, without sacrificing the game’s roots.«

We also decided to go with a 3rd person cover-­based system and animated augmentation sequences that unfolded in 3rd person so you could see the powers of your character in action more clearly. Our main theme was »Transhumanism« and we wanted to be sure you’d see your character as often as possible without sacrificing immersion. Adam Jensen is the embodiment of Man 2.0 and we wanted the player to experience this more directly. Switching to a 3rd person camera at key moments was our way to bring more spectacle on screen and make the game experience more dynamic, but it was never done so gratuitously. Seeing Jensen in 3rd person was always done for a reason.
These are just simple examples of how we streamlined or gave a different spin to some aspects of DX while respecting the type of game it is in the first place. There are more, but the key point I want to make is that, without our process, it would have been harder to understand the ramifications and impacts of our vision on what is expected from a DX game. It would have been harder to attract a new audience while still keeping the old fans on board.

The original Deus Ex games strongly influenced the creation of Deus Ex: Human Revolution. In the end, the game was very hard to master for newer players to the franchise, as stealth, and the proper use of Adam Jensen’s abilities, ended up being more effective than relying on traditional shooter gameplay.

The Human Factor

For me, developing Deus Ex: Human Revolution over the course of 4 years was a great learning experience – both as a developer and as a human being. I like to think that it came at the right time in all of the core team’s careers. We were a bunch of people with a solid decade of experience in the industry, but we’d never had the chance to tackle something as ambitious as DX before. Every step of the way, we had to learn how the game’s several pillars would work together, how to make the narrative and the gameplay blend seamlessly, and, finally, how to build the levels that would support it all. There was a lot of trial and error. We struggled for about 2 years before finally getting the global game recipe right and seeing a light at the end of the tunnel.

One of the things we needed to learn first was how to work together as a team. Most of the group knew each other from past projects in other companies, but most had never actually worked together – or at least, we hadn’t done so in a long time. We needed to develop our own culture, our own way of working as a group, and, as the team and the studio grew, our own way of getting new people on board with our vision. It turned out we were a passionate group that worked extremely well together. We all had very strong work ethics and rarely, if ever, wanted to settle for »good enough«. We shared similar values and constantly challenged ourselves and each other, encouraging everyone to push higher and higher. We had to adapt to each other’s quirks, of course – but when you work so closely together every day, the bonds and trust are created much faster.

Dealing with the rest of the team, however, presented some interesting challenges. We in the »creative bubble« were considered to be a very demanding bunch of people, and this created friction during development. We were very hard bosses to please. I believe most people on the production floor wanted to do a great job but if we believed the work they’d done wasn’t good enough, we wouldn’t shy away from saying so. In return, the production team was also very vocal when they didn’t agree with our direction. At times, it was hard to handle.

Ultimately, trust did get earned over the course of the project, for the most part on all sides. And once the results started to show, most people understood that success could only be achieved by everyone insisting on excellence – both individually and collectively as a team.
Let’s be honest, no one likes to be challenged. It’s hard on the ego. But if we were to reach the standards we’d set out to achieve, we all needed to learn to accept criticism. To do this, we needed to grow as individuals, learn how to separate the criticism from ourselves, and understand that none of it was personal. I truly believe this was key to the success of the group and the project.

The Game Factor

From a game standpoint, we also had a great deal to learn. Specifically, about that game’s underlying pillar – choices and consequences – and how it tied to gameplay, narrative, level design, and a million other unforeseen factors. A lot of the choices you see in the game were crafted on paper, but it’s only when you play a game that you realize what works, what doesn’t, and what falls short of its ambitions. Since we’d never made a game like this before, we didn’t fully realize that supporting player choices can lead to a number of unsuspected

10 years back, when the team at Eidos-Montréal initially undertook their journey of reviving the Deus Ex franchise, they determined the vital elements that a Deus Ex game must include, along with which aspects could potentially be changed, by analyzing the original games.

conditions. We didn’t understand how the rami­fications of those choices would create ripple effects that had to be tied up in coherent ways. Even the smallest choice can end up becoming a big headache to figure out. Nothing is ever easy when it comes to Deus Ex. Every little thing connects to other little or bigger things, yet you have to maintain consistency across the board in order to get the immersion right. We had to dive in and learn this the hard way.

For example, we wanted to support the player’s ability to create chaos everywhere. It’s part of the choice and consequence logic: if we give the player the ability to pull out a gun, he or she should be able to shoot it. And once they start shooting, the world needs to react in a credible way. Sounds simple on the surface but building all the systems – the AI, the animation, the writing, the level design needed to support this freedom of choice – is one of the biggest challenges to achieve. For us, it became a trap that was hard to escape from.
Eventually we realized that we wouldn’t be able to support chaos everywhere, as we had initially intended (specifically, in LIMB clinics and Sarif HQ). We spent days and days questioning how to make it logical that a player suddenly wouldn’t be able to use weapons in these environments. We even got to the point where we added turrets in both locations so players would get killed if they did anything crazy. But this didn’t feel right, either. The logic was not that strong; it felt forced and very »game-y«, in addition to being hard to pull off.
After weeks of arguments, we finally realized the logic was simple and had been in our faces all along. Jensen works for Sarif Industries and LIMB clinics are his employer’s business partners. Therefore, he would not condone any violence in either location. It’s just not in his character. This could have been perceived as inconsistent from a design standpoint, but the narrative logic was solid enough to justify it. And when the game finally got released, players never even questioned it.

»Even the smallest choice can end up becoming a big headache to figure out.«

As I write this example, it feels so simple and easy to solve now. But when you’re in the midst of creating the game, with all the challenges you deal with every day, suddenly nothing is ever that simple or easy. The great learning here was that we needed the rigor we had for such a title but we also needed to understand that overthinking sometimes puts us in corners that are hard to get out of. It makes our lives miserable in areas where the added value is virtually nonexistent. It’s all about finding the right solution for the right challenge while maintaining overall consistency.

The Scope Factor

Here’s a nice rule of thumb that every game developer should hold near and dear to their hearts: in order to cut efficiently, you needed a solid plan to begin with. We learned that lesson a few times during the course of Human Revolution. It’s where our blueprint process came in.

The blueprint is to the developer what an architect’s drawing is to a construction team: a plan that tells you what you’re building and how you have to achieve it. For us, it was a docu­ment that described every part of the story from a gameplay perspective and vice versa. We spent roughly 2 to 3 months building all the game’s narrative beats, gameplay sequences, and possible player actions on paper. Where is the player? What is he or she doing? What are the twists? What are the narrative goals? What are the choices? What are the consequences? And so on. We set everything out on paper from start to finish before even beginning production. Once that was done, we could see where the game was weaker and where it needed fine-tuning. The production team was also able to start planning and building schedules based on what was in the document.

Because of that plan, production leads were able to tell us that three distinct city hubs (Detroit, Hengsha, and Bangalore) were too much to create. We had to cut one even before production began. So, we looked at the blueprint and ultimately decided to fold the story of Bangalore into a second visit to Hengsha. Once production got underway, we started building the game we had planned out on paper, only to hit some additional bumps on the road. New schedules revealed that the Bangalore cut wasn’t enough; we still had too much on our plate. I was asked to cut another 25 percent of the game. Hengsha had been planned to have two playable levels: an upper and a lower city. The upper level was falling behind and was the weakest point of the narrative experience we had built so far. So, even though the map was fully playable from start to finish and was one of the most stunning maps we had, we ended up cutting it. Since we had our blueprint on hand and constantly kept it up-to-date, we were able to analyze all the affected narrative sequences and determine how to retrofit them into lower Hengsha.

To better illustrate Jensen’s augmentations, and to have some spectacular scenes during gameplay, the team decided to retain a 3rd person view whenever the player used special abilities.

For example, during our first cut, locations that we were supposed to visit in Bangalore – like Belltower Headquarters – had been moved to Upper Hengsha. When we had to cut again and Upper Hengsha was placed on the chopping block, we couldn’t fit the compound into the lower part of the city hub. So, we ended up cutting the Belltower map entirely. Pieces of storytelling developed for the map were then moved to another level, Omega Ranch.
Deciding how and what to cut was a process that took Mary DeMarle, the narrative director, and me about three weeks to complete. We were able to do it without too much pain because the blueprint showed us which pieces of storytelling weren’t critical to the narrative experience, and which could be moved to other maps without damaging everything that had been built so far. Because we had a solid plan, we were able to play with the pieces individually and always keep a good understanding of how everything worked together.
Spoiler alert: We thought we were done with cuts at this point, but a few months later we were asked to do it again. We had planned for a small compound in Utah that was supposed to help people recover from having their augmentations removed. By visiting it late in the second act, players would discover a key plot twist: Purity First, an anti-augmentation group, were hiding a terrorist training camp deep inside the facility. We looked at our story at that point in the blueprint and realized we could only cut the map, not its narrative beat, because the reveal was directly linked to Jensen’s investigation. So, we found an area in Detroit’s sewers where players could encounter the Purity First leader and discover the crucial plot twist during their second visit to the city.

The experience of returning to the blueprint time and time again in order to see what to cut showed us how instrumental blueprinting a game is. Especially for titles as big as Deus Ex. The return on investment is much higher than the effort you need to put in at the beginning of development in order to get there. At least, it’s the way of working that we embraced as a team.

Groundhog Day

On Human Revolution, we were often saying it takes a lot of naiveté to create such an ambitious game. By the end of the project, our philosophy had changed: naiveté was no longer possible. We told ourselves that if we got the chance to make another Deus Ex, this time we’d need to have courage.

Mankind Divided was to be the 4th title in the DX franchise, and while we still wanted to retain the core values we had identified with Human Revolution, this time our goal was to bring the experience to the next level – specifically in terms of production values and accessibility. We knew that Human Revolution’s game mechanics were a bit stiff, that the game balancing was punishing, and that combat wasn’t perceived as a truly viable option. We also knew that, visually-speaking, the game was artistically beautiful but disappointing from a technical point-of-view. If we wanted to appeal to a broader audience, we needed to work on these aspects without compromising what had made the last game stand out in the first place. Key to achieving our new goals for Mankind Divided would be to ensure that the team stayed somewhat the same as with Human Revolution. We managed to pull this off to some degree, but we also brought new blood in to several key positions.

For Deus Ex: Mankind Divided, the team thought about the balancing of stealth and combat from the get go, ensuring that both ways of playing would be equally fun.

In some ways, doing so made it easier to work on Mankind Divided because when team members started talking about core principles, the »old-timers« immediately understood what was meant. They recognized how things should feel, play, or look. But it also made it harder because the »new blood« often wanted to challenge the status quo. Sometimes they were right to do so, sometimes they weren’t. This created some interesting dilemmas to manage but in the end, having a mix of new- and old-timers working together prevented us from becoming complacent and enabled us to bring in different ideas. To make the game feel immediately more gratifying and in line with its fantasy, we decided to make Adam Jensen proficient right from the get-go, instead of relying on the classic route in which you have to build your character from the start. The new story showcased an Adam Jensen taking charge, fully embracing his augmented body, and assuming the role of Man 2.0. The game progression, controls, and mechanics needed to reflect this, too. If they didn’t, they’d end up subverting the narrative. The problem we faced here was two-fold: how do we present an empowered Jensen who has some new and exciting abilities, while also letting players customize the character any way they want? We solved it by building an intriguing narrative around an unexpected terror attack and subsequent accidental disco­very that removed some of Jensen’s augmentations early in the game, while simultaneously exposing a new array of mysterious abilities and gun-arms.

»In order to cut efficiently, you needed a solid plan to begin with.«

The gun-arm augmentations were introduced as experimental tactical weapons with lethal and non-lethal capabilities. They offered a »one-two punch« kind of firepower that could be used on-the-go in combination with basic weapons. We also allowed players to either go for the kill or knock-out their enemies non-lethally no matter which mode they were playing in: combat or stealth. Human Revolution had been way more rigid than this, by associating kills to combat and knock-outs to stealth. With Mankind Divided, though, we really wanted to give players a sense of control, immediacy, and flexibility.
Brainstorming such elements was only a part of the equation because we also needed control schemes that supported our new philosophy. We had to work hard to standardize the default control scheme because Mankind Divided is not your typical shooter experience. It might have seemed unconventional at times, but we wanted players to always feel in control of their character, like they really were playing as Man 2.0. Therefore, we needed to reduce the learning curve as much as possible.
We also needed to balance the new game differently. On Human Revolution, we had been so scared that people would not use their augmentations, the environments, etc., that we balanced things in a way that »encouraged« you (i.e., forced you) to use the power at hand. For the most part, this wasn’t a problem with core fans. But a lot of franchise newcomers had their asses kicked so hard, they felt left out. When we made the DXHR’s Director’s Cut version, we corrected this by giving more default energy and loot, rebalancing health points etc. Doing so made the game more accessible, and the feedback we received afterwards proved we could craft a more forgiving adventure without sacrificing the Deus Ex experience. So, when it came time to tackle balancing in Mankind Divided, we moved forward with confidence.

All of the changes discussed so far also helped to make Mankind Divided’s combat and overall experience more visceral, more immediate, and impactful. Combat had been very static and punishing in Human Revolution and we really wanted to do more to improve it. So, we worked

The team knew that Human Revolution was, visually speaking, artistically beautiful. However, they were also aware that it was disappointing from a technical point-of-view, and that’s something that Eidos-Montréal worked hard to improve on in Mankind Divided.

on several other systems as well, including the enemy bark system, environmental reactions to bullets, and enemy varieties and their ability to use  augmentations.
In the end, bringing Deus Ex to 2016 wasn’t about changing everything. Rather, it was about understanding what we had achieved with Human Revolution – the good and the bad – and knowing where we could shake things up a bit without compromising the core experience. Mankind Divided was built on the same core values as Human Revolution. The difference was that we now had experience in building this type of game, so our experiments could be undertaken with more confidence.

I remember how, after the release of some action-oriented trailers and interviews in which we said combat would be more exciting this time around, several fans worried that we might sacrifice stealth and make an action game. But the game is released now and fans are still finding the recipe they loved in Human Revolution, except with one added note: This time around, combat is a valid option.

Closing Words

Looking back at the last decade spent working on Deus Ex, I can honestly say that the team is proud of what it’s achieved. Are Human Revolution and Mankind Divided perfect games? Hell no. But they were made by a group of passionate people who took the endeavor very seriously. We were extremely honored to be given a chance to give new life to one of the most influential games ever made – a game crea­ted by a different bunch of passionate people who came before us. The least we could do was tackle the challenge with the same energy and commitment they did when it all started. I can proudly say that the Deus Ex legacy lives on.

About the author:

Jean-François Dugas
joined Eidos-Montréal in 2007 after working for Ubisoft for about 10 years. He was Lead Game Designer on several projects for the N64, PS1, and Xbox, focusing on establishing the vision, creating design documents, and tweaking gameplay. During his time as Creative Director on projects for the Xbox and Xbox 360, he focused on establishing the vision, then on working with designers, artists, and programmers to ensure that the vision for each game was fully realized.