The RPG Codex managed to snag an interview with Warren Spector recently, and asked him a number of in-depth questions about his time at Origin Systems (with a particular focus on his work on the Ultima series). Spector also too the time to comment on his game design philosophy, noting an interesting dissonance between his own perceptions and those of his audience:

RPG Codex: Looking back at your career, one can observe two Warren Spectors – the early 90s storyteller Spector of Origin, who took the setting created by Richard Garriott and helped flesh it out with detail and amazing stories, and the late 90s designer Spector of Looking Glass and Ion Storm, famous for a philosophy of game design emphasizing player choice and immersive simulation. That’s an interesting transformation because if we look at Martian Dreams or Serpent Isle, those games were actually more linear and “railroaded” than Ultima VI or VII (developed under Garriott’s supervision) and more focused on storytelling at the expense of some of the player freedom characteristic of the mainline Ultimas. How do you explain this shift?

Warren Spector: Interestingly, I’ve never really seen that much of a shift. I’ve always been interested in interactive storytelling of a sort – but a very specific sort. I’m not much interested in branching tree storylines, where players get to decide only what branch of a tree they choose to follow. I’ve always been more interested in offering players a single path through a game but giving them a variety of tools with which they can solve problems along that path. So the high level story arc belongs to me and the team – and players can’t affect it – while the minute to minute of the story belongs to each player and they have total control over it. I thought that’s what my teams and I were doing in Martian Dreams and Serpent Isle, as much as in Deus Ex and other later games. If that isn’t the case, or isn’t apparent, I guess we didn’t do such a good job, did we?…?

Naturally, Serpent Isle comes up for discussion at several points:

From what we know, Serpent Isle was originally conceived by Jeff George as a pirate-themed game that wasn’t supposed to be a part of the core Ultima series. Do you recall anything about George’s original design? How challenging was it for you to take over the project after George had left the company, and what influence did you have on turning the game into a more traditional Ultima title?

If memory serves, Serpent Isle was always meant to be an Ultima universe game, but it was a pirate adventure at first. Honestly, it wasn’t really making the kind of progress, creatively, I hoped it would – who wouldn’t want to play a pirate RPG, right? But when Jeff left, I started working with a fellow named Bill Armintrout on the creative and it became a direct sequel to Ultima VII. The game was made entirely in my unit, so I had a lot to say about the decision and the way the creative [process] unfolded! I don’t recall the transition being terribly difficult but that project was by far the biggest I’d worked on at that time and the team grew along with the game’s scope. None of us had a clue how to manage a team that size which led to the worst crunch mode I’ve personally experienced, to this day. That team worked incredibly hard… At the end of the day, the game ran over 100 hours for most players. It even took QA nearly a full 24-hour day to play through. We actually didn’t get our first cheat-free playthrough until the day before we signed off and all of us were terrified we might be shipping the buggiest game of our career. Didn’t turn out that way, luckily!

Serpent Isle was your last “old school” party-based RPG. In the late 90s, while at Looking Glass, you developed a design philosophy emphasizing player choice, and then you continued with that approach in Deus Ex. During the same period, however, Black Isle Studios was developing its own signature gameplay style which also emphasized player choice, albeit in a different way – games like Tim Cain’s Fallout or Chris Avellone’s Planescape: Torment were traditional party-based CRPGs with an isometric perspective, deep dialogue trees, etc. One could imagine that, had you continued making games like Serpent Isle, they would have turned out a lot like those titles. Do you ever regret not having been able to pursue that path? Do you think you could have married the form of Serpent Isle with the essence of Deus Ex, so to speak?

Interesting question… I think I could have married Serpent Isle’s party basis with DX, but I wouldn’t have done it with dialogue trees and traditional RPG tropes. The key thing about games like Underworld and System Shock and Deus Ex and, yes, even Disney Epic Mickey, is that they don’t rely as much on scripting (dialogue or interaction scripting), as on simulation. I think it’d be possible to make an isometric, party-based game that offers all the player choice and consequence stuff, for sure. I’ve often thought about giving that a try. You never know – it just might happen some day!

The interesting thing to me, though, is that you really see a radical difference between the philosophy underlying Serpent Isle and the DX philosophy. I see them both as being on the same evolutionary path.

There’s a lot more to the interview — the Worlds of Ultima games, Arthurian Legends, Deus Ex, and more — so I’ll suggest that you all just click on over and read the whole thing. One last observation from Spector, however, is good enough to merit use as a closing quote here:

…when I started, “going gold” didn’t mean “shipping a game” – it meant you sold 100,000 copies. And when you did that, you went and bought yourself a Ferrari. By the mid-90s, 100,000 copies was a dismal failure.

This comports with Richard Garriott’s observations in his “three eras of gaming” talks, more or less, except that he tends to denominate games in the ones/tens/hundreds of millions. The basic message is the same: with each subsequent era, the sales threshold and reach needed to deem a game a success grows…perhaps exponentially. One wonders if there are limits to that?