Gamasutra: The Art & Business of Making Gamesspacer
View All     RSS
August 31, 2016
arrowPress Releases
August 31, 2016
PR Newswire
View All

If you enjoy reading this site, you might also want to check out these UBM Tech sites:

Deadlight Analysis
by Eric Schwarz on 07/17/13 08:00:00 am   Expert Blogs   Featured Blogs

The following blog post, unless otherwise noted, was written by a member of Gamasutra’s community.
The thoughts and opinions expressed are those of the writer and not Gamasutra or its parent company.


Truth be told, I haven't been playing too many new games lately.  Although I've been ensconcing myself within a veritable cloak of older titles, specifically Spiderweb Software's entire product backlog, I've been patiently waiting for two things: several Kickstarter projects (like Shadowrun Returns) to get off the ground, and for the Steam Summer Sale to pick up any titles I may have missed over the last several months.

One of these titles I picked up is a little game called Deadlight.  Released nearly a year ago in mid-2012, Deadlight is somewhat of a throwback to an era of platform games gone by.  Taking more inspiration from the original Prince of Persia, or Another World, than more typical games in the Super Mario Bros. lineage, it features heavy, deliberate controls, over-the-top puzzle- and trap-laden levels, and sets it all to the theme of "zombie apocalypse" that's all the rage these days.

It's rare for me to come across a game like Deadlight.  Not because it's a bad game, or because it's a good game - it's one that sits right there in the middle, occasionally reaching for greatness yet also failing regularly enough to be a frustrating game to play at times.  It's a game of extremes, and unlike a lot of more modern titles which tend not to work, its flaws aren't a matter of simplification, of bad interface design, of poor storytelling... this is a game where the biggest deficiencies come directly from its mechanics, controls and level design just not quite working.


Deadlight pays homage to classic slower-paced platform games by having a protagonist who feels less like a point in space with a character model attached, and more like a physical human being.  There's a strong feeling of momentum as you build up speed, you can trip and fall or get knocked over by enemies, and movement is weighty and deliberate.

It's very refreshing to see a platformer that actually does play like this.  In recent years, with games like Super Meat Boy, platformers have tended towards ultra-difficult micro-challenges which require complete and total mastery of the controls, but in doing so they have also reduced the complexity of the controls and the sense that the player's character is a real physical actor.  Stopping on a dime, extreme air control and so on are all things that we've become used to, so much so that going back to something just a little more "realistic" is actually pretty challenging, and requires significant re-learning.

Deadlight's controls and character physics are generally competent.  The controller layout takes a few minutes to get used to, with the buttons shuffled around just a little bit to make room for the various combat-related functions, but other than that, it's pretty easy to get to grips with.  Movement is analog, not digital, and the momentum and speed that you move at directly determine how far and high you can jump, with no air control to speak of.  You have to line up and time your movement very carefully, and the precise controls mean there is little to no room for ambiguity.  If you fail a jump, usually it's your fault.

Often something as simple as picking up an object becomes difficult, as you slide past it as you run and need to walk back and forth across it a few times to trigger the pickup icon - so much easier if you could just stop at it.

That problem is "usually."  Despite the easy control layout, Deadlight has issues with the precision of character movement.  Because of the momentum attached to how you move around the game world, it can be quite difficult to stop, slow down or line up a jump as necessary, which is a problem in time-critical situations.  There's a noticeable lag to moving your character around, one that's only a split second. but often leads to miscalculation and, frequently, a game over screen.  Because the recovery time from an action is often based on your current momentum, and can vary significantly, you're often never quite sure when you'll be able to turn around, or slow enough to make a standing vertical jump.

Then there's input queuing.  Some degree of input queuing in games is always necessary.  Humans do not have frame-perfect reflexes and it's usually necessary to queue up input from the controller accordingly to do things not exactly when the user presses a button, but when the user expects something to happen.  Getting this delay correct, even if it's just a few milliseconds, is necessary to create controls that feel responsive. 

Furthermore, if a game requires the player to perform complex control inputs quickly, such as pointing in a direction with the analog stick and pressing a button at the same time, there should be a grace period around those inputs so the action can be completed reliably under duress.  Last, there's input cancellation - if the player provides input for an action before the previous action is complete, should the game cancel what the player is doing and perform the new action immediately?  Does this happen always?  Sometimes?  For some inputs and not others?  When it gets right down to it, handling this sort of thing is surprisingly complicated and not at all something the average player is actually aware is happening during gameplay.

Deadlight handles this fairly well, but it doesn't quite hit the mark when it comes to allowing the player to cancel input.  It doesn't appear that there is any way to cancel an action once it's been initiated, until it's been completed.  This can be readily demonstrated by hanging over the side of a ledge - it's very common to end up accidentally mantling up the ledge back onto solid ground, rather than dropping down below, and if you press several button inputs even a second in advance, those inputs will be processed in direct order with no way to interrupt them.  If you press the B button to drop down from the ledge only to change your mind, even before the animation begins to happen, too bad - you're going down.  Though the controls are indeed precise, they don't quite have that perfect one-to-one connection between what you want to do, and what your character actually does.

Level Design

Deadlight's resemblance to older platformers isn't just in the controls, but in the level design as well.  Sometimes it draws influence from Prince of Persia, throwing the player into complex strings of traps and deadly bottomless pits, with success in avoiding these obstacles a matter of timing and accuracy.  Other times it more resembles a game like Metroid, with action elements that require the player to explore a large open room full of enemies and find the way forward.  Yet other times, it's a puzzle game, resembling a 2D Tomb Raider, where the player has to pull levers in the correct order, push crates around to form platforms, and so on.

This pacing and variety is Deadlight's greatest strength.  At any given time you could be partaking in a different kind of action, and when it works, it's grand.  The game is occasionally thrilling and exciting, or you'll have an "aha!" moment as finally the solution to a puzzle clicks into place.  Unfortunately, this variety can also be its biggest weakness.  The game frequently throws challenges at you that the controls simply aren't prepared to handle.  Deathtraps, enemy ambushes, fast-moving segments where the player moves forward at a constant pace, all of these things have little room for error and the controls simply aren't up to snuff.  When you're solving a puzzle or methodically trying to figure out how to move forward, the deliberate nature of the controls is appreciated, but the game tries to make them work in different contexts and fails, often miserably.

It might seem easy once you've done it, but try figuring out this brand-new puzzle and set of mechanics while under assault from several enemies.

Deadlight also often does a poor job of teaching the player new mechanics.  There's one especially egregious example found about an hour into the game - perhaps appropriately, it's set in a lengthy sewer level.  In this bit of the game, the player (for poorly-justified reasons) has to navigate a complex system of traps and obstacles, where the player learns about several new play elements:

  1. Bottomless pits full of spikes
  2. Large spike traps which stick out of the floor
  3. Swinging spike traps that come out of the background to impale the player
  4. Spike platforms that fall down and crush the player after running over a timed trigger
  5. Pressing a button to perform a roll upon falling onto the ground, allowing the player to avoid damage
  6. Signaling an NPC to operate a mechanism, such as an elevator

The problem with introducing so many mechanics at once, literally one room after another, is that the player often doesn't have time to get fully comfortable with one before moving on to the next.  Many of the challenges presented toss the player into time-limited situations, such as putting the player between a horde of zombies and a trap - figuring out how to avoid the trap and escape the enemies is extremely difficult when you only have about one or two seconds to do it.

Another even worse example is seen when the game tries to teach the player the roll mechanic described above.  The player falls down into a pipeline, sliding down at a set pace that can't be controlled, and has to contend with not one, but two floor spike traps, and the brand-new swinging spike trap that comes in from the background.  While this is happening (not before), tutorial text appears on screen to teach the player the roll move - a move that the player's never seen or done before that point.  In the crucial learning phase, not only will the player die trying to read the tutorial text, the player may even die multiple times just trying to navigate the previous obstacles before the real test - the new roll move - can ever be attempted.  And then, even when the player gets far enough to try the move, it might not even work simply thanks to the player's incorrect execution.

That's the real problem with Deadlight's level design through and through.  Even if the controls and mechanics are solid and working just fine, it's often not clear what the player is supposed to be doing at a given time, or whether failure is the result of bad execution on the player's part, or of simply not attempting the correct solution.  When you choose poorly, you die, and have to try again, going through a several-second-long death and load screen, and sometimes unskippable scripted scenes too.  These instant death traps are fine when you're putting the player to the test and it's clear exactly what needs to be done, but just casually, or during crucial learning phases?  It's not just unfair, it's also not fun.

Visual Language

The greatest success of a platformer often doesn't come down to its controls, to its gameplay mechanics, or to other obvious things.  Instead, it's often the visual language that a platformer uses that is able to make or break it.  By visual language, I'm specifically referring to the way in which the game is able to communicate mechanical details through its artwork to the player, in a way which can be intuitively grasped on a near-unconscious level.  If the game can't present critical information correctly at all times, then its visual design has failed.

The original Super Mario Bros. remains a fantastic example of this done right.  Despite the limited technology of the time, Super Mario Bros. successfully communicates everything the player needs to know about the game by creating a visual language which is easy to understand and to identify at a glance.  For instance, the bright blue or dark backgrounds contrast naturally with the darker brown and green bricks used in the foreground, making it extremely easy to tell which parts of the game world the player can actually platform across.  The world itself is divided up into blocks, which share a uniform size and shape that remains consistent throughout the game - making it very easy to judge distances or heights.  The game always moves left to right, so it's always clear which direction you should be going to make progress.

Deadlight tries to create its own visual language, albeit somewhat in reverse.  The game uses a bold silhouette art style that keeps everything in the foreground black, while the background remains brightly lit and often quite colorful and vibrant (for a zombie-themed game, that is).  The game is lusciously detailed and a ton of care and effort has clearly gone into making sure that the core game mechanics remain easy to understand no matter how pretty the complex backdrops are.

The problem is that Deadlight then sees fit to frequently violate the very tenets it sets for itself.  While playing the game, I found myself constantly finding it difficult to tell exactly where to go, what objects I could jump on, whether a wall could be climbed over or not, and so on.  Although the game does keep its "silhouette" idea intact the majority of the time, it's very frequent to see objects with significant depth being used on the main 2D gameplay plane - an overturned truck might actually be a platform, for instance, or a chain link fence could be a hidden ladder.  There are also moments in the game where "middleground" details are important, like overturned bookshelves that can be used as cover from gunfire coming from the background, or traps which swing out and smack your protagonist as he walks by, or zombies that shuffle into the 2D plane.  These aren't highlighted any differently by the game's visuals, so it's not clear if they're important or not until you've learned the hard way - usually by dying, or at least taking some damage.

Can I grab onto those air conditioners on the right wall?  All of them, or just some?  Can I grab onto that wire or is it just scenery? Will that broken scaffolding kill me?  Only one way to find out, and it probably includes a game over screen.

The issue, I think, mostly comes down to the perspective introduced with 3D graphics.  The game is able to maintain its silhouetted visual style most of the time, but when the time comes for the camera to move or for the game to start to tell its story or develop atmosphere through the graphics, the silhouetted foreground and well-lit background tend to get jumbled up with each other.  Furthermore, some objects are interactive - such as elevators, boxes and so on - but they aren't silhouetted despite occupying the same 2D plane as the protagonist and most of the "platforms" you need to navigate.  You can't have them silhouetted because the player needs to know that they're distinct from the static elements of the game world, but with no clear "middleground" language defined it becomes very difficult at times to separate the "game" parts from the graphics.

The developers were evidently aware of this problem, because several objects in the game are indicated using a blue sheen which stands out pretty well from everything else.  However, this highlights the fact that the game's art direction isn't as strong as it needs to be, and it also means that sometimes it's actually too easy to figure out what to do in a level, as you can simply look at what's highlighted in the scene and interact with it appropriately to move on, effectively diminishing the challenge that puzzle-solving might otherwise pose.  That you rely on this highlight to figure out how to move forward speaks to the generally poor visual design overall - if things worked to begin with, this band-aid wouldn't be necessary.

I'm not saying that you can't do this kind of game in 3D, but it's far more difficult when things like shifting camera angles and perspectives get involved.  When you compare Deadlight to Mark of the Ninja, for example, there's no contest.  In Mark of the Ninja there is clear contrast between interactive objects of different types, the platforms you can jump on, and the background.  The game's visual language is such that there is virtually no ambiguity and thus scenes can be read quickly.  The solution to having the player adapt to challenges in time-critical situations for Mark of the Ninja was to make the game easy to read at a glance, allowing immediate skill-based execution to begin; for Deadlight, it was to have the player die over and over until he or she can figure out what objects are interactive, which parts of the level are platforms and then finally execute.

Unfortunately, this visual confusion is probably Deadlight's biggest weakness, even more so than the occasionally awkward controls or annoying level design.  When you can't see what you're doing, can't differentiate the protagonist from the enemies he's fighting, can't tell whether a platform is a platform or a background object, and so on, you've got a game that's bound to frustrate players and unnecessarily accentuates its trial-and-error nature.

Closing Thoughts

I don't think Deadlight is a bad game, but as I've said, it's one of those games that can be both tons of fun to play and exceptionally, controller-throwingly frustrating.  It's not like the game is buggy, or broken, or ugly, or unplayable - it's just not quite good enough to hit the goals it strives for, and that makes it disappointing more than anything else.  However, I'm very glad I played it - because it's a learning experience in seeing both what works, and what doesn't work about it.  It also demonstrates that a successful platformer is often balanced on a razor's edge between frustrations of different sorts, and shows what happens when sometimes a foot slips one way or the other.

Related Jobs

Insomniac Games
Insomniac Games — Durham, North Carolina, United States

Character TD
Insomniac Games
Insomniac Games — Durham, North Carolina, United States

Insomniac Games
Insomniac Games — Durham, North Carolina, United States

FX Artist
Skydance Interactive
Skydance Interactive — Marina del Rey, California, United States


Loading Comments

loader image