NPCs are rubbish. They are simplistic vending machines that lack the sophistication of even the simplest natural language processing system. Over on MUD-DEV Amanda Walker recently proposed a tangential solution..
.. get rid of the ‘N’, or in her words (quoted with permission):
- Since many of the most addicted^H^H^H^H^H^H^H^Hloyal players are those with a lot of time and not a lot of money, why not hire them? "Game company has entry level positions for experienced bartenders, no cleanup duties, must have own broadband connection and type 50wpm." It does open up new horizons for grief play, but think of the quests: "hey! that new guy I hired stole all my stock! Reward!"
Given the cost of content generation in story / quest driven MMOs, I wonder if the hiring, training, QA, Customers Service & vetting costs of this solution would ever stack up.
Then there are legal issues of responsibility – it’s not like Disney is allegation free when it comes to the conduct of their ‘cast members’; and what if a character became really popular where are the lines between actor and avatar, could the actor jump VWs, if so, how much of their performance could they take with them (cf Beard - Clones, Bones and Twilight Zones).
Arrrg it’s a minefield – but it would be neat wouldn’t it, just think of the depths of game play and immersion that would be possible.
Already having Ingame Content Providers with large set of tools has prohibitive costs. Not to speak about having actors performing NPCs.
There was a try by EA to use volunteers as ingame actors for events (so at no cost) but that is a taboo topic seen the legal issues behind employing volunteers.
What is left? The old AI, investing more in researching realistic AIs, a better understanding of narrative and storytelling in PWs through algorithm driven mechanisms and related.
At the same time empower the players with better tools to have also player driven ingame content and fiction. Mix altogether with a bit of realism and you get a viable solution and a unique selling point.
Posted by: Luca Girardo | May 31, 2004 at 09:27
My suggestion wasn't to use volunteers. But after minoring in computational linguistics in the 80s (and following the general course of NLP and AI in the years since), I'm very skeptical of the idea of realistic, engaging synthetic human characters. That "bit of realism" can be very difficult to achieve--humans are very good at noticing subtle (or not so subtle) cues that there's not a human at the other end.
My speculation grew out of the observation that many entry level jobs in the real world serve the same function as NPCs in virtual worlds (and alas, get treated as such by 'players' in the real world, but that's another topic). It struck me that playing, for example, an in-game barkeeper is closer to being an actual bartender than it is to being a Hollywood-style actor, and that taking this into account might shift the economics of human-driven "NPCs".
The economics aren't the only factor, of course; that was just the topic that had come up on MUD-Dev.
Posted by: Amanda Walker | May 31, 2004 at 10:11
I've long thought that something along these lines would be a good idea.
I usually don't have time to fully participate in a MMORPG. But would I log in long enough, once in a while, to play a bit part? Sure!
If I could log on for free, I'd play a low-level monster sure of nothing but my own self destruction, a beggar who could trade a wisp of information for a coin, even a wandering salesman with two swords to sell and not enough vitality to wander outside the city walls.
Sure, the room for abuse is there. But it seems there are all kinds of interesting ways to police that.
One other thought, along these lines--give me a $2 a month account that lets me play one of these limited range characters. Maybe I can accrue enough wealth over time that I'd be incented to join the game proper. Or, how about making these PPC (partial player characters?) accessible though simple Web interfaces or the like. I wouldn't be in the world, per se, but could interact with it through a limited interface. I always wanted to check in on my house in the Sims as a household pet--I could see who was there and listen to them, but no do much else.
Anyway, seems like there are a lot of good ideas down this track. The trick is trying to expand the notion of what playing in the world includes. For me, $15 a month a 15 hours a week just isn't possible. But other models interest me!
-- David
Posted by: David Thomas | May 31, 2004 at 11:49
Why not set up policies and procedures for hiring troupes Synthespians? Live content just doesn’t seem to be developer’s core competency, so let more capable people do it. Allocate, say, 5% of the monthly subscriptions to supporting for-profit and not-for-profit Synthespians.
For example, players can get 1 point each month as part of their subscription. They can use these points to vote for the best Synthespians or get tickets to premium Synthespian events. Developers will then use this simple market-based mechanism to allocate the Synthespian fund.
Thus, instead of a relative laissez faire environment like SL, you have a more centralized structure with VW-operators as the taxman and players as citizens with vouchers.
Frank
Posted by: magicback | May 31, 2004 at 12:19
David,
Maybe someone will pay you $1 a month to play a
group of random low-level monsters RTS-style against many single 1PP & 3PP players.
Frank
Posted by: magicback | May 31, 2004 at 12:29
David wrote: Sure, the room for abuse is there. But it seems there are all kinds of interesting ways to police that.
That is probably the biggest problem when designing ingame mechanisms and tools for players driven content. How do you ensure they won’t be abused? That is critical as abusing griefers are a risk for the whole community and it is a deadly killer of retention.
On the other side for example I have experienced directly the use of human NPCs in UO (mostly monsters in dungeons) and AO (general NPCs characters ) and I have to admit that the result for both the players participating was a very intriguing one as it was adding a new dimension to the game for the players.
Posted by: Luca Girardo | May 31, 2004 at 12:29
I don't have a realistic sense of the costs involved. But given what programmers make, it seems to me that you could pay people 6 bucks an hour to plan an 'npc'. People would jump on it. And if the people running the show insist on high standards, the benefits should outweigh the costs.
Just make sure you place people who are going to take it seriously.
Posted by: MM | May 31, 2004 at 13:06
Something I left out of my little screed, but I think is the heart of what I was getting at:
The average online game charges about $15 a month. For that price, the game promises everyone a similar experience--or, at least, the promise of a similar experience.
This seems odd to me. It doesn't recognize a fundamental range of player motivations and time commitments to play. Not everyone wants a game that takes hours a week. Some want games that are much easier and take less of a commitment of time.
I would be inclined to purchase the "Everquest Orc Kit" for $5. For that amount, I could join the EQ world as an orc--or series of regenerating orcs. I would never be able to wear fancy armor or wield a magic weapon. My hit points out be minimal. I might only be able to choose from a limited menu of things I could say. It would be a partial EQ experience.
Why would anyone go for it? Well, like I said, I usually don't have time to take advantage of the all the promise of a $15 a month account. So, realistically, today I'm paying for the promise but am living like an orc in that world.
So, why not take advantage of this marginal curve of time and interest? Why not allow players in at the lower ranks to to provide NPC-like interaction, but as I suggested, partial player character (PPC) involvement?
I know, it's odd to think that there are people who would casually play a MMORPG. Still, most of the game hours clocked in a given year are, by definition, causal. How about a PopCap-like game that interfaces into a virtual world? Call the game "Haggler". The casual gamer drops in and plays a few rounds, the persistent world player interacts with a cantankerous old shopkeep who is always driving a hard bargain.
The real world works like this. Not everyone values their time and their money the same way. So, it creates interesting economies. We need to expand the range of consumer types in virtual worlds, I think.
David
Posted by: David Thomas | May 31, 2004 at 13:36
NPCs don't have to pass a Turing test, they merely need to be entertaining.
Posted by: AFFA | May 31, 2004 at 14:03
I would be inclined to purchase the "Everquest Orc Kit" for $5. For that amount, I could join the EQ world as an orc--or series of regenerating orcs. I would never be able to wear fancy armor or wield a magic weapon. My hit points out be minimal. I might only be able to choose from a limited menu of things I could say. It would be a partial EQ experience.
Could also be another means for folk who are more competitively-inclined to participate constructively in a cooperative game...
Having said that, sounds like it could undermine the "camping" design tradition. My suspicion is that a large number of folk who currently inhabit these worlds would find it unsettling... But maybe mmorpgs should make socializers dance in the cantinas anyway.
Posted by: Nathan Combs | May 31, 2004 at 14:05
The game Final Fantasy XI was also an MMORPG.
I played it for a while and lost the interest in playing with people who had been playing for months before me and were on a scale so grand that I couldnt hope to reach. But I did find enjoyment in roleplaying. My character was a race of all female cat like creatures and would goto the semi private market areas and dance around while taking off clothes. Weird, I know, but it was meant to be a joke to a friend of mine. To my surprise characters/players cheered me on and gave me amounts of money. I did it about three more times and I am sure it added a fun element to those who played.
I would like to see more of that in Online Games. I think things like that add to the gameplay for everybody and programmers may not see the small intersting things that can be done to make games more exotic/memorable for the players.
Posted by: Kendrick | May 31, 2004 at 18:19
Okay, say you have $15/month. Of that, you assign 15%, or $2.25 to pay for a troup of $6/hour thespians. Employment overhead is usually double the wage, so your actual costs will be $12/hour thespians. (You say you want high standards? Then you need people to enforce those standards! 100k players would pay 120 Synthespians working 40 hours weeks) This then comes to 11 minutes per month one-on-one Synthespian time per player.
Of course, you can gain greater ROI by having the synthespian interact with groups of players. One might then expect 1 hour of time per month. Provided, of course, the synthespian content was equally distributed, which it won't be.
1 hour of content per month isn't quite depths of gameplay and immersion.
- Brask Mumei
Posted by: Brask Mumei | May 31, 2004 at 20:58
Brask, you are spot-on with your analysis. I haven't followed this particular Mud-Dev thread, but a similar idea was put forth last December. The problem, as I argued then, is not the relative costs of 'live' and 'static' content, since both pale next to the opportunity cost of the player's time. Rather, the problem is that live content -- unlike its traditional, static counterpart -- is more-or-less rivalrous. Static content may be expensive to produce, but it can be shared amongst an arbitrary number of players.
Posted by: Jeremy Neal Kelly | May 31, 2004 at 22:06
(Having said that, I'd love to play the game this thread describes, but really -- isn't that what pen-and-paper RPG are all about?)
Posted by: Jeremy Neal Kelly | May 31, 2004 at 22:11
Brask,
The internal economics for making a 5% budget does not rely on a 6/hr minimum wage.
There will be many for-profit and non-profit troupes that would be happy to an equitable opportunity to show their "stuff." Thus, it can morph into like a funds for the arts, or at least in structure.
These synthespian art works would probably not add the the treadmill process, but would add to the life of the game. Kendrick has already got the good vote from his impromptu dance.
Moreover, the live synthespian could be utilized as extras in more involved stories and plots that does have entry-level and senior digiteers (I think Disney used this term) running the show.
Frank
Posted by: magicback | May 31, 2004 at 22:12
The rivalous nature would cause camping or hunting for live content, but the content can be structured to have value afterwards.
One example is the sports event. We all still enjoy the game even though we already knew that team A won in triple overtime. So, as a myth-building process.
So the entertainment value is broken into these stages:
1. Live action: rivalous and temporal unless it is a worldwide event.
2. Video replay: non-rivalous external content. Someone can mentioned the event and the game would go into this dreamy flashback mode like in Movie/TV.
3. Written story: non-rivalous external content. People can go to the Great Library and read the story.
4. Synthespian retelling: rivalous and temporal. Synthespian will be around town and put on a play.
It doesn't need that much money to get this all going. Like the tread on guild support, it just need better people organization. Get this going and players will flock to the best "act".
Frank
Posted by: magicback | May 31, 2004 at 22:27
Just requires money, and commitment. Two things that are the most lacking in MMOG projects, especially starting out.
That and ability. Can't forget that. Oh, and an actual story to tell. I also forget that most MMOG developers have zero interest in telling a story; they just want to throw together a world technology so customers can whack foozles in a virtual sandbox.
But even then, it'd be a hard sell. For every player likely to be enticed by a good story, there will be one who will actively work against it, and five who will have a reaction sort of like, "Huh? What?"
And yes, NPCs are vending machines, because that's all most MMOGs allow them to be. They dispense stuff, ostensibly because that's the only way players will even pay attention to them. You'd think there would be a way to dispense some story along with that, or at least try to bridge the gap between bozo sitting at the computer and the living, breathing world he/she paid to experience.
At least City of Heroes is trying that, now. And so did Guild Wars' NPCs, though I was hard pressed to figure out what the hell was going on during the E3 demo.
Posted by: J. | Jun 01, 2004 at 02:13
"There will be many for-profit and non-profit troupes that would be happy to an equitable opportunity to show their "stuff." Thus, it can morph into like a funds for the arts, or at least in structure."
But what then does the game company have to do?
There are already stages in Ultima Online ready for you to put on plays. People *have* put on plays on those stages. There are already bars looking to be manned by barkeeps and drunks alike.
Content can be generated by users. I helped run entirely player created quests which attendees believed were GM run.
So why then does the vast majority of the player base play in ignorance of this work? Why did so many UO players remain convinced that Seers never existed, even when they were actively running events?
I don't think we realize how big "massive" entails. In sub-100 groups, you can have a focussed community and all can feel involved if someone they met slew the evil demon terrozing the land. In the 10,000 people of a shard, the demon slayer will be someone who I never met and never will meet. The story will have no meaning to me, and I will not read it.
I don't hold any hope of a volunteer based system meeting the needs you assign. If you want to gain large penetration, you need to ensure the actors play for a long time and over a socially diverse group. Volunteers will play when they feel like it, and will play *to* their friends. As a synthespian, you do *not* want people to flock to the best act. You must control the crowd so that the content is properly enjoyed. If you have a chess tournament and 500 show up to it, and you don't have enough tables or space, it is not going to work well.
- Brask Mumei
Posted by: Brask Mumei | Jun 01, 2004 at 02:17
Brask Mumei,
You thoughts on the impact of Synthespians on massive games are correct in that the current amount of live content and their relevancy is very low.
"If the tree fall in the forest, I don't care whether it made a sound or not."
For me, this perspective is ok. What I am advocating is a centralized funds of the arts in VWs. 99% may be crap, but that 1% may add great value to the VW. My hope is that the net effect is more subscribers and more than the admin headache.
>But what then does the game company have to do?
OK here's a policy position:
Allocate 5% of gross subscriptions to a descretionary fund managed by the Head of Live Content. Procedures will be outlined for troupes to submit one page concept of their "artistic work". To minimize overhead, application will only be accepted by well organized troupes that have standalone legal entities (partnerships, corps, LLCs, etc.). The Live Team will give the application an quick "accept" or "reject".
Accepted Syntespian will have a special flag that let players know they are Synthespians. Players can vote once a month for their favor synthespian (or for a group of synthespians).
The Head of Head of Live team will payout from the fund monthly based on a performance measure that integrates this player feedback and internal audits by the Live Team.
Each month a post-audit will be performed to see what changes can be made to improve performance of this policy.
At the end of, say, two years the project is evaluated for net effect. If project have negative value then the project is terminated.
Posted by: magicback | Jun 01, 2004 at 03:36
Ahh, the raving.
My opinion is that NPCs could have a better role, in particular if you want to give some depth to a world and not creating just an arcade.
What I would try (and find interesting) is to add a more concrete life to NPCs. For example give them a precise schedule during the day and let them live a meaningful experience. They should sleep, go to work, eat and so on.
At the same time this has to be included in the gameplay. So NPCs could be used not just as "vendors" but as active elements like workers that help you with boring activities, guards that you hire to defend an outpost or your home and so on.
They could tell a (meaningful) story as J. suggests only if they have a real role in the world. Or you have just some form of "lore" you read on a website.
And that's it.
Then I also believe that mmorpgs don't need good stories. That's not the first priority. Mmorpgs need better structures and backbones so that the experience is meaningful and fun. If you offer a structure with some potential the players will build their own dynamics with noone with GM-powers required. I actually believe that the quality of a game resides in the *absence* of the continue need of an "hand" from the outside to "correct" what happens.
You should need just the Customer Support for the technical problems. All the rest of the game must stand on its own.
Posted by: Abalieno | Jun 01, 2004 at 05:35
***NPCs don't have to pass a Turing test, they merely need to be entertaining.***
And I found that line as the most interesting one.
Posted by: Abalieno | Jun 01, 2004 at 05:39
Players in the roles of NPCs are common both in live action role playing and in MUDs. The problem in both cases is this: humans are not reliable, so things will get messed up. Administration frequently gets complex, and many are the GMs to learn that it's easier to do something yourself, than to delegate.
There are some things that need to be equal, in order for a game to be fun. Some standard equipment need to be available, some sources of income need to be there for all, some challenges and some quests. If those services were to be provided by human beings, the restraints humans operate under: connection, language, distraction, time zones etc etc, will immediately apply. This will cause a skewed playing field, which will undermine the important initial level of fairness that makes a game fun.
Add to this what others have mentioned: that you want GOOD players for the shopkeepers, bartenders and even monsters. You don't want players who get bored and start chatting about the latest buffy-episode or wander off looking for an other monster to play furry-games with.
A limited quest run by administrators or with the help of players is very time-consuming and takes a lot of planning, training, administration and a severe amount of playing time.
I think one way to approach this is to give the players a chance to gain additional skills for their characters, that will let them "craft" or "grow" unique items to use as goods in the game. That way you can have peddlers with unique items that will make playing more fun both for seller and buyer, while you will still keep a certain basic level of even access to fun stuff for all. And if those skills are available as options to monster-killing or other useful gaming stuff, the players who would like to be shrewd businessmen or skilled artisans can work for it and specialise, like others do in different directions.
So, I guess, pretty much what Luca Girardo said, really :)
Posted by: Torill | Jun 01, 2004 at 07:10
There are some basic limitations that we have to take in consideration about the amount of resources that are available for live support and live ingame content. The best example of human generated ingame content in the different forms was as far as I know (please correct me if I am wrong) UO. They initially started with a small group of volunteers playing Roles in quests. Then improved and extend it with more paid staff (IGMs), volunteer staff (Seers, Elders and Troubadours (also the so called actors) and increased the numbers (up to 100-200). At the top of the whole they launched a so called massive shard wide event scenario (Trinsic invasion). During that event cycle, there were at least 5 to 10 volunteers creating support events for the event scenario with the main events having up to 10-15 volunteers and IGMs working on them. That was one of the most successful ingame content I have ever seen. It was made possible because of the large use of volunteers (so free work force). And still you could involve just a fraction of players on one shard.
Economically of the 12,99 $ to 14,99$ of the monthly subscription, just 10% to 30% is left after removing all the current costs (for more details one good source is the book written by Jessica Mullgan and Bridgette Patrovsky ). Any solution that would decrease that thin margin has no real chance as it is not realistic.
On the other side player content is free. Why not focusing on that? Why having an employee playing the role of a NPC when we could have it be done by a player?
Do we want intelligent monsters? What about different playable races? That would mean transforming the PvM to PvP. But can you really have a better opponent then a human controlled character?
What about adding more depth to the current NPCs. For sure we won’t have a realistic conversation with a NPC in the next 10 years but already small improvements can add new dimension to the Virtual Worlds.
Example: moving away from repetitive boring tasks done by players (like resource harvesting) and letting players use NPCs to take over the boring tasks while players focuses on the administration side. So a miner role. No longer the player has to mine itself with a series of repetitive tasks, but he can hire several NPCs to mine a precise spot. So the miner role would chage from a series of repetitive actions (mine spot A, move minerals to spot B, go back to spot A and so on) to a more managing activity about searching for a good mining spot, organizing the logistic transportation of material between the mine and a secure area, hiring the NPCs necessary for mining and transport and so on. I do not expect to have an interesting chat with the NPC but I expect to be able to achieve complex goals through the use of NPCs where NPCs need to have a better control interface and be able to execute more complex actions. How complex is a problem of developing resources in term of AI and architecture resources in term of how many NPCs you can simulate. But it is just an example how you can increase ingame content without trying too radical solutions.
Posted by: Luca Girardo | Jun 01, 2004 at 08:18
There’s a horrible business model lurking in here somewhere:-
Evil_marketing_ren:
VVG – Vanity Virtual Gaming
The ultimate MMO experience for the cash rich / time poor.
You’ve got your from eBay – now sign up for the entire guild / high level quest experience.
Yes with VVG you too can experience all the fun of being an integral part of an uber-guild, with out all that annoying business of having to spend time playing the game, getting to know people or even, learning how to play.
Yes settle in as our troupe of Synthespians take you along for the ride, giving you tips, not griefing you, marvelling at your exploits and laughing at you jokes – Oooh how they will laugh.
Sign up now, only $50 per hour.
Posted by: Ren | Jun 01, 2004 at 09:27
NPCs are rubbish because they are badly written. They don't have to have bulletproof AI to be engaging and entertaining (which is good because we're nowhere near there yet). And we can't rely entirely on player-generated content to provide our entertainment, even though we should try to encourage it.
NPCs deserve respect. They need to be created with some style and some awareness of what it means to be a character in an ongoing massively multiplayer game by people who actually know what they're doing.
Bad actors, which is what you get 93% of the time when you won't pay, or pay crap wages (or sometimes even pay tons) aren't going to help immersion and "depths of game play" any more than badly written NPCs.
Unpaid advertisement! This seems like a perfect time for a pitch. I just wrote a whole book on this stuff called Character Development and Storytelling for Games. It's published by Course Technology and will be shipped to retailers on June 15th.
Now back to your regularly scheduled discussion...
Lee
Posted by: Lee Sheldon | Jun 01, 2004 at 13:20
An interesting idea, but a large portion of the time I don't want NPCs to be entertaining - I want them to buy my loot at a pre-established price so I can get back to chatting and playing with friends without having to skip 5 pages of menu dialogue from some gabby vendor. I certaining don't want to stand around town going "dude, where's my merchant" because his meaningful experience is calling for him to have a lunch break every 48 min real world. It might be neat to know it's a real person, but I'd probably just feel sorry for them. All in all not a good ROI for players like me (there's a lot of us).
I guess I look for emersion from graphical detail, good "monster" AI (I love DM-controlled bosses, btw) and the often overlooked ambient sound - but otherwise it just gets in the way of my game, which is time spent with friends of my choosing, not some $6 armor vendor that thinks just because I sold him 100 animal skins I want to be his friend.
I would possibly be interested in some minstrels. "He was not in the least bit scared to be mashed into a pulp" and such.
Posted by: Staarkhand | Jun 01, 2004 at 13:37
>> I would possibly be interested in some minstrels.
Now here's a good place for an NPC chatterbot. I mean, how hard is it to algorithmically compose rhyming drivel? :)
Posted by: David Thomas | Jun 01, 2004 at 14:21
I agree with Lee Sheldon.
Add to that, the reason Disney World uses real human actors in its giant Mickey suits is *only* because they are, so far, cheaper than robots.
They will be replaced, probably sooner rather than later. And will we say, "Look, they're usuing bots and they make way more money and entertain way more people than we do. Maybe we should use bots, too!"?
Right now, bots are cheaper than humans for us, *already*. We're ahead of the game here, not behind.
A human in a mouse-suit is suitably convincing to be believable to very young children (as evidenced by the stark terror in their screams on sight of the beast)... and that's about it. The attraction is for parents to collect a souvenir trophy photo of their terrorized tot on the rodent's knee. Comparing NPCs in MMO worlds to theme park 'actors' is comparing cars to carrots.
A human role-playing a bartender in an MMO would be less convincing, rather than more, assuming nothing else about the MMO is different.
If the NPC is unconvincing in terms of seemingly aware of what's going on in the world, it's mostly because nothing is going on in the world.
The gnolls are still invading, the innkeeper's daughter is still missing, this note to her sister still needs to be delivered. Nothing has changed, and the NPC knows it.
Replacing the NPCs with humans doesn't address that issue.
It *will* cost a lot of money, however, and the very business you're attempting to emulate will stop doing it before we're even all on board.
Posted by: Jeff Freeman | Jun 01, 2004 at 14:33
It's a neat idea, but overall I don't think it's a good idea. How much more engaging does your average chat with a shopkeeper need to be? If we're not talking about just spicing up housekeeping tasks between the 'adventure', then our synthespians must be creating content to be useful.
But human interaction just doesn't equate to meaningful and engaging content by itself. It isn't just a matter of plausibly acting like a barkeeper. It's a matter of having the tools, time and talent to come up with and deliver a fun and engaging experience through that interface (pun intended).
A sythespian might be able to interact like a living breathing person - but living breathing people don't create engaging fiction simply through existence. If human interaction were enough by itself, actors wouldn't need writers. And writing begets bad writing, drafts, editing, and polishing. This all increases the need for tools, time, and talent. Even the best writer/synthespian is going to have a hard time juggling a convincing conversation with a live player alongside setting a half dozen triggers and spawns for an ad-hoc adventure.
Troupes of synthespians might be a novel way of fleshing out GM-run events, and make those more engaging. But simply dropping even an extremely talented thespian at the controls of Joe the Smithy doesn't itself create any sort of depth. It makes conversation a little more plausible, certainly. But there's no reason any convincing interaction between a synthespian and player couldn't have been canned.
IMO the most glaringly tragic part of PC/NPC interaction isn't the peculiarities of text parsing or conversation trees. It's the sad state of NPC writing in persistent world games, and the complete lack of expressiveness of NPCs. Add some interactive emotes: have NPCs walk about the room while they talk, fidget, rock in a chair beside a fire, shuffle their feet, pat the PC on the back, throw a mug in disgust, or drop what they're carrying in surprise or horror.
Employing synthespians would only make painfully obvious the lack of expressiveness in our current array of graphical interactions. They might be able to impart a great deal of expressiveness in a text-based setting, but I think the straight-jacket of expressiveness in our graphical games would severely curtail many of the benefit they could bring.
Sure, interactive emoting and creating/maintaining a large animation library is a challenging technical task. It's certainly much easier to make two avatars hack away at each other than to have an NPC drop to her knees to give her wayward son a hug upon his safe return.
But i'd take the latter over every persistent world experience I've had that involved a 'live GM'.
Posted by: weasel | Jun 01, 2004 at 15:00
It seems like the focus and the thought processes are all focused on treadmill games, where the objective IS to wack some fuzzle, get loot, craft loot, sell unused loot, and level up. In these realms, I also want the NPC to be fully transactional without the fluff. Heck a bot that I can personally choose the automation level, automating any of the above listed objectives, would be great. I can own a fully automated EVE Online megacorp!
I'm think about social realms, like SWG relatively to others, where you fund certain types of live socialization activities to jump start something that may reach critical mass-the point where it will be self-sustaining.
Experimenting with new content for new types of players-undefined player segments-require human dynamics. No scripts, but full impromptu and impervisational entertainment.
It's time to think "out of the box" in addition to improving upon the box. AI NPC still have their uses, but we could use more live entertainers. Experimenting by outsourcing this aspect to more capable groups sound like a logical idea. Any one disagree?
Anyway, the parades at Disneyland always gets my attention. It has no meaning to me, but it does get my attention and subconciously I feel "happier" seeing the festive mood around me. It's that intangible I'm looking after.
We got compulsive, questing, character-building, and economic emotional levers right, so let's move on to foster other emotional payoffs from our VW entertainment.
Frank
Posted by: magicback | Jun 01, 2004 at 15:32
Sort of. I think the resources would be better applied to hiring full-time, professional guildmasters.
Posted by: Jeff Freeman | Jun 01, 2004 at 15:37
One limitation with NPCs and PWs is probably the load a complex AI would create for the NPCs servers. Complexer Ais would mean more computing resources and more servers. So again there is a trade-off.
Posted by: Luca Girardo | Jun 01, 2004 at 15:42
Frank: Experimenting by outsourcing this aspect to more capable groups sound like a logical idea. Any one disagree?
Yes.
If there's a social behavior that will reach critical mass within your playerbase, your players will find it before you will. The best thing developers can do is provide tools that encourages the interactions amongst players that they already struggle to pull off.
Parades/walking parties/old-fashioned stationary parties have been a staple player-initiated activity for years. (decades?) Development money would be better spent looking at how to make such activities: easier to organize, easier to advertise, and easier to perpetuate. Then you can look into allowing player 'organizer's to create competitive games (chess tournaments, footraces, raffles, etc), tiered pricing (guildies drink free! orcs pay at the door), etc.
Hell even improvisational acting itself has been taken care of by the playerbase in emergent-gameplay worlds (UO, SWG). They've had (and I believe continue to have) small but dedicated troupes who put on shows for other players. If one would create systems to facilitate player-owned/rented theatres, ticket pricing, lighting controls, and troupe-management - such activities would surely flourish.
What you're describing is more an attempt to kick-start trends or traditions - and that's more a marketing/community management activity than one that requires actors. That said, I don't know of any community manager who has managed to alter in-game player behavior in the slightest (no offense community managers).
In an emergent-gameplay system, rather than a dozen synthespians working at bargain basement prices, I'd spend the money on a talented coder whose sole job was to make player organization easier. (guilds, parties, performances, scavenger hunts, races, etc, etc)
Posted by: weasel | Jun 01, 2004 at 16:12
EQ tried this once. They had a special server where you could play as a mob. I used to log in there regularly as a bear, or a snake, etc. and go looking for other players. Eventually EQ gave it up as a bad idea since it was effectively equivalent to involuntary PvP.
Posted by: B. Smith | Jun 01, 2004 at 17:36
There is an underlying assumption that access to VW content has to be:
1) On a subscription basis
2) Priced at $15 a month
All of the posts on this thread that touch on economics seem to take this for granted but there is plenty of real world evidence to show that its not true.
Not only does Everquest run a premium server for close to $50 dollars a month but we also see people regularly paying hundreds of dollars on auction sites to buy access to higher level content either in the form of items or characters.
Brask said “Okay, say you have $15/month. Of that, you assign 15%, or $2.25 to pay for a troup of $6/hour thespians…..(intelligent discussion of budget issues snipped)…” and comes up with 11 minutes of content per person per month or maybe 1 hour if your content is consumed by groups.
Lets change the underlying assumption
Okay, say you have $50 a month…
Okay, say you get a one time payment of $250
Okay, say 1000 of your subscribers will pay $100 once every 4 months
Okay, say some one offers you $15,000 for custom content, it’s a speculative commission like commissioning art. What can your company offer for that sum? Custom quests? Carving the patrons face into a mountain ala mount Rushmore?
$15 a month is a dead end for virtual worlds. Value pricing will set you free.
Posted by: Tom Hunter | Jun 01, 2004 at 17:41
** One limitation with NPCs and PWs is probably the load a complex AI would create for the NPCs servers. **
Actually a real AI will simply screw a game and not because of the CPU power. Think to the exploits or the unexpected situations.
When you search for new ideas you are on a journey where you MUST know what's your aim and why. In this case you need to know why an AI could be interesting and the scope of it.
The sperimentation is useful when you are trying to achieve a precise result. You need to know already as you begin where you are heading and what are the results you need to produce. In this case you need to know where NPCs with AI fit in the gameplay and what's the use of that AI.
Posted by: Abalieno | Jun 01, 2004 at 18:33
Tom Hunter wrote:
> Value pricing will set you free.
I think you’re right to suggest that players would pay much more for content than they do today, but unfortunately, that does nothing to prove the cost-effectiveness of live content.
Here’s an example. Assume you have 10,000 players who are willing to pay $500 a year for content. Assume also that a dollar of live content can be meaningfully shared amongst 50 players. If the entire content budget is spent on actors, et cetera, players will receive (on average) $25,000 worth of live content per year. Sounds okay, right? But if that same money were spent on static content -- which can theoretically be shared by *every* player -- they would each receive $5,000,000 of content. The disparity grows as the player population increases. A game with 100,000 such content-loving players could support $50,000,000 of static content a year -- but still no more than $25,000 per player of live content.
Everyone prefers live content to static content. However, if n is the number of players in the game, and s is the number who can share a dollar of live content, then live content must be n/s times more compelling than static content to be cost-effective. When you’re playing D&D in the kitchen, n equals five; in EverQuest, n equals 400,000.
Posted by: Jeremy Neal Kelly | Jun 01, 2004 at 21:42
Jeff> Sort of. I think the resources would be better applied to hiring full-time, professional guildmasters.
If they are company employees, then they are GMs. If they are contractors, then they are a Synthespian. If they are fully independent guildmasters, without performance covenants, then you are financing SIGS. They are all possible actions, but I don't recommend the third option.
Luca> One limitation with NPCs and PWs is probably the load a complex AI would create for the NPCs servers. Complexer Ais would mean more computing resources and more servers. So again there is a trade-off.
A Dev can plot a marginal utility schedule for the cost and benefit of more complex AI. The common law (I don't know who's law it is) is the the more complex the software, the more bugs! Again, they have their uses. Experimentation with an above average AI on a less frequently seen Boss would draw some attention. "That Boss was wack! Was that a GM or some sort of new AI?" Sending the experimentation to some unfrequented test server gets less attention and feedback.
I don't have the analysis of EQ's experiment on playing Monsters, but I guess that without proper management it just ends up being a PvP or griefing tool.
Weasel> The best thing developers can do is provide tools that encourages the interactions amongst players that they already struggle to pull off.
And also give them a slight nudge... toward any direction. But, hopefully dev will have a vision for which direction to go without being the absolute gods they are: so flexible vision.
Once you have many different smaller networks you'll need to get someone to connect the smaller networks to the large whole. This is a Community Manager's role, but most really are not high enough on the organization to take action. All they can do is send the feedback up to the chain of command and hope the game gods agree and the marketing suits to allocate funds.
Weasel> In an emergent-gameplay system, rather than a dozen synthespians working at bargain basement prices, I'd spend the money on a talented coder whose sole job was to make player organization easier. (guilds, parties, performances, scavenger hunts, races, etc, etc)
The synthespians need these tools too. Thus, we can forecast the trend to look like the Mod community. But, instead of funding at the back-end of the quality-cycle (Counterstrike) I suggest funding at the front-end of the cycle: mod competitions with trophies and prize monies.
The difficulty of more and powerful tools is the requirement to give more control of the world to the players. Some are loath to do so.
Tom> $15 a month is a dead end for virtual worlds. Value pricing will set you free.
I agree. The $15 monthly cost just ignore any potential players below that level of utility. People who do pay the $15 are the relatively more hardcore, thus the direction of gameplay.
So, different visions will require different financing and cost structures. It's really hard to be first to try something new. The recent news about There.com, a well-financed dev, is discouraging to would-be innovators.
But at least talking is free.... and cheap :)
Frank
Posted by: magicback | Jun 01, 2004 at 22:33
Jeremy> Everyone prefers live content to static content. However, if n is the number of players in the game, and s is the number who can share a dollar of live content, then live content must be n/s times more compelling than static content to be cost-effective. When you’re playing D&D in the kitchen, n equals five; in EverQuest, n equals 400,000.
Jeremy, I don't dispute your numbers based on your assumptions, but do offer another perspective: the balanced offer.
Let's view the coin-op arcade as static content. You can run a business on this model. It's a steady cash flow business.
Let's view theme parks as the "balanced offer". You get a mix of static contents (they got arcades and rides) and live contents (shows, parades, and stuff). Theme parks are a conglomeration of lots of different activities.
The reason I think theme parks are larger than coin-op arcades is the scale issue and the the combined effects of economy of scale and conglomeration. Sure you can have Arcade World, the biggest coin-op arcade in the world, but I haven't seen it.
Thus, your model based on your assumptions come to the right conclusion, but there are still lots of stuff missing in the equation. I don't know what they are, but I do think that the value of static content (as in fedex quests) is getting lower and lower each day while the pricing power of live content or recorded content (like comics and videos) are getting higher.
Frank
Posted by: magicback | Jun 01, 2004 at 22:47
Ok, I am getting too worked up...
Here's another idea to increase the cost/benefit equation.
You finance a talented group of students to create cut-scenes and other machinima. They can be placed in the world as part of a hybrid of static/synthespian content.
Example: you touch a magical pool and you're shown a short and small cut-scene. You go to a in-game movie theater and get to see an in-game movie.
Example: as a Synthespian you sorted and edited through all the replays of a major combat and create an instant replay like a sports event. Initially, players can only see the "Telling" from a Syntespian. But once they seen it, they have already download it to their computer and now can share it with others via P2P.
These are more structured hybrids which probably will get pass the board of directors better than impromptu live content that only a few will see (the issue of the movie version vs. the play version).
Frank
Posted by: magicback | Jun 01, 2004 at 23:08
That is complete and utter bullshit, and you are wrong.
Posted by: Jeff Freeman | Jun 02, 2004 at 02:59
Frank> Experimentation with an above average AI on a less frequently seen Boss would draw some attention. "That Boss was wack! Was that a GM or some sort of new AI?" Sending the experimentation to some unfrequented test server gets less attention and feedback.
YOu would assume in that case that you have a flexible software design for the NPC subsystem in the architecture. So flexible that just with the change of a script you could alter the AI behavior of the NPC. But it does not happen so often. At least for what I have seen till now.
Posted by: Luca Girardo | Jun 02, 2004 at 03:45
Just as an aside, the way we're using the term "sythespians" here (real people who act in a virtual environment) is not how it's used generally (ie. virtual people acting alongside real people).
Richard
Posted by: Richard Bartle | Jun 02, 2004 at 04:01
Frank> The synthespians need these tools too.
And that is why they're redundant. Players will invent, and organize by themselves. Using thespians of any sort to 'kick-start' such activities isn't necessary or cost effective. Since supporting the thespians-as-community-encouragement requires just as much special attention as simply supporting the playerbase's desire to organize itself - they provide no cost benefit.
Not in a content production sense, and not in a community 'encouragement' sense.
Make a tool that allows players to form 'Parades' and register with a town (to cue the NPCs to line up/throw confetti/etc) and players will do it.
Posted by: weasel | Jun 02, 2004 at 10:06
Weasel,
I agree that they are redundant if the level players will invent, organize by themselves and produce “social goods” is high. Except for SL, my experience is that guilds/group self-organize in most part to aggregate power.
I also agree that on paper the costs out-weight the tangible benefits, but I am going after the intangibles. I’m thinking from the perspective where a mall will arrange a concert, book signing, or a special event to get more traffic. This idea is not proven in this space, but I think it has been tested extensively either.
So, incorporating all the opinions so far, then the hypothetical game will have the open architecture of SL and the gameplay of WoW. In this system, aspiring Synthespian (employees, outsourced contractors, or players) would have an incentive structure to create content.
Richard,
I use the term interchangeably: real people portraying virtual people and, eventually, virtual people portraying real people.
Luca,
New MMORPGs already have very good NPC scripting engines. But after reading the postmortem on Neverwinter Nights, I think the hard part for massive games is balancing all of the different emergent NPC behaviors. The devs at EVE have thought about implement what one called Dynamic Story System (DSS). The guy successfully implemented the system on a single player game, but it was much harder to implement on a massive persistent world.
So, I go back to what Bill Roper, ex-Blizzard, said 3 years about about World of Warcraft:
Bill Roper: Yes, we have Gamemaster in the concept, who will support players and also control NPCs, but our goal is to have a strong scripting engine that gives NPCs a wide range of possible reactions. We are taking a lot of time to write strong tools and scripting processes so that we can maintain the world with as few people as possible. Our main goal after completion of the game is to create new content and not to keep the game running through our help.
So with the right tools, GMs can play a Dungeon Keeper-type game against Players [evil grin].
Jeff,
Huh? You explain to me how my simple three categories don’t work? They are simplistic, but they do fit from my perspective.
Frank
Posted by: magicback | Jun 02, 2004 at 10:46
Because a person's job description isn't determined by his employment status.
"If they are company employees, then they are GMs" is demonstratably false. *Some* company employees are called GMs by certain companies. They're the ones that have "GM duties" assigned to them, meaning in-game customer service. Other companies refer to those people as "Customer Service Reps".
Companies have employees whom are customer service people, community relations, quality assurance/testers, producers, designers, artists, managers, whatever the people on the events-team are called, and so on. It's worthless to apply the term 'GM' to all these people. The term itself is pretty bad, too, since 'guildmaster' is also abbreviated to 'GM'.
When I say you'd be better off hiring full time professional guildmasters, I mean you'd be better off hiring people, to work as full time company employees, people whose job responsibilities would be best described as 'guildmaster'. They'd be the guys coordinating associations of players, scheduling regular guild events, recruiting, and working with the live events team and the design team to incorporate the guild into the game world.
Players do much of this for free now, but the game world is largely unaware of the player's guilds and there's not enough bandwidth to make all the npcs cognizant of all these various ad hoc guilds, especially since they don't exist on multiple servers (we can make all the npcs aware of 'the warrior's guild', but not of the 'kewlioz' guild that only exists on one server, etc.).
Also, the quality of guildmastership is hit or miss (some of them just don't ever organize guild events to entertain their members), and since the guildmaster isn't an employee of the company he can one day say "Screw this, guys, lets all go play that new game."
So I'll repeat:
* We'd be better off hiring full time professional guildmasters, if we're going to hire anyone. They might make enough difference in retention to pay for themselves. I'm not convinced they would, but if I were to hire a bunch of folk this would be what I hired them to do. And even then I wouldn't hire as many of them as there are player-gms now. They'd have to structure their guilds into sub-guilds and delegate.
Now, many companies already have a live events team of some sort, and they do some roleplaying as part of their job. I gather *this* idea is to hire *lots* of those people though.
* Hiring scores of actors is a really terrible idea. Probably one of the worst ideas since 'Permanent monster death' got pitched on MUD-Dev.
I don't know how to be more blunt than that without hurting people's feelings, so I'm going to flee from this conversation now.
Posted by: Jeff Freeman | Jun 02, 2004 at 11:43
Jeff,
I see your point. What I meant was that if professional guildmasters are hired, they are in essence employed "gamemasters", who perform broad in-game duties. This is the category (out of three) I collectively place them in. The other two could be categorized as "freelancers" and "funded interest-group leaders."
If I am readying you correctly, you are saying that if more staff are to be hired to perform in-game functions they should be allocated to more guildmaster-related functions. This, in your opinion, may have relatively better, but still negative, ROI than hiring in-game actors. I don't dispute this.
My idea is not to hire scores of actors.
My ideas are for VW operators to focus on their core strengths and:
(1) outsource live content; or
(2) fund and support "freelancers", independent groups, in creating live content.
One can look at the two ideas as open tendering of outsource contracts to produce live content. If the company decides their world needs actors (Synthespian in the strict sense), that's ok.
My ideas are much broader. Hope this clears things up and bring you back to the conversation.
Frank
Posted by: magicback | Jun 02, 2004 at 12:17
Do you really think professional guildmasters is a good idea given many players' inherent distrust of anyone receiving a dollar from the company? Maybe for n00b or stock guilds, but I really hope you're not suggesting that all guildmasters would have to be employees.
Employees working full-time to support and empower guildmasters might be a good idea.
I see eye to eye with the weasel here. To wit:
"NPCs are particularly delineated from players because they aren't intriguing characters in the grand stories that RPGs are all about. They aren't supposed to be engaging or interesting. The story isn't about them, it's about us. If the shopkeeper is really interesting, if she's more interesting than we are, she becomes the actor and we become dull and disposable."
I'm all for more dynamic and expressive NPCs, ambiance, etc, but hiring actors would be an ill-advised use of funds imho. Instead of telling me your story, let me tell mine.
Posted by: Staarkhand | Jun 02, 2004 at 12:46
Frank,
Ok, we're not that far off. It's a matter of saying, "Ok, you've for x-number of new employees, what do you want them to do?" Whether they're outsourced, contract, full-time, whatever, isn't really relevant. I mean, it'll be relevant to someone, but he's a suit and that's his problem. I'd use them to create content by having them patron guilds, groups of guilds, and so on. I *wouldn't* have then man the booth in the tavern.
I'm not convinced out-sourcing is a great idea. Politically, I mean, you have so many obstacles you have to overcome there. They *have* to be able to work well with the design team. Getting a design team to work well with an events team is really pretty challenging. And I don't think your systems should be that divorced from your content. Meaning, say you have a design team working just on systems, game balance, adding new zones or whatever - all that *is* content. Would you outsource that too? Seems like at that point you're just hiring someone else to run your game for you.
Staarkhand:
It would be impossible to force players not to form their own guilds, and have their own guildmasters. So, that's still going to happen.
But yeah, "stock guilds", the in-game type currently staffed by mindless vendortrainer NPCs who don't actually *do* anything - I'd put real people in charge of that, and their job would be to "support and empower" the player's guilds.
You're still going to have players form-up into their own cliques, migrate into the game as a guild and remain a guild, but you can put a man on-staff to act as a sort of 'patron' for the player's guild.
Trust issues can be overcome, I think. Even if it means using some smoke-n-mirrors to make it appear that animosity between various guildmasters (co-workers) exists where it actually doesn't. It's a good idea to manufacture that sort of conflict anyway, and more immersive to not to let-on that it's just manufactured.
But these folks primary job duty wouldn't be 'to act', but rather to do the more organizational type duties that player-gms current handle (or mis-handle, depending).
And no, I'm not convinced its a *really* good idea either. It's just not a terrible one.
Posted by: Jeff Freeman | Jun 02, 2004 at 13:13
Surely having a player playing a non-player character (NPC) is just a contradiction in terms? Isnt the whole draw of MMORPGs over single player games, the fact that some of the other characters have humans working them, and not an AI? Every player is already a part time actor for the other players.
Perhaps what you are really saying is that more ingame functions need humans to be able to run them?
In games where player-crafted items are actually useful/valuable, then some of the merchants are already PCs and not NPCs. But this is dependant on the thing that the psuedo-NPC delivers, being valued in the game system. A stripping cat in FF9? An art gallery in There? A raid organiser in EQ? The synthespians are already there, but are getting paid in IC currencies, both money and reputational.
Maybe the trick is to look at other ingame functions and see how to provide payoffs to players for performing them.
The karma/pvp system of Lineage2 appears to be an attempt to reduce the requirement for GMs to react to griefers, by encouraging the players to deal with them, and providing game functions to assist in identifying them.
My recollection of the player-running-EQ-monsters was that they were only deployed on the test and PvP servers, where there was already a presumed assent to random and dangerous behavior from the enemy toons.
But this begs the question of how much randomness we want from our NPC AIs? Would we prefer the questgiver to sometimes randomly change the quest? We like complex monster AI and pathing, but would we want the boss to arbitrarily pick one player to kill over and over (as a human directed monster might choose to do?)
PvP delivers the unpredictability of human opponents for fighting encounters. Player to player trading does it for the economic arena. I am less clear on how it could be done for the cultural and artistic realm (in an environment where there may be no tangible payoffs for such contributions)
Posted by: Estariel | Jun 02, 2004 at 13:28
Jeff> the game world is largely unaware of the player's guilds and there's not enough bandwidth to make all the npcs cognizant of all these various ad hoc guilds, especially since they don't exist on multiple servers (we can make all the npcs aware of 'the warrior's guild', but not of the 'kewlioz' guild that only exists on one server, etc.).
I fail to see where this would be a bandwidth issue.
While the NPCs certainly couldn't be made aware of individual actions of the players within a guild, an aggregate faction standing for the guild writ large would be quite feasible and straightforward to implement.
The fact that player guilds don't span shards is, if anything, simplifying the lookup problem by resulting in a smaller set of records in the player guild table (smaller than if guilds did span shards).
It's essentially the same process borne out every time a PC prompts an NPC in Everquest (either for conversation, combat, or by presence). The only difference being that SOE has their NPC checking only the PC's character record for faction standing to determine its reaction - rather than possibly consulting a guild record as well.
Posted by: weasel | Jun 02, 2004 at 16:43
Not literal bandwidth.
True, and I'd do that before I'd hire humans to do it. But if I were to hire humans to do it, I'd expect them to cater their dialog, demands, quests, conflicts and so on to be better than what a machine can do automatically.
For that particular sort of implementation, yes.
A human guildmaster would need to do quite a bit more than this to justify his job.
Better example would be a guildwar, or even some other sort of indirect conflict. You'd be hard pressed to automate a method by which one player-guild comes into conflict with another player-guild. But with your own employees 'leading' those guilds, it'd be pretty easy to establish a context for the conflict.
With willy nilly ad hoc guilds that differ from server to server, you can *still* find guilds that ought to be cooperating or competing with other guilds, based on their 'flavor', let's call it. But that requires either a self-flagging system the guilds use to identify their flavor to the system, or a lot of research on a server-by-server basis.
The bandwidth I'm talking about is dev' time to script convos, quests and add content specific to certain guilds, and whatever narrative they're being steered into. It's just a LOT easier if that can be pushed-out to all servers, exactly the same.
Not impossible with the various other guilds and such, but easier not to worry about it.
And potentially, it helps stick the guild to the game, so they can't just depart as a group when the next big thing comes out.
I'm not really the person to argue in favor of employee-run guilds, though. I just don't feel that strongly about it.
Posted by: Jeff Freeman | Jun 02, 2004 at 17:16
Outsource vs inhouse.
I think there are arguments for both sides, but as inhouse is the default action, I'm pushing for outsource. Managing virtual teams is not a easy task and there could be security among other issues, but hey it could be done right.
Nevertheless, while I think reality is stacked against outsourcing at this time, a small discretionary funding to experiment on this front seems appropriate.
I agree that any paid staff should work to benefit the overall gameplay (say guild administration) instead of manning a bar...well unless they are as good as the actors on Cheers and I can watch reruns on my in-game TV.
So, let me argue more for contracting freelance Synthespians:
Let say you want to introduce a server-wide event and you need people to play key NPCs. Who do you want to play that role? (a) a team member with the best RP skills or (2) a freelancers with better RP skills? Discuss.
Frank
Posted by: magicback | Jun 02, 2004 at 22:08
Jeff> The bandwidth I'm talking about is dev' time to script convos, quests and add content specific to certain guilds, and whatever narrative they're being steered into
Well then that's where our preconceived notions of design diverge. I'd sooner just create content that's available to certain ranges of faction approval.
E.g. The NPC Trading Guild in Littleton would offer a player with +400 faction a delivery quest, while a player with +600 faction could choose between the delivery, or an escort quest.
Again, specific reaction to players or player guilds wouldn't exist, but I don't think it's necessary.
Though I agree with your overall point. If i were to hire these theoretical low-wage content providers - I'd put them to work much like you would long before I'd send them out to serve beer and make proverbial balloon animals for the players. That said, I'd likely never spend the money on either. I'd spend it on content available to everyone.
Frank>Let say you want to introduce a server-wide event and you need people to play key NPCs. Who do you want to play that role?
It wouldn't matter whether they are a 'team member' or a 'freelancer'. You'd pick someone who has the right skills at the right cost for the job, If you were to run a server-wide event that required live actors.
Maybe there's a dev on your team who'd do a good enough job and would cost effectively nothing. Maybe you ask a friend or colleague to do it gratis, or maybe you do hire a gaming SCA type. Their employment status has little to do with the decision - only their cost and capability.
Of course, I'm of the mind that 'live' content should never (or at the least extremely rarely) be one-off improv. It's just not cost effective, and it's exclusive by nature.
Posted by: weasel | Jun 03, 2004 at 09:30
The "live-action" RPG run by the Lorien Trust in the UK has NPC guildmasters. They are unpaid, but are able to attend events for free. They are recruited from among the most experienced players. Their not-quite-employees not-quite-players status leads somewhat predictably to conflicts of interest and accusations of abuse. The turnover rate can be quite high
Posted by: Estariel | Jun 03, 2004 at 11:41
Estariel,
I familiar with Lorien Trust. The perception of conflict is enough to start accusations.
The LARP I am involved with have a feudal structure where a player can prove themselves worthy to be a noble in charge of a productive land. These nobles play a realm management meta-game that affect the fate of their citizens and major plots. There are a fair amount of levels that makes it their priority to keep their citizens and their competitors "happy". There are still abuses and high turnover rate for certain land. However, a few player nobles have raised their ranks over 10+ years of play (and this is with permadeath and full PvP). His early courage won over the citizens, but as he (the character) is on his last death he has become quite conservative. He still hold his position as he has loyal lieutenants that are happy to have him as a figurehead :) Anyway, just an nice anecdote.
Thus, the management decision is:
(1) rely on internal pool of talent,
(2) rely on external staff with no performance covenants, but very low cost, or
(3) rely on outside help if they are more effective then the other options.
Feel free to add options I didn't cover.
Frank
Posted by: magicback | Jun 03, 2004 at 12:16
The basic problem is the people skills of the people you let be NPC. Some people can be extrovert active and wonderful, alert to other players, new people and problems, these people do not even need to be NPC's to do this, they just find a way of doing it.
However as in the example mentioned in the previous post then in the real world the NPC's are selected not for ability and their people skills, but just for knowing the right people often through family, or going out with people and social circles. In the example the majority of NPC's are male, too often they only NPC for the perks such as free entry playing only lip service to the vast amount of work being an NPC involves (leading the wonderful ones do more than their fair share and sadly too often burn out). IN LARP many of the poor NPC's have a tendencies such as only 'helping' new female players settle in, sometimes justifying such actions claiming they are for example deliberate attempts to increase the numbers of female players in a male dominated hobby, this is not a fair point as positive discrimination is still discrimination. They do not aid ordinary male players who are trying to get involved.
Some of those kinds of events have a history of using powergamers who they have noticed because their powergaming has brought them into contact with the people at the upper levels of the game and organisation. As powergamers they do not have any affinity with the ordinary mass of players, as such they are not good people o be got involved in a role which is meant to help others. Even worse is that too often they have brought their prejudices against others into how they act towards them. This has in the past led to feelings of elitism, racism (as some people have views which they have expressed when undertaking their NPC role) and discrimination, especially amongst ethnic minority gay and female players (female players sometimes feel that all the extra attention they receive from the NPC's over others is quite frankly creepy).
Conversely sometimes volunteer NPC's works great, some systems especially those which have lots of smaller events can identify good people with commitment and the skills to make the event better for everyone, including both long term and recent players. As these events are smaller players fall into helping and getting involved by 'accident' right from the first few minutes when they arrive, new players are not just pushed out by the mass but quickly feel they are a part of the event. In a computer environment the selection process is the same, a person selects people, and that persons view of the people they are selecting is biased being drawn on their experience of the person and not on the persons suitability.
Posted by: Silversmith | May 24, 2005 at 17:34