Catalyst chopped up its own city
8 min read
By Eduardo Orellana ·
DICE, 2016, fixed a lot about Mirror Edge and still lost the run. Burnout Paradise, Criterion, 2008, kept every event on the same streets. Catalyst did not.
The city was cut into levels
The first Mirror Edge is an obstacle course. You learn a line and you run it. Catalyst, DICE, 2016, repairs a stack of complaints about that game and puts you in a city, and the runs get worse. The city is asked to hold missions, travel, and the fantasy of choosing a line, and it holds them as separate slices. You fast-travel. You start a mission. The city between is a menu you can fall off.
Burnout Paradise, Criterion, 2008, is the counterexample sitting in the same decade. The events are in the city. You drive to them on the streets you race on. The map is not a lobby. Catalyst had the better movement and the worse decision about what the open world was for. An open world that does not host the best verb is an expensive corridor. Put the time trials on the streets people already know, or admit you wanted courses and build courses.
What the sequel got right, and what it still missed
Catalyst repaired most of the complaints people had about the first game, and still did not come together. The one change that explains the gap is the open world.
Mirror's Edge always looked like a game begging for a second attempt. The first-person platforming needed sharpening, the story needed a better delivery, the guns needed to go, and there was not enough to do.
On that list, Mirror's Edge Catalyst scores well. Movement responds more crisply, the cutscenes drop the stylized Flash look, firearms are cut, and the amount of content jumps.
So the trouble sits somewhere else. It sits in the most argued-over change the series ever made: turning the city into a sandbox.
The first game was built as an obstacle course
Levels in the original Mirror's Edge were obstacle courses. One route ran through each space, and the job was to keep momentum alive by chaining parkour across springboards, fences, pipes, walls, and zip lines.
Runner's Vision usually painted that route somewhere ahead of you. Reading it at full speed was the skill, along with spotting a quicker line and landing the jumps, tucks, rolls, wall runs, and hard turns cleanly.
Then the pace would drop and the game turned into a platforming puzzle in first person. You stopped and worked out how to climb a tower or thread a knot of platforms and obstacles.
Catalyst is still good at this, and the reason is telling. Nearly every story mission and a fair number of side missions sit outside the open world, in linear, carefully shaped locations. The skyscrapers, offices, underground facilities, and server hubs are made to be run through once, in one direction.
The city is left holding the jobs nobody else wanted
With the campaign's best missions happening elsewhere, free roaming gets only leftover duties. It ferries you between missions, gives you somewhere to practice parkour, stores collectibles, and hosts side activities.
So the question that decides the whole design is narrow: do those side activities actually need the City of Glass to be a sandbox?
Open worlds earn their keep in two familiar ways. They let missions be approached from many angles, or they let you revisit the same ground until you own it. For a game about speed, the sharpest point of comparison is Burnout Paradise.
Burnout Paradise puts every event through the whole map
Burnout Paradise made the jump from fixed tracks to an open city and kept working, because each of its events leans on the sandbox rather than ignoring it.
Races and marked man challenges throw you across the map with no rule about how you get there. Road rage and stunt run let you hunt takedowns or stack combos in whatever direction you like, with no track edge waiting to end the run.
And because all of it happens in one city, you drive the same roads over and over. Routes, shortcuts, boost refills, and cut corners accumulate in your head without anyone teaching them to you.
Races also end at one of a handful of finish lines, so the approaches to those endpoints become familiar too. The city stops being scenery between events and turns into the thing being tested.
Catalyst chops its open world into slivers
Catalyst does the opposite. Its dashes, fragile deliveries, covert deliveries, and dead drops are point-to-point runs, but they head for arbitrary spots along short routes that are largely decided in advance.
Faster lines exist, and they usually sit just to the side of that one event's path rather than anywhere else in the city.
What you end up with is an open world sliced into small pieces. Drop those events into a menu instead and little would change; they are barely less linear than the races in the first Mirror's Edge, which carved speed runs out of chunks of story levels.
The shortcuts you learn come from grinding a single event, not from knowing the City of Glass.
Why the City of Glass never sinks in
Broad knowledge of this city is hard to build, for three reasons that stack on top of each other.
The first is exposure: main missions happen outside the open world, so you never log meaningful time reading its layout. The second is memory: there are few landmarks strong enough to orient by.
The third matters most. Playing without locking onto the thin red Runner's Vision trail is extremely difficult.
You can dial Runner's Vision down or switch it off, but the layout makes it feel compulsory. The open world keeps squeezing you into linear corridors, hemmed in by towers too tall to climb, buildings too low to land on, and rooftop gaps too wide to cross.
That does reproduce the old obstacle-course flow, and it is dreadful for navigation. It bites hardest between big clusters of buildings, where only a couple of connections actually exist.
A waypoint can sit in plain sight ahead of you while the only real path loops far around the city. If the world is chunked that way for streaming reasons, that is a fair technical explanation, and it still wrecks how you read the space.
The layout blocks the missions it most needs
That geometry is why Catalyst could never run missions in the mold of Burnout Paradise or Crazy Taxi.
Crazy Taxi points a rough arrow at your destination and leaves the rest to your driving and your grasp of the streets. A Mirror's Edge take on that could have been superb: grab a package, carry it across the open world, and land it on time by navigating well and never breaking flow.
It would also have been genuinely different from the campaign. Linear missions test execution; this would test whether you understand the city well enough to hold speed across a route you picked yourself.
Inside a fractured, maze-like City of Glass, that mission collapses into two outcomes: trace the red line, or fall.
Player-built time trials hint at the better game
One feature does need the open world to exist. Players can author their own routes across the City of Glass, and that idea would lose most of its point in a linear game whose best stretches have already been turned into official races.
Burnout Paradise still has the better version of the same instinct. Leaderboards sit on every road, so the competition is passive and never switches off, and you are always being nudged to push harder and learn the map.
Its collectibles keep you moving too. Gates and billboards get smashed at speed. Catalyst tends to halt you instead, to open a fuse box or fiddle with an object.
Even the people handing out events in Burnout are in motion, tearing past and daring you to follow, rather than standing still waiting to be switched on.
Security hubs come closest, then hand you the answer
Catalyst has one mission type that genuinely uses the open world: the security hub side quests.
They ask you to clear the guards, wreck the tower, and then lose a helicopter. As in Burnout Paradise's stunt run events, the escape can go anywhere, along any line that keeps enough speed to stay ahead.
Nothing is laid out in advance and no direction is the correct one, so whatever you know about the City of Glass pays off in avoiding dead ends and drops.
Then the optimal play turns out to be following the red line to the closest safe house, where the chase simply stops. The design walks right up to a strong open-world parkour mission and lets the guidance system finish it for you.
Parkour and open worlds are not the problem
None of this proves Mirror's Edge and open worlds are incompatible. The movement suits free roaming, and an open space could carry mission types that sit alongside the linear and puzzle-shaped campaign rather than competing with it.
What breaks it is the pile of decisions taken together: a fractured layout, a hard lean on Runner's Vision, side quests that behave like short linear slices, and main missions sealed off in bespoke spaces.
Burnout Paradise is the working version. Go open world and the city has to be load-bearing for events, shortcuts, navigation, collectibles, competition, and mastery alike. You find out whether yours is by running the build and driving your own routes through it, not by reading the mission list.
The flashes are there in Catalyst, mostly in player-built time trials and helicopter escapes. Most of the time, though, the City of Glass is not somewhere you learn and exploit. It is a tangled route network that the red line solves on your behalf.
Build it in Flockbay
In the Flockbay app, build a few streets and put the race on those streets, starting from where the player already is. Play the trip to the start as part of the event. If you need a loading screen between the city and the fun, the city is a lobby. Delete the lobby or move the fun.
An AI racing game maker and an arcade racer template are the Burnout lesson. A puzzle platformer template is the course, if you would rather be honest.
Keep reading
- The dotted line ate the trip
- Dread steers and calls it freedom
- Rapture explains itself in the props
- Screenbound makes you play 2D and 3D at the same time
- Build an open map worth driving around with the AI racing game maker
- Make a racing game with AI, starting from how the car feels
An open world owes you more than travel time between missions. Build one your players get better at reading, then play it yourself and check whether knowing the map is what wins.
More notes on designing places players learn