GDC 2005: Clint Hocking on Narrative

Here is another in our series of GDC session summaries.

Deconstructing Sam: Narrative in Splinter Cell
Clint Hocking

Clint’s talk was a great overview of how he sees storytelling in games. He broke things down using a lot of analysis that I have heard before, but the difference here is that he has put much of Marc LeBlanc and Jesse Schell’s theory into practice in Splinter Cell: Chaos Theory. It will be very hard to summarize his entire talk so I’m going to concentrate on the part where he discusses his variant of Schell’s Theory that Story and Gameplay are the same thing. Clint explained that games have four types of narrative: Emergent, Embedded, Strong Procedural and Weak Procedural.

Emergent narrative comes from the difference in what a player tells himself will happen and what actually happens. For example, a player observes an NPC’s behavior and creates an expectation of what will happen when he attempts to subdue that NPC. But if what actually happens is different in someway (that isn’t perceived as cheating) then the player’s experience will be richer.

Embedded narrative is anything that is authored. This is anything written, modeled, scripted or hard-coded. His example was blood on the walls in Doom 3. He also described how in SC:CT they worked from the assumption that the ‘who’ or ‘what’ of a scenario was not that important, but concentrated on the ‘how’ and ‘why.’

The two kinds of procedural narrative are closely related. Weak procedural is when a game system generates changes to the flow of the game in the micro, in order to increase dramatic tension. This can be as simple as spawning a few extra guys if the player is cruising through a level. Strong procedural is when game systems generate responses to possible player actions in order to change the narrative, by adding complications and strengthening the rising action

This is a small slice of his whole talk, but I thought the definitions and examples he uses were worth relaying.