Sunday, March 11, 2007

"Three Layers of Immersion"

I recently talked here about swamping and all sorts of other geeky things. Basically, it leads to this post.

On these forums someone posted a theory about "three layers of immersion" or "story". Unfortunately, my search-fu is weak as hell. It's one of the major posters, but I can't remember which one. I think he even wrote a PDF about it, but I never got to read it, so I'm running on some mentions he made a few months ago.

( I bow to Eric Poulton's search fu. It was "Arkaeyn", in this thread. )

It got me curious, so I've done some running with it - to the extent of testing it and brutally adapting it to my own greedy purposes. So, remembering that someone else came up with it (or something like it) originally, here's the way it runs:

You can split the experience of a game into three tiers. The lowest tier would be immediate experience - shooting the Nazi, buying a potion, leaping a pit. The top tier would be the long-term arcs - avenging your father, saving the planet, getting the girl, whatever.

The middle experience would be short term arcs: "Defend this one city", "get through the alien maze", "get the sword of Good Thoughts".

Obviously, the lines can get very blurry: is a four-hour arc in a twenty-hour game a top or middle tier? So, the distinction I came up with is this:

Layer 1: Moment to moment direct play.

Layer 2: Situations that direct the player's overall progression through first-layer situations.

Layer 3: Situations that direct layer 2 situations.

"Wow, Craig's doing more worthless theorizing."

Well, any theory should be useful, so I poked around and discovered that, yes, this theory can be very useful, especially in how it relates to LARPs, open-ended one player games, and multiplayer games of all varieties.

Which part? The second layer, of course.

The second layer controls swamping.

See, if your second layer is too loose, your content begins to melt together. All content will be judged simply as to how effective it is in moment to moment play (layer 1). If it can't be judged that way, it isn't judged at all. The second layer's purpose is to assign a specific value to all the things that have minimal or counter-purposes play value. Examples of this might be whether to kill or save villagers, whether to march on Rome or France, etc. It is also there to limit - or at least heavily tint - which content is capable of being judged, either by restricting access or making something more or less efficient than simple gameplay indicates. (For example: enemies weak against lightning, or lots of railgun ammo.)

All games have this second tier, but the second tier is a very muddy system as it stands. I'm sure there's a better way to think about it, but as it stands, level design, drops, subquests, primary missions, and many other kinds of play are various kinds of second-tier elements.

"Loose" second tier elements are level design, enemy drops, and so forth. Things which assign local values but don't provide real revaluation or direction. "Tight" second tier elements would be things like quests.

MMORPGs show this. WoW, for example, thrives on quests. A lot of its success relates to the fact that players take many quests and submit to their directions. This actually isn't as tight-fisted as it sounds, because it's not really an infrastructure for second tier elements. Instead of the game forcing players into certain activities, it's a loose web that allows the players to catch themselves. And if you don't like that kind of thing, you won't much like WoW.

Normally, tabletop RPGs have excessively tight second-tier elements. The GM says, "The old guy says to kill the monster in the labyrinth!" and the players say, "well, okay, not like we had anything better to do..."

I'm running a Star-Warsish game right now where I essentially did away with second tier elements. It's not the only reason I'm running the game (or the only thing I'm testing), but it fit in well. I was betting that the players would invent a second tier fairly easily. I was wrong. They did, but it was not something that came naturally or without a bit of pushing. It was a very "swampy" game - none of the content could be rated because there was no innate reason any of it was more important than anything else.

This rambling little theory is getting too long by half, so I'll cut my meanderings short. Er. Shorter.

Basically, when you're designing a game, you need to think about the second tier. And, like most things I do, there are three pieces to consider:

1) Level, enemy, power, and item design. Any given part of the game should be relatively unique as to what it looks like and the tactical options it gives. Make certain powers more inherently useful in some places and less so in others.

2) Quests. Allow the players to take missions. I would suggest even giving them occasional choices to do random non-major-arc-related quests, such as bounty hunting or saving a useless village. Every quest should have a different feel than the previous one in terms of pacing and all the piece 1 elements listed above.

3) Barriers. As the players progress through the game, the gameplay needs to slowly change to keep things spicy. It shouldn't be so much that any character is ever rendered irritatingly underpowered! However, giving your players an airship, or letting them learn magic when they didn't have any before... it's a very powerful tool to change the way they look at the world. It also lets you do radically new and interesting versions of the two earlier pieces.

Did you get this far? Does your boss know you're wasting time on this blog? Leave a comment!


Eric Poulton said...

It's Arkaeyn's theory you're referring to. Originally conceived in the "Writing, or plotting?" thread:

I always liked this theory. It's easy to grasp and seems to be a pretty useful way of thinking about story in games.

Craig Perko said...

Fixed, thanks!

Chill said...

I suppose it depends on the group your playing with whether they can come up with their own second tier. The group I played with in high school in a Star Wars game would, in the absence of anything else either:

1) Start a pirate group
2) Start a illegitimate business of some sort

which obviously doesn't work with some characters but the point is, players tend to have things they want to do, even if they don't realize it.

I'm curious about third layer and second layer interactions, though. It seems obvious that given a third layer, players will invent second tier elements that serve third level goals. "Oh, we need to save the princess. Well we should take over this castle because it would weaken the enemy's power plus there might be information on where the princess is." I wonder though if that's the extant of it or there is something else there that would be interesting to play with.

Craig Perko said...

Chill: Yeah, that was my theory. I was wrong. It appears that the more cohesive a group is - the more they have played together - the easier it is for them to invent second tier goals. However, even my very cohesive group had trouble when I didn't give any value feedback. (IE, didn't say whether their second tier goal was good or not.)

As to third tier goals, second tier goals only get invented to serve them if there is a clear path. "Re-establish the Jedi order" is apparently too muddy a third-tier goal.

Patrick said...

My favorite instantiation of this in computer games comes from stuff like SMAC or Civ, where the second tier is genuienely emergent. The short term is move your units and adjust your cities production queues, the long terms is go into space or conquer the world or whatever - but the mid-term, you'd never guess. Wars, technology races, evironmental crises; its like the game is drawing cards dictating carefully designed missions, but those missions are actually the products of entirely deterministic calculations combining into full fledged gameplay scenarios.

I also found the pre-designed scenarios of those games to be far less rich.

Craig Perko said...

That's a good point: the Civ games (and other such games) use a tech tree system to limit content. The tech tree is basically a very tight second tier element. :)