February 27, 2007

String of Pearls in the Sandbox

by Andrew Stern · , 7:12 pm

Among mainstream game developers, Warren Spector is one of those putting significant effort into building more sophisticated interactive stories. To the list of exciting GDC talks next week, we need to add Spector’s newly announced talk “The Future of Storytelling In Next-Generation Game Development” (as well as Ernest Adams’ “Rethinking Challenges in Games and Stories” for that matter). Recall Spector was a participant on a GDC panel on interactive story I moderated two years ago, although his attitude was somewhat dour at the time.

These days, after founding and reportly getting funding for his new studio Junction Point, Spector seems more optimistic. He was interviewed in this month’s print magazine Game Developer (owned by CMP, the same company that runs GDC and Gamasutra.com). Since few of you may get the magazine, I’ll type in a few highlights from the article:

There is a middle ground [between sandbox games like The Sims, and roller coaster rides like Half Life], and I don’t think it involves the choose-your-own-adventure approach. … The key for me is creating linked sandboxes and letting players explore those little narrative chunks on their own. [As the game developer] I’ll determine why it’s important that you get through a door, but how you get through it, what happens, and whether you kill, talk to, or ignore everyone on the other side belongs to you. That concept of sharing authorship is where the sweet spot of game narrative is… it’s a hybrid of a linear string of pearls game structure and a sandbox approach.

About Spector’s concept of an “RPG which deals with a single city block and all the people and experiences involved in that”:

I think there are huge design problems that need to be solved there. … It would be a real interesting challenge. It would be extremely risky… I think it could be done. I think it’s a mistake to look for gigantic, revolutionary steps — or plan for them. There’s probably somebody in a garage somewhere that has all that working and will change the world completely. … At the end of my career… I want to be able to look back and see that every game I did was some logical, evolutionary step towards some clearly defined goal.

The interview concludes with:

The reason very few people do what we’re doing at Junction Point, and what we did at Ion Storm [with Deus Ex], Looking Glass [with Thief, etc.], and Origin [with Ultima Underworld etc.], is because it’s really hard to do, and it’s a lot of extra work that’s satisfying to developers and players who get it, but it’s not necessarily something that immediately increases sales, and it certainly costs money. It’s more expensive to make games that have choice and consequence than it is to make a game that has an illusion of choice and one scripted consequence.

I like a lot what Spector has to say, and very much look forward to Junction Point’s games. Admittedly we’re among those garage developers that believe that more-than-evolutionary steps may be required to make faster progress here (but not one of the ones who has it all figured out ;-). My preferred game design metaphor is not to have the player hop from little sandbox to little sandbox, but to have the player and system collaborate to make sandcastles. More literally, a system where “the rules of the simulation [sandbox] are regularly being updated in an attempt to give the player a well-formed overall experience with unity, efficiency and pacing” (see pages 4-6 of our GDC03 pdf).

Also worth linking back to is my debate with Eyejinx over his essay “Pissing in the Sandbox”.