The Holy Grails of Everquest Next: Emergent AI

When I wrote about Storybricks a year ago last May, I could not easily conceptualize how it would make MMOs better, though I thought it would. Then David Georgeson described what I shall dub the Tale of the Wandering Orc bandits. I’m not quoting verbatim, but it went something like this:

Orcs don’t like guards, because guards don’t like orcs. So orcs stay away from cities, because cities have guards. Orcs also don’t like PC’s because PC’s also kill orcs. What orcs do like are roads that don’t have much traffic, but just the occasional traveller that can be relieved of their possessions if not their life.

So orcs will travel around trying to find a spot that they like and set up camp. And should things change so that that camp is no longer suitable, they will move on.

That vision, all by itself, is pretty compelling. Like things actually breaking when you hit them, it seems likely to make the world seem a lot more real. This breaks the “there’s an orc spawn point there and there and there” logic. Orcs were there yesterday, will they be there today? We don’t really know. It depends on what other players did, and just how much the orcs liked or didn’t like it.

This is a world that will be different, perhaps from day to day. This by itself is powerful. Gamedevs just don’t have the bandwidth to do this. So you could look at this as “players will do the work of gamedevs” if you so chose. But really, it’s “players will do the work that nobody else ever did”.

But the responsiveness to player action, in this case collective action, is icing. Maybe you can clear the orcs out of an area, rather than pretending that the 13 of them you killed solved the problem even though you can see them respawning as you leave. In fact, what defines something as “grinding” is the fact that it didn’t mean anything in terms of the game world, or the other people playing it.

There are more layers to this, of course. (Remember Shrek?) Perhaps not all half-abandoned roads are the same. Perhaps some are under the protection of an orc King, who is not going to take kindly to harassment by uppity PCs, and will strike back at nearby settlements. Or perhaps the orcs have allied with a dragon nearby, and point out to the dragon where some delicious snacks are to be found. There’s lots of possibilities.

Now it’s possible that this kind of thing could be done as ad-hoc code in some generic programming language, because Alan Turing. The value of a Storybricks is that it puts the structure of the AI into terms that allow the gamedevs to concentrate on what should be happening in their game, and gets rid of details that aren’t all that relevant. That’s what any good library or language should do.

When people say something is “emergent” what they mean is “we have no idea what will happen”. And the reason that they don’t know is that it will depend, in part on what we the players do. That’s exciting.

My Long Slumber

When you’re three thousand years old, you need a lot of sleep.  Also, the fabulousness is on full display here, though green is really not my color normally.  (The artwork is by Heli Härkönen, from here.

I’m still playing a lot of Civ V, the new expansion is really fun.  Fun to play, but it never seemed to inspire all that much writing.   But something did inspire me to do more writing.  Actually, two things. Continue reading

In Which I Invoke Robert Johnson to Talk About Storybricks

Let’s get the cliches out of the way.   With the failure of 38 Studios, and the, well, mixed success would be a polite way of describing Star Wars: The Old Republic, it’s clear that the MMO world is at a bit of a crossroads.   (This is where Robert Johnson, he who learned to play guitar when he met the Devil at the crossroads, comes in).  These failures are very, very different, but still, it gets you wondering whether MMO’s have any future at all.

Here’s stuff that I know to be true:   People like adventuring and computer gaming.   That’s still true.   People also like doing it together.

I see two big issues with the MMO model.    First is the lack of social contract between players.

In any tabletop game, there is a social contract.    The players and GM will agree to meet at certain times, to play with certain rules, and often, on what the focus of the group will be within the world of the setting, be it home-brew or store-bought.   There was also some agreement as to decorum and language.    I’m not just talking about coarse language, but also about homophobia, or political commentary, and so on.    What sort of things are talked about in OOC, and in what way…

Not all groups start this way, but any good tabletop game ends up with a sense of shared mission among the players.    MMO’s completely lacked this.   Small groups could recreate this, and the raiding game could give a guild this.   The lack of social contract in the population of an MMO at large has been highly corrosive to the public aspect of these games, which is why pick-up groups have become such a problem.   The best pickup group these days just pounds through an instance at top speed, saying almost nothing to each other.

The second problem with MMO’s is their persistent nature.   Player characters can defeat great evils and destroy threats, and they will respawn in 15 minutes, so that the next group of PC’s can defeat them.

The drive to have an impact has shown up in such things as appearance gear and house (and guild hall) decoration.   Players absolutely loved these features in EQ2, who did them better than anything else I’ve seen, though I’m by no means a sampler of all things MMO.

I think the drive to have an impact on the world is also what drives people to grief.   Developmental psychologists will tell you that children will repeat whatever behavior got them the most attention, regardless of whether that attention is positive or negative.    There’s also a bit of regression involved as well – since most gamers these days, even though they are now adults, started gaming when they were children, they continue to behave as children while they are gaming.    But that’s back to the social contract issue.

In single-player games like Skyrim or Mass Effect (I, II, and III) the player’s choices have consequences to the world.   You get to decide whether the Stormcloaks win or the Empire wins.   It’s not prejudged for you either, there’s no obvious “good guy”.    In ME, what you do will have consequences later on, consequences which you might not have been able to foresee.

You can’t have this experience in most MMO’s.     Of course, the one exception to all this is EVE Online.   You can build space stations in EVE, and you can blow them up.   You can build ships, and you can destroy them.   When you mine out an asteroid field it’s gone, although another one will pop up soon.  However, since there are a large, but limited number of spaces for space stations and mining and so on, it is absolutely necessary that EVE be PVP, and that the game mechanics not preclude players blowing up other players stuff.

Which means that EVE has a social contract, of sorts.   The contract is roughly a lowest-common-denominator contract – anything goes, as long as you’re not hacking the game system itself.  Lie, cheat, steal, ambush – go for it.   If you’re ok with that as a social contract within the game, then EVE offers you the possibility of doing things that impact everyone.   Things like Jita Burns

However, if that’s not the social contract you’re interested in, there isn’t much out there for you.

So, those are the problems that we face today.   However,  there are some things that are still true.

  • People like to play computer games.
  • People like to play those games with other people sometimes.
  • People like to have adventures.
  • People like to express themselves.

I have a long background of tabletop gaming.   I’ve played in weekend-long face-to-face live-action RP games.   We played with about 50 other people, with pre-set characters and a one-shot scenario and plot.   It was one of the most exhilarating experiences of my life.    However, LARPS of this character require an extraordinary effort on the part of the writers, with not much payback.

I think that to go where I want to go, then, will require more participation by non-paid GM’s, and worlds (servers you could call them) that have smaller populations.   And those GM’s are going to need lots of tools to help them create interesting worlds for their friends to play in.

And that’s where Storybricks comes in.   The point is to build tools that will let non-professionals populate a world with characters that do things that seem human and reasonable and not simply “one-path”  amusement park rides.   It’s an interesting approach.

After getting walked through the tool by Kelly Heckman and Brian “Psychochild” Green,  I spent some time last week (while on vacation) playing with their demo toy.

Before I tell you my reaction to it, I want to tell you a story from my mundane software development days.   Another guy in my group took over a parallel-C programming language product about the time I started work.   We had a parallel FORTRAN project that was very popular.    He found that there were very few customer reported bugs, but as he worked on it, he found a lot of problems on his own and fixed them.   With a new release, the customer-reported bug rate went way, way up.

At first this disappointed him, until he realized that what it all meant is that before his work, everyone had written off the product, and weren’t trying to use it at all.  Now, they were encouraged enough to file bug reports.   They cared.

With regard to Storybricks, I found that I had several frustrations with the demo, but on reflection, all those frustrations mean one thing, I want to use this tool more, and I want it to be better, so that I can do more with it.   And that’s a good thing.

The 3D browser package Unity3D isn’t very stable.   I had several crashes and freezes, which were frustrating.   Currently the only models available in the game are male, and this was an issue for me.   I wanted to be able to name the characters myself.   I wanted to be able to create new objects.   I had problems with the program not saving the dialog I typed into certain boxes.   I want to create a character named Toldain with fabulous red hair, you know I do!

With my software developer hat on, I understand all of these problems.  It’s a demo, not even an alpha.   And my frustration means, among other things, that I want to use this more.

My personal efforts, in my spare time, are to push tabletop gaming into the cloud more – to go turn-based, social, and maybe even mobile, to exploit what computers can do to make things simpler for users and GM’s.    I would preserve the “live” GM, if for no other reason than I want players to be able to pick up a fork at the dinner table and stab someone with it.    In a programmed world, forks are for eating and swords are for stabbing.    Bridging the gap is a deep, deep AI problem, which as far as I know has been outstanding for 30 years.   So my approach is “keep the GM and give them tools to make their life easier”.

Storybricks is working on the same problem, but from another direction.   They are trying to push MMO’s in the direction of tabletop RPG’s.  Or as the phrase is in the MMO world, towards more “user generated content”.  It’s a welcome approach to me, and I support it.  The world is full of very creative people who do really cool stuff in their spare time.  If you don’t believe me, cruise YouTube some time.  So let’s give some of those creative people the ability to create multiplayer online roleplaying experiences.

Anyway, I’m pledging to their Kickstarter campaign, and so should you.   As I write this, there’s about 25 hours left.    Show some support.

Pizza and Storybricks


Last Wednesday, I had Pizza in Palo Alto with Brian “Psychochild” Green, along with spouses, girlfriends and daughter. We had a great time swapping stories of RPG’s and MMORPG’s and crazy stuff we’d done in them. Brian works for Namaste, which is running a demo at GenCon this week of Storybricks, a brand new technology and approach which Namaste is trying to bring to the market.

It’s my sense that Brian (and possibly others at Namaste) want to make MMO’s more like tabletop RPG’s. Which I love, so that’s good.

Phil Carlisle of Namaste says this:

What I’m more interested in, is the ability to actually feel like the world is allowing me to play a role. That I’m part of a story and can explore the world while the story unfolds, where the drama of the world evolves over time and where the mechanics of play in the world are less about accruing items and more about social play. Which I guess is why I’m here working for Namaste.

Yeah, that’s the cool thing about a tabletop campaign – the stuff your characters do matters. This could play out at the level of lore. Stephane Bura, also of Namaste, writes:

lore is useful for giving some context to the players’ goals: there are Demon Princes and Forger Kings, pick a side and kill the other one. It’s the wrapping paper on the quests. There is some reason, somewhere, why it’s important to slay demons in this game. This is comforting for some players because it brings a sense of order to the world – a sense that the developers know where they’re going. This kind of lore is also useful for setting up worldwide events and giving context to new content. However, even if lore distilled through quest texts can be very well written, most players skip it (trust me) because, in the end, it’s inconsequential. Players have no influence over such lore and its details have rarely any bearing on what they effectively have to accomplish.

But that’s only one level of lore. Stephane goes on to write about others:

So, there’s this magic item called the Scepter of Life and it control plants. A King owns it. You can bet that the farmers in this kingdom have a completely different life from your stock farmers’. They don’t fear droughts and they don’t need to take care of their lands so much. They’re much better at harvesting, since they do that all year-long. Inns serve soups, salads, jardinières and pies, as much as you want. Commerce is based on the exports of virtually free food, with dozens of caravans and shipments leaving the kingdom every week. Nobles fight over the amount of woodland and arable lands they control. The woodcutters have the most powerful guild. No imagine how all this would change if the Scepter of Life were to be stolen by some Demon Prince…

And the world is ready for virtual worlds that are like this. Liz Danforth, who is joining with the Namaste crowd at GenCon, says this:

Something has fundamentally changed in our expectations about entertainment and our interactions with the things we love. We expect to participate, tinker with someone else’s creations, to contribute and to share what we make. When Time magazine featured an article about fan fiction and the writer gets it entirely right, warts and all … Time magazine for heavensake! … the world has truly changed (and is continuing to do so).

So there’s a hunger for worlds that players can change, and canvases which are collaborative and expandable. That hunger goes back to Tolkien, who imagined Middle Earth to be a place where others would dally, and write songs or stories about. This is an entirely different approach to fiction from that of, say, Lois McMaster Bujold, who ascribes to “just in time” world creation. There’s a lot fewer continuity issues to fuss about her way, to be sure.

Anyway, I got a demo today of the Storybricks system pretty much as it will be demoed at GenCon. Kelly Heckman of Namaste took time out of her own preparations for GenCon to show me the latest build. What I saw was in two parts. The first was a screen that looks like the graphic I’ve posted above. The second was a typical Medieval street scene, with a guard, a citizen, and a thief. I watched as Kelly added Storybricks to say that the citizen would be happy to see me and the thief would be angry. The guard stayed neutral. As the player character approached each of these characters, it triggered animations that portrayed those attitudes. The guard turned to look at me, but showed little emotion. The thief glared and put a hand up in the universal stop sign. The citizen seemed very happy to see me.

These are the basic building blocks of storytelling. I’ve seen this kind of thing in a few places before. Specifically, once you did enough writs for the Qeynos Guard in Everquest 2, guard members would sometimes stop and salute as you ran past them. I have to say, that felt fantastic. Presumably this was done, for EQ2 via special ad-hoc programming in a scripting language and added to the guards. Storybricks, at the level that was demoed to me, would codify that kind of thing, and make it easy for content creators, be they professional or amateur, to add this kind of thing to a virtual world.

Tools matter a lot. I don’t think it’s a coincidence that writer productivity is a lot higher now than they were in the days of Tolkien, who wrote all of his manuscripts by hand in soft pencil and then wrote over them in ink. And at that level, Storybricks already looks to be something useful. The system will allow one to describe changes in attitude. For example, you might retrieve a Philtre from a crypt on behalf of a shopkeeper, and he will be grateful when he finds love. You might think that the point of the episode was that sword of smiting that the player gets, but it might be that the point was to make people happy, because that would foil some other plot. Or make someone like you. The potential for matchmaking is there, and that will definitely encourage competitive shipping. I can see it now: Factions competing between making Bella like Edward or Jacob more.

Err, never mind. Namaste isn’t really promising that. But they’re dreaming about it. Psychochild described to me, over that pizza, how you could have instances in which Bella preferred Edward and instances in which Bella preferred Jacob. (Actually he was talking about the dictator of Freeport Lucan D’Lere who has a crush on Bella and …. never mind) And by their interactions with those instances and the characters, one of those realities would get promoted to the default reality. ( I think we can safely say that three-ways are out of the question.)

So, I think they are on to something that players will like, and that’s possible. It’s a big job, though. But they have the sort of goals that, even if they don’t make all of them, they will still probably get something very cool. My only request is a Storybrick that says: “If (elf has red hair) then (NPC thinks he’s fabulous)”