In praise of a BBC Micro classic: Forever playing in Exile

Once when I was small I went to the shops to buy an explosive - a tiny microdet designed for stage pyrotechnics. I've been fortunate that I've never really had blood on my hands when playing with fire, but that day I did get a lot of raspberry jam on them.

A school friend held a balloon open while I crammed in as much jam as possible before knotting it sealed. Then we pierced a small hole in a puppet's head and inserted the microdet followed by the balloon of clotted blood-like substance. I hooked the wires up to a 9-volt battery behind the camcorder, which was masking-taped to a makeshift dolly (a pram on rails). In exquisite anticipation I barked "Action!" But things didn't quite go to plan.

At that age, life only made sense when you were having fun. The freedom to play was why blood pumped through your veins. It was necessary to experiment and push boundaries, to test and meddle with the laws of physics. Teasing spiders, climbing trees, disturbing bee's nests, throwing snowballs at passing cars and building follies to set fire to. There was always a thrill to be had on the edge of mischief. That was instinctive, that... was boyhood.

But a back garden wasn't the only playground on offer in the mid to late 80s. Computer games were fast becoming portals into entire worlds with their own laws and abstracted physics. I was 11 years old and I was attempting to dock a rotating space station. Not Starbuck or Skywalker... me! I could barrel-roll a Spitfire, pilot fragile pods through planetary caves or hack into and fly remote-controlled robots. Exploring those digital playgrounds was my generation's undiscovered country and the thrills were intoxicating. Then at some point in those life-shaping years I found myself spellbound by an unusual game titled Exile, for the BBC Micro. And to this day part of me still hasn't been able to leave.

In 1988 games had to run in only 32k of memory. My phone now has 60,000 times that amount. And yet... there I was 30 years ago exploring a vast underground world simulating realistic physics with its own ecosystem, all running inside 32k of RAM. The wildlife and machines all had their own abilities and behaviours and even emitted digitised speech if you had the sideways RAM. This was absolute heaven to me, but it wasn't until I started making my own games that I came to fully appreciate how unlikely it was that all this ever came together in the first place.

Somehow the developers created a responsive 2D scrolling platformer with pixel-perfect collision, that worked in an engine where objects had properties and a mass effected by gravity, inertia, shock-waves and the elements earth, wind, fire and water. These days it's hard to appreciate now that physics engines are commonplace, but this was relatively new back then and Exile went all in to unprecedented lengths.

This was a new blend of platform game with true physical principles giving birth to a peculiar quality that sparked the imagination like nothing before - physical emergent narrative. It offered the player the ability to experiment and discover things outside the remit of whatever challenges the developer laid out for them. This alchemy for real tangible depth to interactive worlds can still prove painfully elusive for developers today. How can one offer the flexibility to experiment while maintaining a balanced ecosystem in which delicate puzzles have been woven? Maybe this is why it took so long for games like Minecraft and Disney Infinity to happen.

But even these days, how often do players find themselves trying to disarm a primed grenade that has escaped their grasp in a circular wind tunnel? How do you keep a flask from spilling water as you're caked by jetpack-clogging mushrooms thrown by cheeky imps? What do you do with a killer bee you've caught from its nest but stings while you flip through your inventory? What happens if you hold a frightened pink ball of fluff under red drops of acid jam? These just weren't the questions gamers were used to pondering. Equally important, it turned out nor were they questions a developer needed to contrive.

The intelligence required for path-finding and strategy alone must have been a challenge in itself, but the indignance of that Darlek-esque sentinel as I landed politely on its head, the audacity of the villain Triax teleporting in and out to shoot me in the back, that endearing desperation whenever Fluffy clung to me for dear life... Was any of that real or did I imagine it? Advanced AI has often proved a poor investment in games. It can be so very clever but the bottom line is: if it's not noticed, it's completely wasted. The AI in Exile probably doesn't compare to what exists in today's games, but again... this isn't what's important. The secret is frequently staring right at us, through the eyes of a great movie star or even a wooden actor who knows when to exploit that unfortunate quality. Explicit emotional performances or Shakespearean monologues aren't necessary to immerse the audience in the mind of the character. Many actors have made the point that the power of ambiguity can be far more powerful. If there's drama in the situation, the audience can do all that leg work for the actor, who only has to project the illusion of thought. Artificial intelligence in entertainment is as much about anthropomorphism as it is about explicit communication or action. If it's done well it can be concocted in the eye of the beholder, emerging from situation, emerging from conflict, as an emergent... narrative.

Exile demonstrated a strong sense for crafting these natural unpredictable behaviour patterns, giving the illusion of a deeper intelligence and appearing to show changes in mood or temperament. Even if that sometimes just meant knowing when to hold still for a few seconds and do absolutely nothing, like when you give a spider a little poke to watch it play dead. How can something so motionless be so captivating? The AI was superb, but so was this pseudo AI.

For many reasons Exile was the game that made the biggest impression on me, and it wasn't long before I hooked up with school friend Chris Mullender to make our own game for the Amiga 500. What started out as a simple platformer inspired by Giana Sisters soon ballooned into our own sprawling world of bizarre creatures that followed their own laws and physical abilities. We realised then that we had both been heavily inspired by the land of Exile, even down to logistics such as memory management. It was literally impossible to store a map of that size into 32k. Chris' research uncovered how the map was constructed from selected procedural tile sets. This blew my mind, it felt like the big bang in reverse. Also, I remembered seeing the game played on an even less powerful machine, an Acorn Electron and noticed a large portion of the screen filled with corrupt graphics. It turned out this was no bug, but a technique that harnessed the screen buffer to store data. Genius! So we stored our game's hidden cave map data in the island's negative space - the sky. This instantly halved the size of our map data.

We learnt a lot regarding how so much could be put into so little space. It felt only natural to write a fan letter to the developers of Exile, and I was thrilled to receive a handwritten reply from Peter Irvin. However, it was in that letter I learned the tragic news of the death of his co-developer Jeremy Smith (who had previously created the much loved gravity game Thrust). Also sad was the speculation of what magic such a partnership would have gone on to create next.

In a strange mirroring of fate, a sequence of events followed that saw us releasing our game Odyssey under the same publisher as the Amiga version of Exile. My partner was hired to code for Peter's following projects and by the end of the next decade I had lost Chris, my co-creator and closest friend to a sudden illness. I was most touched when Peter reached out to console me, having been through the very same thing with Jeremy.

When I look back on the development of Odyssey, before it entered that tough phase to bring it to completion, I remember the joyful fleshing out on whim and self indulgence. We drew on what excited us and made us laugh. I ate lots of biscuits and Chris drank lots of tea. We were making it because we enjoyed the process. Who knew that making a game could be as much fun as playing one?

When the red light blinked on the camcorder, I dollied towards the puppet and triggered the microdet. But the jam didn't explode. There was a loud pop and sparks flew from its head but the balloon was so tightly packed that it launched intact, remaining lodged and bulging out of its forehead. We were mucking about... and we mucked it up, but it was hilarious fun. Much like the time we filmed a firework backfiring a cardboard bazooka, singeing the armpit of my brother's favourite teddy bear. Or when a friend painted his face silver and climbed into his mum's washing machine, snapping the door off its hinges. We weren't playing to win, we were playing to play.

Now I find myself in the role of a responsible parent, supplying the back garden and fencing in the boundaries for my own kids to push. But like a lot of games developers still under the spell, I'm really just trying to recreate the playgrounds of my own childhood. If I'm still mucking around at middle age where things make only less sense, then it looks like I'll remain forever in exile, playing happily with grenades, pink balls of fluff and raspberry preserves.