Click "Sleep" for a dark background.
Click "sleep" again if text isn't dark.



Mirror's Edge 2D/3D: Reflection, Review, & Repair pt.3

Part 1. Part 2.

Like with all of my repairs, I try to maintain the original spirit of the game with all of my suggestions. Faith is a runner. She's not a secret agent like Bond, nor is she a trained soldier. Her best skill is moving through complex urban environments. Before she got drawn into the plot concerning her sister, Faith was a message carrier. For my repair of Mirror's Edge, I want to preserve and enhance the spirit of momentum, flow, and mobility. First, I will address 3DME and then make a bridge into possible repairs for the 2D version.

For this repair, I think it will work best to start from outermost design layers and then move inward considering that the core of 3DME is the strongest part (ie. will need the least attention). The order is level design, enemy design, level elements, and core mechanic. Keep in mind that this repair only offers suggestions on how 3DME could have been a better Mirror's Edge.



Before addressing any other part of 3DME's design, I want to remove/address all the design features that function as game design crutches with its level design. With these features at play, the developers were free to design levels filled with clutter, trial and error, and an overall lack of well balanced challenges.

  • Remove the check point system. Check points, especially in first person games, have been used to a fault recently. Essentially, infinite check points take away almost all of the punishment and consequence from playing a game poorly. At the same time, developers rely on it to keep players playing when they design their games poorly. Whether a section requires a lot of trail and error/memorization to progress or there are simply too many random occurrences and elements to ensure the player's safety, if players gets unlimited trys at a challenge, they are much more likely to keep trying long after they would have quit otherwise. With many games that feature infinite check point systems, the level designers have settled with designing levels filled will holes instead of designing levels that empower and give responsibility to the player. 
  • There are two ways to go about removing the check points from 3DME/ fixing the difficulty-punishment system. Faith can "do the Mario," which means having smaller, more manageable levels with check points and a limited number of lives. Or Faith can only have one life/one shot at completing a level however large, from start to finish with no check points. Perfect Dark, my favorite FPS, has arguably the best single player level design of any FPS because Rare didn't use check points, regenerating health, or health packs. From the start of a mission to the end, you only have one chance to make it happen. Because of this high penalty, the game is designed to inform the player and give him/her advantages tactically much like Super Mario Brothers is designed so that the player can see upcoming dangers with enough time to react accordingly and effectively. Such a design is what makes the difference between an interactive challenge that's ridged with limited variability and expression and a game where the player is free to play around while completing the challenge. It's the difference between your death being the game's fault and your fault.
  • Adjust fall deaths. Now that 3DME has removed the frequent, infinite check points, the difficulty-penalty design must change across the board. One feature that must change are the heights that Faith dies from. Falling off buildings should still end in death, but just about every other type of large fall in the game shouldn't kill players. By adjusting the level and enemy design, large falls can still have large consequences. The change can be something as simple as taking a lot of stun from large falls. The lack of mobility/control from the stun time will give us plenty to work with latter on.
  • Take away the dependency on runner's vision and the hint mechanic. Like the check point system, the runner's vision and hint mechanic are game design short cuts/crutches. In many 2D side scrolling games, the player instinctively knows where to go. This is not because there's practically only one direction the player can go (left or right). It's because the levels are designed in such a way as to guide the player in the right direction. Some games do this very well (SMB3, Yoshi's Island) and some not so well (Super Princess Peach, Henry Hatsworth). 2D games are very effective at doing this because the level designers generally let go of making the visuals representation, and they embrace the abstract, interpretational quality of 2D visuals.
  • To remove the runner's vision and hint mechanic, the level artists and designers would have to work closely together to design spaces that are easy to understand from the player's perspective. Instead of complex jungle gyms of pipes and hand-holds, gameplay would be better suited in simpler 3D spaces. The runner's vision was designed to help players navigate through the environment by highlighting what level elements players should interact with to reach their destination. So, if Faith could interact with more of the environment and the destination was expanded into a larger area, players wouldn't need the runner's vision as much. The goal with this suggestion is to design a game where the runner's vision is available but not needed without memorizing the level layout.


  • Create environments instead of race tracks. In general, the 3DME levels should be designed as playgrounds instead of race tracks. They should be designed like Mario 64/Sunshine/Galaxy levels instead of Mario Kart tracks. Faith is a runner, but she's not a racer. According to the story, her original job was to carry packages across the city from one point/runner to another. Like in the explore level areas, Faith would have to find her own way through the city. Though I believe 3DME works best with a balance between puzzle sections with limited solutions and explore sections with many solutions/paths, creating environments that can be traversed forwards, backwards, and with intersecting paths would be ideal for my next suggestion.
  • Use folded level design and organic timers. Folded level design is an efficient type of design that makes more use of a given space. With every crease not only does the level make more use of the space, but the dynamics and conditions of the challenge increase. As I've commented on in part 1 of this series, the time trials mode in 3DME gets more use out of 3D spaces because the game presents multiple courses through a single area. Sometimes the course will run in a large circle through the space. Other times, you'll have to stop, turn around, double back, and move through areas heading in new directions, which inherently changes the type of platforming challenge.
  • The crane that lifts up an escape platform, the runaway target, and the helicopter pursuit are all good examples of organic timers in 3DME. Organic timers are great because the player can visually determine the timing of conditions and events. Also elements that make up the organic timer tend to have dynamic effects. For example, in the world if Mirror's Edge, if you only have a limited amount of time to find a way out a building because a team of soldiers are rushing up the stairs to your position, you can either..
  1. Use the time to figure a way out of the building from the floor that you're on.
  2. Decrease your available time by going down the stairs and taking the soldiers head on.
  3. Increase your time by platforming down to the first floor forcing the soldiers to double back down the stairs.
  • Dynamic objectives/game types. Faith may not be able to break bricks with her head, rotate the world using her mind, or throw ropes around for spelunking, but there are other ways to design level transformation. One way to give the player the ability to change the environment is to incorporate more switches/buttons that affect parts of the environement (elevators, cranes, etc.). 3DME already does this.
  • Another more interesting way to have the player's actions affect the types of challenges he/she must face. For example, instead of dying/restarting the level when you fail to keep up with a target because you just couldn't make it through a closing door in time, what if the area on your side of the closed door was a puzzle area. In other words, if you fail to complete one type of challenge, the challenge organically shifts into another type. If you fail to outrun your pursuers, you have to battle them. If you go out of your way to figure out how to get on top of a structure (puzzle area), then you get access to an entirely new area (explore). 
  • Dynamic radio chatter. To compensate from the inherent visual limitations of 3D spaces especially from the first person perspective, the radio chatter that informs Faith and players can be designed to respond to the gameplay situations instead of simply making the game feel more cinematic like a Hollywood action film. Basically, instead of scripted blathering from headquarters, players would play an active part in the information they receive.
  • It would also be interesting if players could gain visual vantage points to feed information back to their home base by going out of their way to scout around off the beaten path. This design would then provide interesting gameplay incentives for exploring areas. In more dangerous/battle situations, players can decided to take the risk and scout or cut their losses and escape. In this way, interacting with headquarters is a way of controlling one's difficulty and gaining information advantages. The 2D coin design from Super Mario Brothers has proved to be difficult to translate into 3D. Many 3D games avoid peppering their 3D game worlds with floating powerups/incentives. But a design element like dynamic radio chatter would be the best balance of the function of SMB coins in a 3D space and the realistic presentation of 3DME.




With all the changes mentioned above the gameplay of 3DME is more focused on what it does best. At this point, it is up to the enemies to create organic timers, and provide enough contrary motion and interplay to keep things interesting. The problem is that in 3DME, most of the enemies use guns. In general, guns in first person games can put a great stress on the design. They're powerful, can be fired remotely, and when you're hit it can be very difficult to tell where the bullets came from. Guns are also hard to design interplay around. Once someone has you in their sights, there's typically nothing you can do to dodge the bullet. Furthermore, for a game like 3DME, the fiction is stressed by guns. After all, how many bullets are we supposed to assume a thin, Asian girl can take before falling? I say one bullet. The game tells me otherwise.

There are a few ways to address the issue of gun wielding enemy elements in 3DME.

  • Enemies can behave less like a well trained SWAT team and more like guards. Because Faith doesn't carry a gun, she is always at a disadvantage in a firefight. So if the enemies were fewer in number, their aim less accurate, and/or they could be caught by surprise the game would be better for it.
  • The enemies should be predictable, even dumb. Combat is only one of the four types of 3DME level challenges. The balance of the entire game would be upset if the player had to spend a lot of time learning the complex AI/weapons of the enemies in combat. It's the same way with Zelda games, which are made up of puzzle, battle, and exploration level types. There's simply no need for a Zelda game to adopt a deeper or more complex battle system. Doing so would only work to draw in more time and focus from the rest of the game into battling. In other words, after learning a deeper battle system, players would begin to look forward to testing their skills/weapons in the next battle and probably grow to resent their time doing anything else in the game. Or, it's possible that the player may fail at battling. So, when the game stages battles that are designed to test the player's proficiency with the deep/complex combat system, the player might grow to resent the battle sections. Balance is key. 
  • The FPS conventions and the Mirror's Edge fiction must come second to the gameplay. This means, it may be necessary to change the nature of the enemies. Instead of trained killers, perhaps the game would be better served if they used non lethal weapons in attempt to capture Faith alive. In such a scenario, players wouldn't be shot full of holes, but stunned from the pain of rubber bullets. The more pain Faith's in, the harder it would be for her to run and maneuver around. When enemies see that Faith's mobility is reduced, they could close in for the capture thus engaging in hand to hand combat. Already, with this simple change, the enemy design has become more interesting and dynamic. Fighting from long range and then closing in to attempt a capture uses space in a more interesting and player-relative way than simply being shot to death from positions outside of view.
  • Guns should be designed as a powerup. Currently, while wielding a gun players gain the ability to engage enemies from a distance. Guns only come from enemies so players have to defeat one to acquire the weapon. Without ammunition clips, the player is limited only to the bullets inside the gun. This type of decay not only functions to encourage players to take careful aim and to move from enemy to enemy to maintain supplies, but it also gives the gun a very temporary function encouraging players to toss the gun without much hesitation.
  • Still, the gun should be more useful to the player as a mobility powerup and not just a combat powerup. A few simple ideas are...
    1. Shooting out lights to blind or stall enemies.
    2. Shooting electronic panels/computer consoles to disable security systems.
    3. Shooting off locks to access new areas.
    4. Shooting out thick glass to access new areas.
    5. Throwing the gun to cause a distraction.
  • When holding a gun Faith's platforming abilities are stunted. While 3DME used this design feature to ensure that players dropped their guns and continued on with the game (effectively reducing the amount of suspension carrying the gun created), there is a much more ingenious way to utilize this feature. Like in Spelunky (and nearly all games with gravity), it's much easier to go down than up. This inherent quality of gravity shapes gameplay in interesting ways because gravity makes going up a very different challenge than going down especially when trying to take external items/objects with you.
  • So if 3DME embraces this dyanamic, additional cases of dynamic game types can be created that revolve around the gun powerup. For example, a level can be designed where Faith is running away from pursuing enemies. If players successfully run away, the level would fall under the chase/escape type. If the player can't run fast enough or decides to stay and fight, the game type will change into a battle. After fighting, the player is free to explore around to find his/her way to the end goal. But, if the player chooses to hold onto a gun, he/she will be unable to uses Faith's highest jump to climb out of the area. At this point, the player would have to decide to ditch the gun, or embrace the limitations of carrying the firearm and find a new/less obvious way out. Choosing to keep the gun would turn the explore challenge into a unique puzzle challenge, which the player can freely abandon by dropping the gun and climbing out of the room. In this way, not only is the level design dynamic, but it also has player controlled variable difficulty. The reward for working through such dynamic puzzle challenges is having a gun in an area that you wouldn't have one otherwise. It's this kind of suspension along with sequence breaking level design that could really open up the level design if 3DME greatly.




Many have commented on the barren world of Mirror's Edge. Pedestrians don't walk the streets and they can't be found in the malls or the subways. The office rooms are furnished but they exude a sense of unuse. Even the graphic style has a hospital like sterility to it. The white are ultra clean, and the colors are flat and sharp. Though I'm a big fan of the look of Mirror's Edge, I do think that the game world should respond to player's actions/mechanics more.

If the clean look to the world matches the cold, impersonal control of information that exists in the oppressive world of Mirror's Edge, then Faith is the opposite of such things. So, if Faith represents freedom and disorder in an ordered world, then I think it's only fitting that her effect on the environment should match as well. To some degree, Faith should smug up the pristine rooftops with her footprints, knock over office furniture with her vaults, and send bricks careening to the earth after jumping onto scaffolding. 

It shouldn't just be cosmetic effects either. More elements need to respond to the player's primary mechanics. There should be many types of glass. Some you can kick through. Some you can smash through with enough momentum. And still some you must shoot with a gun to weaken before smashing through it. Faith needs to be able to kick open doors, kick over tables, and knock over people. Perhaps players would even be able to create vaults or steps by knocking over the right item. At this point, a little level transformation would go a long way.



As I've said before, the core mechanics of 3DME are pretty well designed. A few changes I would make are...

  • Possibly integrating the runner's vision and the REACTION TIME mechanic into the RUN/MOVE mechanic. The more momentum the player builds, the more the runner's vision will kick in and the more the rest of the world will seem to slow down in comparison. Tweaking such an idea is a complicated matter that involves looking closely at the new level design described above. So I'll leave the idea at that. 
  • While running at high speeds, Faith's agility should increase. In other words, when running really fast the enemies shouldn't be able to shoot Faith hardly at all. This will give players a way to counter gunfire by doing what Faith does best. Being able to get up to top speeds requires space. Because moving through space uses the player's primary mechanics, the game becomes more engaging (ie. fighting fire with running).
  • Everything from the ATTACKs (punches and kicks), the player stun, to the enemy stun animations should be smooth, dynamic, and more realistic. A full Euphoria system may not be entirely necessary. But Mirror's Edge would greatly benefit from a level of procedural animation that at least matches level of animation in Wii Sports Boxing. Perhaps a stiff, rag doll system would work best like in Halo 3?



So what can I say about repairing the 2D Mirror's Edge? Aside from the issues I have with the forms and interactions in the game, 2DME is not a good candidate for repair. The core mechanics of 2DME are much more limited than in 3DME. In fact, everything is much smaller in scope and scale from the level design, enemy elements, to level elements. Though the 2D version captures a similar feel of momentum and parkour action of the 3D game, it's ultimately held back by the 2D perspective. Exploring the space in 2DME isn't as engaging or interesting as in 3DME. Also, the 2D visuals naturally create more abstract environments, while in 3D, the buildings and spaces are more representational/realistic. Furthermore, the 3D first person perspective gives players a wide view of the environment that stretches as far as the eye can see. In a 2D space, the player can only see what can fit on the screen. Comparatively the 2D vision range is much less.

Both Brad Borne and DICE did a great job kicking off the entire Mirror's Edge IP. Now, only solid, Classical game design can help Mirror's Edge reach its potential.

« Exploit: Hackers Unite! | Main | Mirror's Edge 2D/3D: Reflection, Review, & Repair pt.2 »

Reader Comments (1)

Very well thought out, KirbyKid! Its amazing how you got something complex like this up and running. Anyway, cheers to more of your projects!

PostPost a New Comment

Enter your information below to add a new comment.

My response is on my own website »
Author Email (optional):
Author URL (optional):
Some HTML allowed: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <code> <em> <i> <strike> <strong>