GDC Transcript: James Portnow, User Generated Story: The Promsie of Unsharded Worlds

by Darius Kazemi on March 25, 2009

in gdc,gdc09,transcript

Here are my raw session notes for James Portnow’s GDC talk, User Generated Story: The Promise of Unsharded Worlds. This is my best attempt at a transcription of everything he said. Any mistakes or misinterpretations are mine and mine alone. I had to boot up my computer so my note taking started a teeny bit late.
Most of the time we’re building content for a story and then outside forces (marketing) make the story take a sharp turn.
Branching story is hard in sharded worlds, because if on server A they kill a dragon and on server B they make friends with it, then all of a sudden all the servers have different stories.
Unsharding is the solution to this problem. By unsharding I simply mean that all your players participate in same story space. Not necessarily the same world space.
Battletech 3025: fascinating game. Instanced FPS-style battles. Take a group of mechs and fight another group of meches, 10-15 minutes of combat, but Kesmai overlaid it with a simple 2D map of territory control. Every battle affected what territory your clan controlled. Territory affected what you could buy, where you could fight, and affected the story. All they did was lay over a 2D map. That was 6 years ago, so this is technologically feasible.
Some types of unsharded games:
* Instanced unsharded worlds give you the most flexibility
* Cross-server unsharded worlds give you a group of mirrored servers that affect a storyline that takes the aggregate of all actions across servers
* EVE-style worlds, which use clever tech to cram everyone into the same worldspace
Arguably, you could do a WoW-style unsharded fantasy mmo too in the future
Once you figure out how to get all these people in the same story space, what do you do?
We’re defining a term called a Massive Choice Event: basically, any stuff that fundamentally affects the world. Kill a dragon, it doesn’t respawn, that’s MCE.
That is maybe not the best example of MCE. Legend of the Five Rings: collectible card game. It’s a totally physical product, but they manage to get all their players in same story space. Do a series of tournaments, as they go through, they aggregate the results, who won where, which clans, and then integrate that into storyline of the next expansion. Some cards have names of specific players on cards!! People feel like they can really affect that world.
Ahn-Quiraj from WoW is an example of something that is not a MCE. The AQ war effort allowed players on server to turn in goods, eventually gates unlocked, world changed. But there was no alternative. Players could not actually do anything but open the gates.
How to make MCE good? TRUST. If you’re making a shared story world, it’s not your story, it’s your players’.
You have to be willing to give players real choices, real consequences.
Of course, when Lord British got killed in UO, they just pretended it didn’t happen. Would have been cool to integrate it into story: let’s all band together to resurrect LB, etc. You need to let that kind of thing happen, roll with it.
Making MCE matter
MCE has to be tied to mechanics. The consequences should affect where your players can go, what they get, etc. Pure story consequences not so interesting. Should be grand scale, affect everyone.
Universal participation: everyone should be able to be involved in massive choice event. If you allow even your lowliest players to feel like they’re touching a living world, the’re going to realize that as they progress they’ll have an even greater affect on the world. And it gives them a “Here’s what I did” story to tell.
Patching according to MCEs
When do you patch? How many patches can you stretch it over? I encourage you to make MCE effects patch-independent: spawn rates, drops, etc. If you can deliver those it’s plenty for players, and over a series of patches you can unfold the results.
MCEs are governed by meaning. You have to be willing to get past our paradigm of building high-level content. What about stuff for social players? [Darius note: reminds me of taverns in DDO, they were just dead spaces with quest givers, even though the tavern is such an integral part of the D&D social experience.] If you have a player who rallies people to heal the front lines of the war, they should be a part of the story.
Writing MCEs: simple entry points. Look at how fantasy novels are written. The protagonist is generally not out to go change the world. They get swept up into massive events. It should be that easy for your players to get into your MCEs. What I’m doing in my standard everyday activities should get me into these. When I talk to my merchants to buy and sell, you should alter the text a little to talk about world events. That’s how you draw everyone in. A high barrier of entry leaves engagement on the table.
You want to direct your players’ experience. User generated content will be 90% bad, but with a little direction and constraint, it presents a much better experience. You DO need writers. You can’t force a particular experience, though.
Learn from ARGs! Creators of ARGs gahve done more work on massive audience engagement than MMo devs have. 
Use MCEs to bring players into your storylines.
Seesaw events as much as possible. Everything should be a near thing [like scoring in basketball]. Blowouts in sports cause walkouts, close games cause tension and excitement.
CHEAT! Without breaking your contract to your player–let’s say there’s a war. One side is getting stomped. Write in that the side has found a dark alliance or something and increase their spawn rate.
Sometimes there are controversial results. Players wont like some choices. Write for your disenfranchised players: give them a quest to go take back what they lost! Incentivize them.
Okay, but we have to implement this. How do we make these cheap?
Refactor existing content. You don’t have to patch in stuff like changing text, spawn rates, loot tables. When you have to create content, it’s going to become part of your world. Make it reusable!
Ignore what I said above: every once in a while you want a big culminating once-in-a-lifetime experiecne. Dont’ be afraid to throw out content. But you have to balance when it’s worth it.
Make the outcome as easy to anticipate as possible. The longer events last, you have more time to predict the way player base is leaning so you can predict what to patch.
Use everything you’ve got to make sure people knokw about MCEs. Drop mysterious note in mailbox, town criers, text for vendors, just email ‘em!
I want to talk about a youtube video from EVE online. This guy is sitting in the station, he buys 1B tritatinum (basic mineral) stock price is falling. Then he’s talking to talking to guild leaders. Convinces them both to go to war. Shows stock chart: tritanium is rising!! Great story of your evil war profiteer. The story is lost, though, because it didn’t affect the world.
First, incentivize player politics. Guilds in WoW provide no mechanical benefit (only group  coordination / social benefit). Provide material benefits. Allow player interaction to affect your storyline. Recently someone turned traitor and destroyed their own guild in EVE. If you had factions at war and guilds could ally with them, maybe whoever wins, that faction gets new ships, that traitor is now a huge thing in the world. Make his name a curse word for the losers, name a ship after him for the  winners…
Stir up trouble. Put a resource they both need in a place where players have to fight over it or resovle their differences.
Force people to choose sides. People want a sports team or a political party, larger than themselves. Give people something to root for and associate with. People dont’ want to play in a nebulous undelineated world. Once these sides have locked into certain patterns, create common enemies to bring them together!
A brief recap: story in mmos won’t improve until we can affect the world. Need to be shared story space. Players need to direct story (in a guided way), give them events aggregated over the long term that you can build for. And then let player political interactions affect the landscape.
Q+A: no questions, let’s talk about unsharding a wow-type mmo instead!
The key to unsharding your fantasy mmo: well, you need procedural content. Terrain system. Socially designed resource system. Take cues from real cities and grouping. Adaptable encounter generation: a system that takes into account what players are where, their level, what they’re doing.
User content: players are going to have to create whole cities and run them, build outputs. Player driven economy. MAKE SOCIAL FRAMEWORK FOR PLAYER CREATED CONFLICT.
Socially designed resource system: why do real cities form? Why do they leave? People gather around natural resources or trade routes. They leave when resources go away or no longer support the population as much as other cities might.
Once you have people in cities, you direct them with resources, we need to keep gaqmeplay. What do players want out of encounters? Loot, fun, and feel something vibrant and different. Designers want to keep balance.
View your monsters as resources. Have spawn pattern adapt. Rather than spawning individual monsters or mobs: spawn set pieces (2 orcs, an elf, and a caravan) [like Fallout-style encounters but procedural!]. Have the elf fight back. Or have him helpless. Change it up.
Make town building rewarding and easy to do. Town management should be interesting.  You have to have isntant travel because your world is going to be enormous. This is a hard problem because you don’t want o your world to be largely menu-driven.  Maybe onlyl allow towns of certain size to have instant travel or something.
Lastly: it will break. If you’re doing a procedural mmorpg, the system will break down, the players are smarter than any system you can build. But limit the amount of time people can break your system. Breaking the game is about the creativity of players.


David Sahlin March 25, 2009 at 8:08 pm

Yes. Massive Choice Event is the phrase which summarizes my philosophy towards MMOG design.

Thanks for the notes, Darius.

Austin Grossman March 25, 2009 at 8:38 pm

Thanks for throwing that up! The GDC exile appreciate it. Anything else you want to put up there would be awesome.

solipsistnation March 25, 2009 at 8:40 pm

Ha ha noobs. We solved a lot of these problems for LARPing in the 90′s!

More seriously, some of the issues are very similar, just larger. You also give your players more freedom to do stuff, since you don’t hand them character sheets to start with and people are playing for more than a weekend. Although in an MMO you do hand people archetypes to work with, and if there’s an explore/fight/loot loop going on then there’s more to the game than pure character interaction and role-playing, which can sustain players for longer than a weekend or week-long game.

This is also interesting in the LARP context:
If you allow even your lowliest players to feel like they’re touching a living world, the’re going to realize that as they progress they’ll have an even greater affect on the world.

In almost every LARPs I’ve played (and remember, these are theatre-style role-playing games rather than boffer fights), there has been some new player who is given a fairly light role because the GMs weren’t sure what to do with him (or her). Sometimes they play it light or don’t get into it or whatever, but often those new players in light roles expand to fill some kind of space left by, say, a player in a major role who isn’t that into it. Or on some fantastic (but sadly rare) occasions they make their own space in the world of the game and have that greater affect.

There was a flip side to this, of course– inexperienced GMs writing games occasionally put in character who are, on the surface, pointless powerless window dressing but who, upon closer examination, turn out to be REALLY pointless REALLY powerless window dressing. The best example I have of this was a game based in the Star Wars universe where there was a character of a “broken droid.” This character was pretty much written to wander through, do something silly, and wander off. The GMs made it explicitly powerless and unable to communicate. And then they asked a player to do that for 3 days– wander around in costume annoying other players until they finally just ignored him. I think he quit Saturday afternoon and I doubt he ever LARPed again. This led to an article I wrote for the New England LARP association magazine about ways to write interesting stories for your characters and how to make sure you don’t have characters who are window-dressing. When you play a game like that, everyone is the protagonist of their own story. In any MMO, all those people are protagonists, they’re all the hero of something (or would like to be), and there’s nobody there who doesn’t have something interesting going on.

The good thing about an online MMO is that you can have NPCs to do the window-dressing stuff. Need a crazy kobold who runs through town and annoys people? Well, the internet is big and full of weirdos so you might have people who do that particular thing on their own… But you don’t NEED to have people paying you (and using their own time) to do something that’s window-dressing. Now you just have to make sure that everyone who starts out as a random guy on the street is totally sure that random guys on the street can aspire to godhood and get as far along that path as their ambition and skills can take them.

Anyway, that whole “nobody is a side character” thing is one of my favorite things to get kinda drunk and talk about. I think it’s interesting that the MMO world is reaching the same sorts of problems from a different angle than the LARP world. I don’t think it’s a coincidence that some of the people I gamed with at InterCon worked at Looking Glass…

Oh, and this part:

Socially designed resource system: why do real cities form? Why do they leave? People gather around natural resources or trade routes. They leave when resources go away or no longer support the population as much as other cities might.

…Mr. Tepper? Phone call for you.

solipsistnation March 25, 2009 at 8:41 pm

holy crap that got long. sorry. 8)

Amanda Cosmos March 26, 2009 at 2:17 am

Interesting. I’m definitely going to show this to my Darkfall-obsessed coworker tomorrow because a lot of what was mentioned here (player-built entities, ever-shifting alliances) definitely are ringing bells when I think back to what he’s told me about the game. I’m sure it’s to a simpler degree but I’m now curious about his take on it.

David Sahlin March 26, 2009 at 6:20 am

Darkfall certainly has the player-driven conflict down, Amanda, but influencing the world outside of that looks to be pretty non-existant.

edubois March 27, 2009 at 5:55 am

I’m surprised someone besides me remembers the canceled BT:3025

Oh what could have been.

Comments on this entry are closed.

Previous post:

Next post: