Gameplay Research & Development at Ubisoft

How I got there:

 

After completing my double degree in Georgia Tech and failed to get a job offer in the US in time for an easy VISA extension, I went back to France and submitted my resume to a few companies, Ubisoft included. I got an offer from 2 of their teams in Paris : the Ghost Recon Team and Gameplay Engineering & Entertainment Analysis (GEEA). I took the one at GEEA because working on a shooter wasn’t extremely appealing and on the the other hand; GEEA was a very exciting prospect.

GEEA used to be a division at Ubisoft Editorial in Paris which worked directly under Serge Hascoët – Ubisoft Chief Creative Director – its task being to experiment daringly on self and mandate-based topics to inspire creative directors and production team worldwide .
I was part of the 3rd ‘wave’ of GEEA members, and we were a bunch of youngsters with little to no production experience, sometimes big egos and always high motor.

What I got to work on:

 

  • LUA Prototyping engine: helmed by Cyril Laguarigue, our internal engine had the ambition to become a prototyping alternative for the heavy, iteration averse engines used in production circa 2008. It glued together Ogre3D, OpenAL, PhysX and any other 3rd party API we were interested in with Lua, the game engine logic being in LUA itself. I worked on the underlying lua component system, making a more efficient ‘hot reload’ system as well as providing a Nodal Editor/Visualiser to document logic and objects structure within a prototype.

 

  • Narrative Structure & Activities:
    • Gameplay Grammar: I analyzed existing games to come up with a sequence based grammar that identified roles for both player and NPCs, their tools and their winning/losing conditions expressed as object/world state plus extra constraints.
    • Sequence Creator: A .NET tool to crowd-source game situations, allowing anybody to describe game sequences either as one shot or as made from a template (that used the framework described above). Multiple outcomes per roles could be defined and each outcome could be linked to another role/sequence, creating a browsable space ala Choose Your Own Adventure.

 

  • 3D Camera (ZCam, Primesense and eventually Natal/Kinect): Yves Guillemot took notice of the 3D Camera before really anybody else and tasked us to make as many toy and game prototypes as we could. I personally did a few
    • Self-slingshot:(inspired by Pain on PS3): player enters the slingshot, cocks it by stepping backward then make sa wide arm gesture to release themselves, they can control their body in air by moving around; landing on elements with different properties to try and catch points.
    • Full body inventory: touch part of your body to retrieve/put back weapon and tools: left shoulder->bow, right shoulder->arrow, right hamstring pistol, left elbow crease-super strength buff, hand on chest->heal, Make a specific full body pose to turn into a robot.
    • Rollerblade: player balance their arm in a swing motion to go forward and could use their feet to interact with a rail (but with little effect).
    • Match character in poster: players tries to mimic in space, the posture of an actor in a poster. As they do, a Video RGB gets more revealed as they match the character position; a terrible mashup being screenshots for History when they’re close enough.
    • A Voice in the West: western/mafia themed FPS with 2 weapons selected depending on hand position : a pistol in each when they are quite apart and a two handed Thompson 1928 when they are fairly together. Key point: the player use their voice to trigger firing the weapons. So you can go ‘pewpewpew’ and it spouts out bullets.
    • Body armor: as you enter a certain distance of the camera and stay still, robotic arms put armor elements all over your 3d avatar
    • Soccer penalty/kick (with Olivier Maurel des Vallons): trying to use good recognition of velocity to make a very good penalty/kick game
    • Soccer Golf: a game of golf except players use their arm to turn the camera around and a kick to hit the ball
    • And the whole team collaboated on an FPS adventure for a behind-closed-door demo at E3 2009 – using assets from pre-reboot I Am Alive.
  • A lot of interrupted attempts at memory based AI and collaborative storytelling between parents/kids

Take-aways:

 

  • Prototyping for new input is fun: I entered the game industry wanting to work on storytelling structures and interesting player tools. When the focus of the team changed to 3D Cam, I discovered that working with novel input methods beyond was also very rewarding; and that I was also pretty interested in lowering the barrier to entry that gamepad literacy can represent.
  • Too much focus on engineering, not enough on experience: I failed to deliver experiences as compelling as I wanted because I focused too much on the systems and writing reusable code than on defining an actual target experience and shortcutting my way to that.
  • Learning to kill your work: despite this, moving from topic to topic and from technology to technology, I learned how to ride the high of an idea, prototype it and then leave it where it stood, without second guessing. This proved very useful to future projects.
  • Project management is always needed: our work was good but we lacked discipline in terms of scheduling and deliverables, which meant to me, like we were making less out of the opportunities than we could and weren’t smart with the resources that we had.

Tools

LUA, C++, C#, a dash of Unreal Engine 3