Elderly Dev Wakes Up Long Enough to Prattle A Bit About Some Stuff

[Signe]

RP vs. G

Massively Multiplayer Online Role-Playing Game, that's what Lord of the Rings Online: Shadows of Angmar is. Massively Multiplayer, which distinguishes it from games that support mere dozens of players. Online, demanded by the Department of Redundancy Department, since Massively Multiplayer pretty much implies being online. Role-Playing Game: ah, finally we're at what kind of game it actually is. In RPGs, players take on the role of a character in a different world, and play within that world according to defined game rules.

For some, "RP" means a great deal: they take the roles seriously, and want the escapism of truly acting out another persona. For others, "RPG" may as well be Ruthless Power-Gamer: they drive their avatar around accomplishing the goals in the game in a mechanical fashion, speaking in their "real" voice and associating with other players as players, not as characters. There is a strong tension between these two factions, as each has different motives and goals within the game. What is a developer to do?

What is Role-Playing? And what should the game do about it?

It's odd how "role-playing games" will have "normal" and "role-playing" servers: it's an admission that there is a fundamental tension between the "RP" and the "G" in the genre. Being a game means that there are a priori rules that govern success and failure within the game environment. Conversely, there are no rules governing what makes for good role-playing: role-playing takes place entirely between players, who determine for themselves the quality of the role-playing experience. Without formal rules, role-playing is essentially not a game: it is performance, and the world and systems we design need to recognize that distinction.

Should a game attempt to cater to role-players via dedicated game rules or systems? I am of the opinion that games should not. For one thing, any game distinction between "role-players" and "typical players" is often a magnet for griefers. RP chat channels, visual indicators of RP or Non-RP players, RP "ranking" systems, and RP banning policies all introduce more problems than they solve for RPers, for they are all accessible to people who like to disrupt them. Even RP servers generally fulfill their roles with mixed success: many players join them solely to interfere with the serious role-players, or often simply for network latency or server population reasons.

It is of minor consequence to provide a role-playing chat channel or a role-playing server if the players ask for one, but as a practical matter, it has been my experience that these things generally hurt more than they help. They're relatively easy to provide, but represent a "be careful what you wish for" choice, under even the best of circumstances. They cannot be effectively policed by customer service, because as I said previously, there are no clear-cut rules by which role-play can be efficiently and effectively governed.

Another reason is that any code-governed system dedicated towards encouraging role-play as its goal can itself usually be gamed in a non-role-play fashion. If there are tangible benefits to being rated as a "good role-player," then gamers will find a way to achieve that rating without putting in the time and effort of actually role-playing. If there are no tangible benefits ("You get a gold star!"), then the system seems cheap and irrelevant and will generally be ignored.

But those are practical rationales: the more fundamental reason is that role-play is entirely a function of players, not code. Role-play is about people connecting with each other, not with bytes of code or icons. It is about turning what might otherwise be isolated imagination into a shared experience. There are no rules or systems that truly make that more meaningful: it is the personal interaction that makes it significant.

So what can the game do to help role-players?

Give players a world worth role-playing in. Make a Middle-earth that is attractive and interesting to role-players. We are working to make the world rich, weaving the lore of Middle-earth throughout the game, so that players are willing to invest themselves more fully into the game experience. An accurate Middle-earth is a big step in that direction, and accuracy in depicting the world is our priority in building the environment of the game. The Ivy Bush pub will be on the Bywater Road; Athelas plants will have long leaves; the hobbit holes of Bree are higher up the hill, above the houses of the men below: every detail will be as accurate as we can make them, and things that are unmentioned in the books will be made to fit the feel of Middle-earth as best we can.

Graphics cannot make a bad game into a good game, but they can help create an immersive environment, and we're striving for a deep, convincing representation of Middle-earth: from the farms of the Shire to the underground cities of the Dwarves, everything is being built and polished and detailed all with the ultimate goal of creating a Middle-earth to meet the scrutiny of fans of the books for years to come.

Taken further, the story woven into the game is intended to not only provide good direction for adventuring gamers, it is meant to be narratively fulfilling, providing fertile ground in which the seeds of role-play can be planted. "Sandbox" games, where the world and systems are just plunked down and players are supposed to construct their own drama, have been shown to be difficult scenes for players to construct compelling stories in. Our hope with the epic story of LotRO is that there is context to all of the player interactions with the world, such that players have more to latch onto in playing out their own characters.

The best a designer can (and should) do for the dedicated role-player is to provide them with a world that meets their needs and expectations, and then get out of their way.

--Floon Beetle

http://lotro.turbine.com/?page_id=96

[brian]

It's interesting to compare that with this from over at Skotos: http://www.skotos.net/articles/guestvoices3.phtml

The primary impetus of a dramatic story is conflict, and while an internal conflict can provide a fine reading experience, interactive storytelling requires the presence of external conflicts in which two or more players can participate to create stories that are effective and mutually satisfying...

- Brian

[Signe]

Hey!  Thanks for that link.  It looks like an interesting read but it'll have to wait until tomorrow when my eyes can stay open.   

[Fargull]

The article is interesting, and I am currently of the opinion that Level is wrong in more ways that just division of the player base, it narrows the gaming world as content becomes obsolete.  It is a funnel, not a journey.

I would also like to see a world outside of crafting/kill foozle fifteen.  Why can I not ski down that virtual mountain?  Or go surfing?  Or race horses and have virtual betting?

[AOFanboi]

The article is interesting, and I am currently of the opinion that Level is wrong in more ways that just division of the player base, it narrows the gaming world as content becomes obsolete.  It is a funnel, not a journey.

It's also a virtual money sink for the games where equipment also has "levels". Which appears to be all of them.

(I never understood why the baddies in the 40-50 level range zone didn't invade the cozy newbie (1-10) zone and 0wn everything. Your "funnel" is as good an explanation as any.)

[HaemishM]

The game systems DO need to support role-playing, or any role-playing you have will be hopeless mental masturbation, or mutual mental masturbation of the gayest variety. And by game systems supporting role-playing, I mean people need to be able to have a tangible effect on the world around them and they need to have to face the consequences of those effects. Kill all the furry critters in the area? No more fur, which means no more fur-lined armor. Also, people need to be able to effect each other.

But you can't have any of that in MMOG's or even MMORPG's, because people don't pay for consequences, they pay to have their shiney handed to them by Frodo.

[schild]

No game system that has stats on weapons will really support roleplaying. Particularly when all of the images come out of an artists mind. Real roleplaying will happen when someone invents a twitch game and weapons act as they look. A Soul Caliber/Ninja Gaiden MMORPG would rule the world.

[Samwise]

With zombies!  And roller coasters!  And ice cream!

[Fargull]

I mean people need to be able to have a tangible effect on the world around them and they need to have to face the consequences of those effects. Kill all the furry critters in the area? No more fur, which means no more fur-lined armor. Also, people need to be able to effect each other.

But you can't have any of that in MMOG's or even MMORPG's, because people don't pay for consequences, they pay to have their shiney handed to them by Frodo.

Everyone effects everyone in a MMOG, issues logging, the now common email issues, camping, idiots in general chat, people in position to make decisions need to stop being pussies and follow the logic that people don't pay for consequences.  I think that is bullshit.  Hell, the level grind is paying for that...  I would just much rather pay to have that fur hunting dry up and the armor going poof than having to grind up the levels.  I would like to not watch the game world go from 100% useful to 10% useful.

[AOFanboi]

I think the devs and players have different opinion when it comes to "persistent world".

A player dreaming of what he wants the MMORPG to be wants his character's actions to persist: If he heroically kills Ogre-King Fuglybastich, Fuglyubastich should be dead the next time he logs in. This perception is usually based on years of playing single-player games.

To the developer, this is an antithesis because this would mean player A's actions would prevent player B from experiencing that content, for all B coming after A. This would either mean a shitton more content than reasonable, or many bored players cancelingas more and more content was "completed". So they instead interprets persistance as: the world persist in spite of player actions; Monsters respawn, quest NPCs give the same quest to a million players (literally for WoW), landscape is unmoving, etc.

Guild Wars seems to manage to land somewhere between the two, where "the world" only holds the state relevant for your character: NPCs and monsters for your quests, monsters not respawning (until you respawn the entire instanced "sub-world" by reentering it), and where the progress moves you into a different "phase" of the world (pre- to post-Searing Ascalon in particular).

[Margalis]

I think devs know what people mean by persistence. I just think it's hard. It's pretty simple to make it so that if you kill a secluded boss he is dead to you (but alive to other folks) but beyond that it gets really complicated. Whenever people can change the world they can also create problems. And changing the world also means an entire layer of real interaction that would have to be built into most games.

[Evangolis]

Sleeper's Tomb.

[tazelbain]

Guild Wars seems to manage to land somewhere between the two, where "the world" only holds the state relevant for your character: NPCs and monsters for your quests, monsters not respawning (until you respawn the entire instanced "sub-world" by reentering it), and where the progress moves you into a different "phase" of the world (pre- to post-Searing Ascalon in particular).

GW makes no real progress on this issue. Sure "progress moves you into a different phases" but you can always go back to old phases.  It's really just static.  A.net has talked about all the cool stuff they could do with persistance, but as far as I can see none of that made it into the game.

[SomeKindOfMoron]

I agree with a lot of what this Floon guy is saying, especially the last line.

The best a designer can (and should) do for the dedicated role-player is to provide them with a world that meets their needs and expectations, and then get out of their way.

This is true because, firstly, players will ultimately never care much about a world so static as that of WoW or EQ, where the world only can ever change when the developers want it to.  Yeah, the graphics can be gorgeous and everything can reek of lore, but so long as the players are unable to institute any meaningful change, they will never have stake in it as a world and as a place. 

Second, as the Skotos article that brian linked to said, for "interactive storytelling" (or role-playing, if you'd rather call it that) to develop, you need to facilitate conflict between two or more players.  Yeah, you do already see conflict between players in the EQ-clones in the form of competition for resources (loot, mobs) and that sort of stuff, but that's not exactly the stuff of legends.  PvP ganking is a little bit more compelling, but still not quite there.  We need to see stuff on the scale of epic wars and struggles between player-run organisations wherein resources with in-game significance can actually be won and lost.

So, as I see it, for "role-playing" or "interactive storytelling" or whatever the hell you want to call it, you need a world with meaningful persistence and a compelling outlet for PvP.  You don't need swarms of AI-controlled beasties and increasingly uber armor and dungeons that require 40+ players to win. 

All that said, to go back to this Floon guy's article, I find it pretty amusing that, as much as I agree with him, the solutions he provides for the "role-player" question are pretty disappointing.  A game world steeped in Middle-Earth lore and "immersive environments".  He does mention player interaction having "context", but who knows what that will actually end up meaning.

[Alkiera]

Sleeper's Tomb.

By this you mean, what? 

Sleeper's Tomb is an excellent example of how to badly implement a player-changable environment.  It's a zone very very few people ever saw, and was filled with mobs that dropped the best items in the game at the time... and when 'changed', dropped relative crap.

Having players be able to effectively remove items from the drop tables for other people in a lewt-based game is just stupid.  While it was a rather impressive idea at the time, it, like many design decisions in that era of EQLive, was not thought out very well.

Alkiera

[Evangolis]

Sleeper's Tomb.

By this you mean, what? 

Sleeper's Tomb is an excellent example of how to badly implement a player-changable environment.  It's a zone very very few people ever saw, and was filled with mobs that dropped the best items in the game at the time... and when 'changed', dropped relative crap.

Having players be able to effectively remove items from the drop tables for other people in a lewt-based game is just stupid.  While it was a rather impressive idea at the time, it, like many design decisions in that era of EQLive, was not thought out very well.

Alkiera

See, you understood exactly what I meant.  Anyone who doesn't understand that comment needs to learn to understand it, because that episode embodied so many of the basic flaws in persistantly player-changable environments.  If you can't move the environment back to any alternate state, you are going to have the kinds of problems that Sleeper's Tomb had.

To go futher back, think about the Defense of Trinsic in UO.  The city was forced to fall on the one shard that successfully defended because of miscommunication between programmer and designer (See Mulligan and Petravsky's book), but I suspect that if that that shard had held out, people would have become as unhappy with it as people were with Sleeper's Tomb.  Which is also about how unhappy they were with the forced fall, as I recall, so which is better, I wonder?

[Fargull]

All good arguments; however, I think Sleepers Tomb could be argued that it is the wrong kind of content for a persistent world.  I still think a viral quest system would be a better solution.  Such that slaying x creature has a chance to spawn a quest object that send the character or group off on x goal that might then spawn further goals.  Or random npc x is engaged by a variable on the character causing again a virual quest to spawn.  The meat of the system would be the engine behind the content, not the meat being the content.  Sleepers Tomb is not a MMOG based content, it is a single player (or small group) based content.  Same with most of the raid crap.  The fact all of these diku's are aiming for the raid puke is what turns me off of them for the most part.  Not that such encounters are not fun, but that getting to them funnels out 90% of the rest of the game.  Especially after 1 iteneration.

[Descended]

My response got changed after I read Fargull's post.  I was going to post a response that boiled down to "can today's server infrastructure profitable support a WoW-style world where every combat/quest subzone is statically instanced, allowing your character and perhaps some invited groupies to permanently (for you) kill NamedFoozle003?"  Upon pondering, this feels wrong -- the point and attraction of univerasal persistance isn't just that your character can make a visible mark on the world noticable by you, but that your character can make a visible mark on the world noticable by everyone.  Bragging rights make the hero.  Then I read Fargull's post and remembered something that Mu (site not updated in forever) was commenting about when working on a NWN persistent world with a realistic ecology.

A fantasy virtual ecology of heroes, villains, monsters, nobles, and peasants could have a relatively stable core of civilization (the king of kingdom A is likely to still be king a real world month later) but still have a volatile shell of action.  Imagine how the following set of 'ecology zone rules' might affect your involvement with the world.

ExampleZoneA has a small trading town near it's center.  The town depends mostly on trade for its income.  Characters can get escort missions to escort caravans to adjacent towns (oh noes! boring escort missions!).  ExampleZoneA also has a population of Bandits.  Bandits eat Caravans.  Escort missions take about half an hour.  At ten minutes and twenty minutes, the caravan stops and rests.  Guarding characters have to report to the caravan captain at these stops.  Between those stops, the character can do whatever he wants, wherever he wants.  If bandits attack the caravan, losses taken will reduce payment given.  Bandits aren't spawned at set points to ambush the caravan -- rather, bandit lords order attacks on caravans, and the bandit lords depend on the reports of bandit scouts in order to order attacks.

So the character has options.  He can go find and kill all the bandit scouts along the caravan route before the caravan passes by.  He can kill the bandit lord that the bandit scouts report to.  He can range ahead of the caravan to take out bandits laying in wait, then report back at the set times. He can guard the caravan the whole way and just kill whatever threatens the caravan.  He can build faction with the bandits and offer them gold to leave a particular caravan alone.

And since this is an ecology, the AI running the NPCs choice trees needs to make sense.  Demand for goods in neighboring towns needs to be tracked.  If this statistic isn't tracked, then the merchants in town might as well run on a fixed schedule, regardless of their losses or successes on previous caravan trips.  Merchants need to be smart enough to say "No, demand for Generic Good C isn't high enough to warrant a trip to Neighboring Town B, considering recent bandit activity, unless you, Mr. Hero, are willing to guard us for a payment of X gold".  Or:  "Demand for Generic Good D is so high I have to attempt a trip to Neighboring Town C, I'll pay you X+3 gold to guard my caravan!"

For this whole scheme to work out for the player, there are a couple of things that have to work smoothly but silently.  First, changing player populations within the zone have to immediately affect the ecology of the zone.  Increased player count should cause an increased number of caravans leaving town per hour.  Do this by having population numbers generate a multiplier on external goods demand or whatever -- it is important that the number of player characters (perhaps limited to non-AFK and the right level ragnes) in the zone affect the availability of ecologically generated quests.

A second silent but necessary factor is that all the parts of the ecology need to be able to shrink or grow logically.  The population of bandits needs to drop if a fair number of players log out for the evening and, thus, the number of caravans leaving town per hour goes down.  If there is a high number of caravans leaving town per hour, the bandit population max needs to go up, and perhaps even the bandit respawn rate.  If a bandit lord is killed, but the majority of his bandit camp is still alive, a new bandit lord should respawn more quickly than from a decimated bandit camp.  If all or most of a bandit camp is wiped out, the bandits should be more likely to adjust the location of their camp while they recruit new bandits.

This is an interesting system -- at least it provides a slightly varied play experience, and allows some player choice in how to make money off of caravans.  However, more is needed to allow a character to gain fame.  Another layer of more unique events needs to applied to allow players to feel they are special or to give them a goal to work toward.

Suppose the town mayor or guard captain gave out conditional quests, depending on relatively infrequent conditions.  For example, if a particular caravan route suffered 100% caravan losses for three trips in a row, and the bandit lord in that subzone stayed alive the whole time, the bandit lord might proclaim his studliness on /shout (and perhaps become a named bandit lord, using some table of randomly generated names, like the unique bosses in D2), and the guard captains on either end of the caravan route might post town cryers or wanted posters indicating rich rewards for the bandit lord's death.  Alternately, if bandit lords were being killed off at a high rate, a bandit might show up in a seedy tavern, bemoaning how 'the bandit lords ain't what they used to be' and offering a quest to characters with high bandit faction to become the new bandit lord of some particular group of bandits for as long as the character could stay alive and with the bandits (this obviously suggests the best way to implement PvP in this type of world is to have aggressive targetability dependent on the faction standings of the attacker, the (perhaps temporary) faction membership of the attackee, and the current zone/subzone).  Zone broadcasts or /shouts regarding such events would make the player feel like the center of attention, which is generally all that is needed to make someone feel better-than-thou.  Implementing visible-to-others faction bonuses, like bowing guards and /say greetings from important officials, rather than just price discounts or equipment availability also increases the sense of the heroic.  Slowly degrading faction standing might be important, too, to avoid the feeling that your level 50 character who hasn't even seen the zone he gained fame in at level 15 for two or three months ought to still be just as famous there.

Hrm, I'll finish this up later; I got most the points I wanted to make in, anyway.

[Glazius]

A second silent but necessary factor is that all the parts of the ecology need to be able to shrink or grow logically.  The population of bandits needs to drop if a fair number of players log out for the evening and, thus, the number of caravans leaving town per hour goes down.  If there is a high number of caravans leaving town per hour, the bandit population max needs to go up, and perhaps even the bandit respawn rate.  If a bandit lord is killed, but the majority of his bandit camp is still alive, a new bandit lord should respawn more quickly than from a decimated bandit camp.  If all or most of a bandit camp is wiped out, the bandits should be more likely to adjust the location of their camp while they recruit new bandits.

Something to take care of the first few factors:

City of Heroes spawns at the edge of character vision - an avatar wandering through a previously empty city would find villains appearing outside of his cone of vision doing pre-determined villain things. Though it may never actually apply, there's a belief that parts of the gameworld which haven't had any avatars wandering through them in a while just cease to exist, and are re-created when avatars come near them. This may actually be useful if, to the edge-of-vision spawns, you wish to add static or time-random spawns that have nasty effects on the environment.

I'd suggest a sort of persistent traffic effect - an avatar passing by "taints" the gameworld. In remote stretches of forest there are mostly animals and less benign beasts, and occasional small camps of trappers or barbarians or whatever. As more avatars frequent the "remote stretch" it becomes more populated by "people" - bandits, loggers, whatever. So a mining/trapping town could, with enough avatar traffic, turn into a rudimentary hub of commerce and put another "trapper town" further out in the wilderness.

--GF

[Arnold]

Though it may never actually apply, there's a belief that parts of the gameworld which haven't had any avatars wandering through them in a while just cease to exist, and are re-created when avatars come near them. This may actually be useful if, to the edge-of-vision spawns, you wish to add static or time-random spawns that have nasty effects on the environment.

I'd suggest a sort of persistent traffic effect - an avatar passing by "taints" the gameworld. In remote stretches of forest there are mostly animals and less benign beasts, and occasional small camps of trappers or barbarians or whatever. As more avatars frequent the "remote stretch" it becomes more populated by "people" - bandits, loggers, whatever. So a mining/trapping town could, with enough avatar traffic, turn into a rudimentary hub of commerce and put another "trapper town" further out in the wilderness.

--GF

IIRC, that's how AC1 worked.  The server farm didn't have dedicated machines to the various sectors of the world.  When no one was in one section, no processing power would be devoted to it and that machine would be diverted to other parts of the server that needed help.