Redesign: Mirror's Edge

When you give a game to testers, you have to be critical about the feedback you receive. They aren’t programmers, nor game designers (probably). So what they perceive to be, for example, a problem in one area is actually a side effect of another problem elsewhere.

Sometimes, I like to look at a game I enjoyed, but thought could be better, and carefully analyse why it didn’t work. One of my favourite games to do this with is Mirror’s Edge, because so many people* seemed to believe the controls were a problem, when I think the real issue lies elsewhere.

*No research, I just remember some reviews and people talking about it.

Level Design
I seem to remember that someone more specifically said that the controls did not fit the level design. To me, this sounds as bad as saying that we’ll make the graphics, and build the game around it. Or we’ll write the story and make the gameplay fit that. This is ok, because they aren’t game designers, but in reality, the levels are made to fit the controls. When it becomes awkward to move around a level, the level is often at fault. Sure, the controls can be poorly made and still finalized, but if you are late enough in production that you are making the levels that will appear in the game, you don’t want to mess with the controls, because that will mean redoing many of the maps, which, if you have a deadline, is a great way to get sub-par level design. Since the levels must support the controls, you must be able to detect weaknesses and try to avoid them, if possible. If it’s not possible, then the controls should be adjusted earlier.

If you want to be a level designer, and you haven’t played Half-Life 2, go play it. The level design is superb. Particularly, the skill of their designer(s?) to draw your attention to where they want you to look is admirable. For example, they want you to see an airborne vehicle crash-land, so they put an enemy with a weak weapon in a spot that he can’t reach you from, which happens to be the direction the vehicle crashes from. Sure, he’s no danger, but he gets your attention. In rooms with obstacles, you can often see, and recognise (important bit!), the goal as soon as you enter, even if you don’t yet know there’s a puzzle in the way.

Mirror’s Edge, on the other hand, has very little skill at this. I distinctly remember a room which you fall into, which is around 4x4 meters, with a little passage to a 1x1m room, and even in this tiny area, I remained stuck for almost twenty minutes. I played both the PC and X360 versions of the game, and this room frustrated me both times. You can tell a room is badly designed when you enter it again a few days after solving it the first time, and you get stuck once again. Tip to the ME level designer: nobody ever looks up without a prompt or a hint! There were many other instances that failed to guide the player, and caused me to get stuck on both playthroughs. I suspect they were more concerned with the pretty than the functional.

Controls
That all said, there were a couple of issues with the controls, but only one was really bad at all. That one isn’t even really anything to do with the controls themselves. It’s the problem of precision jumping when you can’t see (or otherwise detect) your feet. Precision jumping is hard enough in third-person games, but in first-person, it’s a nightmare. When you game is based around parkour and freerunning (two different things, by the way, not many people seem to realise) you want to solve this problem effectively. I present two ideas off the top of my head:

1) Use Zelda style jumping. In the 3D Zelda games, jumping was automatic. Just run at an edge and Link would jump at the best moment to reach the furthest. This matches, more or less, how real people jump from an edge. You run, so that you touch (or nearly touch) the edge with your foot, then leap. Many of the mental processes are automatic. To fit with the ME control scheme, let’s say that sliding (crouch button) will let you drop without jumping, and going slowly will let you hang from the edge. The jump button will still function for when there is no edge. Obviously testing is in order before finalization.

2) Add some visual indicator to the HUD, showing how close you are to the edge. This one is a little distasteful, because they did such a good job of removing all other HUD elements (except the timer in time trial, and I found the centre dot annoying, even as I recognized that some people may need it. Glad that was optional). Either an indicator showing general “edge-closeness” or an overlayed, rotating minimap-like thing (except expanded, so that it only appears when you are near an edge). This would allow you an additional top-down view, showing you what your first-person view could not, especially if the centre dot is given the dual use of showing your position too.

Actually, the problem was slightly removed for the PC version, as with a mouse, you can much more easily quickly look down. It’s not ideal, but I had more fun with the PC version in general. In fact, the WASD style was very good for this game, considering that you did not need the control flexibility of a control-stick (forwards to go, and that’s mostly it). The mouse also gave more flexibility than a control stick for something that required it. (I have a rant about how silly direct porting from WASD+mouse to two control-sticks is, but that’s for another time)

Another problem was that, trying to look around when hanging from something was too slow with the X360 control-stick. This can easily be solved by making the camera direction map to the control stick directly when hanging. Actually telling the player you can use the rotate-180 button to look around could also help, because some people didn’t realise this.

Finally, ME could have taken a cue from Trackmania in regards to failure. There was a dedicated button for restarting, and it worked instantly. I imagine adding that feature to ME would take a bit of work; loading may be an issue, and every other object would have to reset, but a good result would go a long way to improving the experience, as anyone who fell from a great height could tell (falling, falling, falling, falling, thud, die, slow fade out, loading screen, fade in, is a great way to break the pace of a game, especially one built around a feeling of flow).

Combat
I wonder how many people would have preferred the game without that. It wasn’t actually bad; the system itself made sense, as the protagonist wasn’t good at fighting. The problem was when they forced it upon you. Giving someone with no weapons (and little skill with them anyway) a fight against multiple shotguns, is a great way to frustrate them. It should have been an option at all times.

Actually, I suspect the only reason why there was combat at all is the same reason they put boss fights in stealth games. They (whoever “they” are in this case) might have thought that the game would not sell if there wasn’t combat. Maybe the poor, mindless gamers won’t be able to cope if a game is a little bit different from normal. Either that, or it was there entirely because of the story. I know that many people will disagree with me, but I think a story should be created to fit the gameplay, not the other way around. You’re selling interactivity, not story. The point of a game is the interactive experience, and the story should support that. If not, you’ll be making bad design decisions because otherwise the story won’t make sense, oh no!

Commercial soundtrack
The final problem I found with Mirror’s Edge was actually with the soundtrack you can buy of it. I wasn’t going to mention this, but I’m here now. If people like the music of a game, and you release a soundtrack, make sure the soundtrack is of the music in the game! You’d think this was simple, but no. For example, I was really disappointed with the Rez one, because my favourite music had been remixed, and wasn’t as good. ME goes worse. The game had 11 remixes of the same tune, which was actually fine because, while not vastly different, they at least appeared different enough (or more specifically, the game had only the parts that were different). The soundtrack had the entire version of each remix, revealing that they were not, in fact, different enough. That starts to get repetitive on its own, but every song had exactly the same singing! And you get so sick of it! The soundtrack was not worth it.

In the end, though, Mirror’s Edge was a good game with lots of potential. It just made a few too many mistakes that weren’t picked up on or solved. I kinda hope there’s a sequel, and that the controls remain largely the same.

Edit: What a coincidence. A couple of days after writing this post, I won Mirror's Edge for PC. It occurred to me that, now that I have my own copy, maybe I can try to mod in my ideas or something. Not that I've ever modded before. Must investigate. Anyway, I tried out some user-made levels, and it just hit home that, no matter how poorly I think the official levels seemed to be designed, they still show a level of expertise that is hidden by the fact that all the levels were made by the same people. The fan-made levels bring forth the aforementioned expertise in comparison, as all of them are exercises in frustration. Common problems include tiny platforms that are hard to stay on because of momentum, situations in which one fall will send you far back to the beginning (exasperated by the fact that most don't have checkpoints), and a general failure to direct the player as to where to go next. That last one wouldn't be a problem if the levels weren't linear, but they are.

Speaking of which, one thing I find odd about Mirror's Edge is that it is so linear for something based on parkour, which is so non-linear. The actual destination is often far less important in real life. Storywise, it makes sense, though. Just like how Half-Life 2 wouldn't be better non-linear. I saw a vote on this forum on what should be in ME2, and "Free Roam" was at the top. Which is odd, because nobody in the community is making a free roam level... Must be difficult to do. If you don't want your level to be boring, you need some kind of challenge or goal, but that defeats the purpose of free-roam in a game about movement. Unless the level is designed with multiplayer in mind... Sorry, I'm just think out loud now...