The End of the Line

Greetings Lichdom fans, John Schuch here, lead level designer on Lichdom Redeemer.  I’m taking a break from the grindstone that is CryEngine Sandbox to talk a little about the work that gets done on game levels as development gets close to alpha – and beyond.

I’d like to begin with a quick overview of how the level design process begins and how you get to where we are now, nearing alpha, to provide perspective.

Get out your graph paper

Many moons ago the level designers sat down with paper and pencil and sketched out the game levels using a few key pieces of information:  Geographical location, major events (i.e. bosses), and story beats.  From these pencil sketches more detailed (and legible!) documents were created that laid out the level in its entirety.  After many revisions and approval by the content, art, and design teams, the levels get blocked out in rough form, gameplay is put in place, and play-testing begins.

Hand-drawn paper map.

Hand-drawn paper map.

Paper map cleaned up prior to blockout work begins in the Sandbox editor.

Paper map cleaned up prior to blockout work begins in the Sandbox editor.

Weeks of play-testing, adjusting, play-testing, tweaking, etc., etc., happen and then the level is handed off to the art team to do their magic (see Dan Brown’s blog post about environment art).

Level after art pass.

Level after art pass.

Skip ahead a few months and we now have the entire game built and looking pretty.  Now what?

Level design isn’t just about getting the player from point A to point B – we’re the end of the line for the whole team.  We’re the ones that take everyone else’s hard work and put it all together.  It’s as though we’re standing at the end of an auto assembly line watching a mountain of parts heading towards us and we have to put them all together.  The end result must be a fully functioning car.  Imagine standing there watching, for months and months, as those parts are slowly being created, one at a time, and then all of them showing up nearly at the same time!  This is how the final months of development happen for level designers.

So we have all the pieces and parts we’ve waited so patiently for, we put them all together, and then we play the game over and over and over again.  And then we play it some more.

Graybox blockout of a level.

Graybox blockout of a level.

Art pass of the same location.

Art pass of the same location.

Shower, rinse, repeat

As a game level progresses through its life there’s a seemingly never ending stream of changes made to it.  A game level is never really “done,” only “as good as time will allow.”  There are always more things a designer wants to do to levels in the final months but doesn’t have the time or resources to do.  Some things, however, are absolutely necessary.

The greatest supply of changes comes from play-testing the game.  Most play-test sessions end with a long list of things for the designer to go back to his or her desk and fix.  Everything from combat that is too hard (or too easy) to areas that are confusing to the player to broken scripting.  The items are taken care of and then it’s back to the play-test lab for the next round.   Rinse, repeat.

I should take a moment here to describe the level designer’s worst enemy:  The Jerk Tester.  This is the play-tester that sits down and proceeds to play the game and break every level in every possible way that not a single level designer on the planet could have ever conceived.  They use the world to take advantage of combat.  They find shortcuts that no one else sees.  They lure enemies to places no one would dare go.  This is a case of your worst enemy being your best friend:  The Jerk Tester is the tester that, in the long run, makes the whole game better by being “that guy.”  It sometimes becomes a game between you and The Jerk Tester to try and be one step ahead of him by anticipating the next unexpected thing they’ll try and prevent them from doing it.  Seeing them try and fail because of the designer’s anticipation is worth every penny.  Again, all this makes the end product a more solid product, despite The Jerk Tester’s annoying game breaking!

There are always parts of the game that a level designer loves and wishes to make work as perfectly as possible, but just don’t.  These are the hardest changes (read: cuts) to make.  You may have spent three or four days on that awesome scripted event and combat setup but it just doesn’t communicate to players well or just flat-out isn’t fun, no matter how hard you try.  Around alpha is when these bits get thrown in the level dumpster to be hauled off and forgotten about.

On the flip side, there are times when parts of levels seem to be headed to the dumpster and a fellow developer will come along and make a simple suggestion that saves the day.  This is one of the advantages to having the entire development team involved in playing the game on a daily basis and being given the freedom to make such suggestions.

Level design isn’t just about pretty landscapes and hoards of baddies to kill.  There’s a vast array of technical work done by the designer under the hood of the level that the player is never aware of.   We’re responsible for everything from making sure the correct areas of the world are showing or hidden at the correct times, all of the NPC navigation data is valid and efficient, and ultimately that the framerate stays at an acceptable speed by using special tools and ingenuity.  This work is especially important towards the end of development when more and more content is put into the levels.  That blocked-out prototype map may have run at an impressive 120 frames per second at the start of development, but now that it’s a full level with art, story content, cinematics, visual effects, etc., it may tank and only run at 25 frames per second. Staying on top of the technical side of level design is an absolute must as the game goes through alpha and towards shipping.

A typical combat setup as seen in the Sandbox editor.

A typical combat setup as seen in the Sandbox editor.

Seconds to liftoff

As alpha, and shipping, come closer and closer, the game — and specifically the levels — really start to come together.  It’s feeling like a real game now and not a disjointed collection of locations to set fire to and freeze NPCs.  Story content is in, cinematic moments are in, progression through from New Game to You Win is possible.  This is when the real fun begins for level designers, when we can go in and polish the levels until our bosses say, “Stop” and yank the mouse and keyboard away from us.

As of writing this blog post, one area of one of the levels in Lichdom Redeemer has had eighty-nine revisions just to the gameplay.  That’s the result of the cycle of play-testing, improving, play-testing, reviewing, removing, adding, etc.   And we’re still not done!  Every level in the game will have this level of revision through to the end.

One of the most advantageous methods to achieving highly polished, exciting, and (most importantly) fun levels is to have each level designer have their own time with all the levels in the lead-up to shipping the game rather than assigning level X,Y, and Z to designer A for the life of the project.  Having multiple brains pouring over the levels always results in much better designs, whether it’s minor combat changes or additional gameplay spaces being added.  Another upside is that it prevents level designer burnout – staring at the same game level for months and months breeds complacency (and boredom) and the level suffers.

The march towards the wonder that is “done” from around alpha is when you start to notice inconsistencies across the levels.  They are usually not apparent until this point when you can play the game, in order, through and through.  Decisions are made, the levels are tweaked, gameplay is updated, all to ensure the player has a consistent experience.   It is vital that a player’s expectations are met and to not change the rules from level to level.  Prior to this time in development designers are so focused on the individual levels that inconsistencies can slip through.

The final months are like Christmas for level designers.  We finally get all those cool toys we’d been promised so long ago.  Toys such as new NPCs, game items, or even just new scripting functionality.  Loading up each level and going through putting all the new toys in is a blast and is rewarding in itself because you, as a designer, finally feel like you’ve gotten what you’ve always wanted.  The ultimate reward is watching players experience everything you’ve worked so hard on and having fun doing it.