Mirror's Edge aka skeleton under the camera

A long time ago, I posted a blog called FPS GAME MAKERS ARE ASLEEP, it was about the PS3 launch title Resistance (Fall of Man's lack of a skeleton under their supposedly-first-person camera). Two years later, if we can believe the pre-press hype, EA DICE a Swedish operation with very classy-looking offices and an even classier looking game called "Mirror's Edge" have pulled out their FPS finger and seem to have thrown a skeletal model under their forthcoming game. I love the game's art style using minimalist architecture, key colour and LOADS of texture resolution but how integrated into the game is their 'skeleton'? Does the player move the head of his 'actor' and the body follows, degree by degree with relevant constraint/adjustment down the hierarchic chain to the feet? Or is it just a clever marketing gimmick?

You can now share with me this vision running in the game engine for the very first time.



The general concensus among my gaming colleagues and I is this is 'twitch gaming', with an instant response from a button press - whereas there was loads of potential in the investigation of other methods of world navigation. I'd personally liked to have seen less of the fancy Sonic speedlines and more effort expended in the interface, e.g. a head control interface where the player can a) feel the weight of the skeleton under his fps camera and b) pre-mark hand/foot holds as he traverses the assault course that is the city.

LOOK-AHEAD INTERFACE aka FREE-FORM RAILS:
if you were one of the few hundred readers who tuned into this blog yesterday, you're gonna miss this. Maybe your friends will tell you about it and you'll come back. I was thinking that this heads-up interface might be a little like the way one plays REZ or PANZER DRAGOON. And it does, except for one very important feature. The rails are free-form.

The game already knows your physical abilities/limitations, it knows if you can kick open the first gate, it knows how high you can currently jump and whether you'd need to put a foot on a wall to scale it, it knows if you have enough stamina/speed to run along the yellow wall and whether you'll clear the final jump in the demo.

And it shows you. By using the red-foot, red-hand, red-action iconography of the design itself projected onto the game world.

Let's give it a try. Stand on a gravelly roof top and look around, it's as simple as moving your mouse/thumbstick. What does the interface show you? Nothing. You can affect nothing from where you are. Why? You're not moving. Push forward to walk/ jog/run suddenly the interface shows you that AT THIS SPEED, (there's a line of ghostly red feet from where you are in the direction that you're looking) you have a slight chance of making that jump from the edge of the building to the building across the way. Look around and you'll see that the transparent footsteps leading to the edge get more opaque as you 'look' towards a special jumping area. Slow down to a walk and the footsteps to the jump area fade to nothing.

Basically it works by 'projecting ahead' and applying your current stats/speed/direction to the task in situ, at the location you're looking towards.

Take, for example, that red pipe Faith lands on in the demo at such an oblique angle late in the demo. No way! It wouldn't even turn red until she was pointing in the correct head-look + movement vector. There're so many reasons why no real parkourist would even attempt that pipe walk like that. A real parcourist would scrape to a stop before attempting such a risky maneouvre. Yes, stopping, that can be used to enhance this form of gameplay. They'd line themselves up properly with the pipe THEN accelerate across it once they'd established their momentum.

There should be other jumps where you'd have to stop at the edge (to highlight the interface) to get the best grip with two feet, jumping off one foot would lead to a death drop as the rolling friction far exceeded the standing friction of your (seemingly 100% friction) shoe - that's what 'real' parkourists would do. I know, I've watched them do it. Imagine trying to run across a metal pipe, having just scrambled across a gravelly rooftop. Imagine gripping onto a swing rope with wet hands. These are issues that have already been addressed in driving simulators but will also be issues in future skeletal games.

LOOK AHEAD is a dead simple interface that shows you the likelihood of success of the current rail you're creating. It allows you to PUZZLE SOLVE on the move.

Back to the demo, I don't like the way she shoots that handgun through her own thigh, oh, and a reflection of the runner in reflective glass would have been a great idea. Not happy to be stating the obvious. A representative of EA Dice yesterday refused to talk about the issues brought up in this post which probably means the 'press up now' and 'press down now' game control model is one that isn't broken and doesn't need fixing (yet) as they put it.