Captive relies heavily on procedural generation. When you start the game, you first have to fly with a spaceship to one of the planets on your star map, land, locate an enemy base, enter, locate a space probe inside the base, destroy the generators and then run like hell. If you don't manage to get out in time, you'll die when the base explodes. The goal of the game is to do this ten more times, allowing you to free yourself.

If you succeed, you get the chance to start over again, and again and again. Hundreds of bases with countless procedurally generated levels, all sprung from the same seed to make sure every player sees the same sets of levels. This is both the game's greatest strength and weakness. Once you understand how the game constructs levels and even puzzles, you'll realise just how bare-bones and repetitive it is.

Playing Captive as a kid, that didn't bother me. What kept me going was seeing something new every base. New monsters, different tile graphics, more weapons, body parts and ingenious tech upgrades. That first run with 11 bases is quite fun and has enough to offer to overcome the simplistic gameplay. But after that it gets tedious.

The sequel, Liberation: Captive 2 (1994), was truly ambitious. Once again in charge of the four droids, you have to investigate a murder cover-up in a futuristic, hostile city. The city is massive – a sprawling open world with shops, libraries, offices, houses, etc. The game also featured fully 3D graphics, a customisable UI and introduced the ability to talk to NPCs, pursuing more peaceful approaches.

Still, just as with the first game, most of it was procedurally generated and, combined with the insanely large city, made for a game easy to admire for its ambition but hard to finish due to sheer size and lack of compelling, handcrafted content.