I crept up, my feet slowly taking me across the stone floor when a lich emerged from around the corner. I checked my gear, said silent prayer to the save game gods, and prepared myself for a grueling battle. I struck with my blade, the magic along its edge cutting into the enchanted bones of the dead thing in front of me. It hit back with cold, knifing ice tearing into my life’s blood. Backing off, it followed me. I could measure what time I had left by how many steps it was taking to catch up with me.
And then it got stuck on a pillar allowing me to rain blow after blow on its dessicated body until I killed it. Me, a low level nothing, had managed to slay one of the mightiest creatures in AD&D after it couldn’t find its way around a support beam.
That was Descent to Undermountain, Interplay’s attempt to bring AD&D to the world of 3D in 1997’s holiday season. It attempted to reach for the same magic that Looking Glass had cast for Origin with Ultima Underworld five years earlier only to become one of the worst CRPGs in the memory of many.
It had the making of an epic. It was going to be based on TSR’s popular Forgotten Realms setting, an AD&D wonderland of busy nations, ancient intrigues, and high fantasy. But not just that, it was going to take place in what was boasted as the Realms’ largest dungeon: Undermountain. A maze of tunnels, tombs, forgotten chambers, and caverns woven together by the mad archmage, Halaster Blackcloak. Players would be called to plumb its depths to seek out the Sword of Lolth and save the city of Waterdeep from a terrible fate.
It was also frustratingly buggy and strangely inconsistent with a lot of things, such as being able to kill a lich at the low level that I was at. I was somewhere below 10th level which, in AD&D terms, still meant that I should have a ways to go before I could even spit shine a lich’s boots much less kill one with ease. Especially when I was a hack ‘n slash army of one since this was a solo journey sans party. Performance was on my box made moving through the strangely scrunched corridors and past oversized furniture of Undermountain something of a muddy slog and this was a better box than the one that ran Ultima Underworld and Bethesda’s Arena without a hitch. Real-time fighting at slo-mo speeds created no end of annoyances.
I managed to get through it and the story had promise, but this vaunted super dungeon didn’t feel very super, not when I went back to Ultima Underworld and its sequel. What went wrong? It wouldn’t be until a few years later that bits and pieces would finally trickle out.
Chris Avellone is a name that many CRPG fans immediately recognize, although his career at Interplay didn’t start off with a bang when he was called to helm Descent to Undermountain halfway through its development. He immediately realized that there were major problems with it, most notably the shoehorning of the Descent engine as a 3D solution, and wanted out in order to take up Tim Cain’s offer to join the Fallout team.
He wasn’t transferred, and like anyone who has a job to do, did what he could. Even today, he looks upon it with some regret, but no one can question how well he’s rebounded since then thanks to his work on Fallout 2 and Planescape.
The first problem which Avellone has pointed to was the Descent engine. This was the same engine that wowed players with its zero-g, shoot ’em up action in subterranean mazes against floating robots throughout the Descent series. Adapting it to more terrestrial pursuits proved to be the cause of most of its issues such as figuring out how to work in gravity. It would be as if Lucasarts decided to ditch the engine for Dark Forces and adapt Tie Fighter’s instead because it was still basically a 3D engine.
There were hints that this wanted to be just like Ultima’s adventure through the Stygian Abyss with NPCs, quests, treasure, and plenty of monsters to stab. The AD&D system’s conventions were peppered throughout the gameplay in a limited sense – you could only pick the basic fighter, thief, mage, and priest classes, for example. A number of different races were still available along with spells and other class-specific bonuses, but it was a far cry from the same depth that SSI’s Gold Box series offered gamers in the late 80s and early 90s. Perhaps the only reason I kept going with the game was to find out how the story would actually end.
The ad shamelessly banks on the hope that fans of Descent will turn in their flight badges and dive into a CRPG making that a risky approach by itself, though it didn’t kid about rendering everything in 3D such as the monsters. That in itself was a technical achievement that neither Ultima Underworlds had done. But the price paid for that was too high despite what must have been a heroic effort by the team behind the game to make the solution work. Complaints about Skyrim’s or Oblivion’s bugs ring hollow when faced with something like this.
Descent to Undermountain wasn’t as bad as Maabus in my experience, and there was some promise buried beneath the mess. But when other games had already done the same thing, and with far better results, it also stands as another example of where leveraging the most cutting edge technology for crispy visuals is no replacement for having gameplay that simply works.