NPCs as workhorses, tools for balance, control and roleplay

[Dark_MadMax]

One of the main problems developers ran with their MMORPG designs is that they rely on delegating many of the in-game functions to player themselves ,without having no control whatsoever over the process.  Players generally have no problems in fulfilling leaders, hero’s roles  and brigands roles :) Typical functions which fail though are : law-enforcement , resource farming, crafting  and roleplay .

Developers have to make unhealthy compromises  to  make those areas “more fun” and more effective . But true is that to have real economy (not whack a mob-get loot parody)  you have to have real workers –miners/farmers/crafters . Being miner is not fun though –very far from being fun (eve-online is good example how really fun is to run mining operations yourself – just like a boring  job  ) .  How about this so exciting “trade bot “ job   -stadnding for hours and /shout’ing items you have for sale?  How about this : “find  X pieces of this resource by killing X^ 10 mobs” and bring it to “the crafter” who has  5% chance to make good stuff out of it?

Same thing is with law enforcement –  log on 7:00 pm ,patrol from old forest  ,to green swamp  till 22 :00 pm ,  encounter absolutely nothing ,log-off  - sounds like a security guy job   ,developers intended to be done by players themselves ( UO ,AC:DT ,SB) !

How about lore ? When was the last time elf player attacked  orc player because of some mythical role-play and not for ganking him and taking his loot? When last time elfs were considering that mass recruiting irekei in their guild is probably not wisest thing to do as they are supposed to be death enemies to each other?


You have to be careful in preventing all sort of player abuses of the systems (making sure there is no hyperinflation of goods and resources ,there  is balanced demand and supply) , no “dirty” roles ( as miner /farmer ) , no dedicated wolf/sheep by trade ( you are miner therefore you are supposed to suck!) .

The obvious problem is that  to have meaningful economics/politics/law somebody have to do the “dirty job” and be this sheep- resource gathering,  resource transportation, patrolling , standing in shops , building  -etc. etc .All the boring time sinks player don’t want to do because its indeed boring and very similar to unpaid second job. But those functions are necessary if we want build good solid political economical and law enforcing functions in the game. What to do if player don’t do it? Well solution lies on the surface : ) :

Make NPC do it –they don’t whine!   NPCs would be workhorses who provide infrastructure for players fun .  And NPC can be easily controlled and balanced by developers with predictable results and outcomes on players community .

I will separate NPCs in two groups : NPC grunts and NPC Guards

NPC  guards/ NPC grunts will be one of the main resources/balancing factors in  game .    
Lets examine their functions:

NPC guards – AI controlled entities dedicated to enforce in–game laws against other players. –E.g. they will directly compete with lone PC ,PC groups and sometimes even organized military PC group (raid) .  NPC should be powerful and “smart” enough to perform their functions of guarding and patrolling . (btw  Eve-online has effective guards )

We have two ways to improve power – improve AI to UT 2004 bots level (not gonna happen anytime soon :) ) , boost their stats (a-la EQ) or combination of both .Depending on how good AI  is you can use less cheating like boosting stats, super awareness , no-clip, etc..  At the very least AI NPC should at least use the abilities they have – having wizards/furies guards in Shadowbane whacking  tanks with their staves is pathetic :(. –More variety and  more creative NPC behavior  are way to improve PC vs NPC interaction and making it more fun! Some team work would be nice too ( scripts for healer AI NPC ,tank AI NPC, Nuker AI NPC are not something stellar  - look at BG ,NWN npc can provide some real challenge with minimal cheating (well not minimal but its at least not 10x HP, 10X dmg )  and sometimes demonstrate teamwork.

“Smart” – means they should not  be subjects to primitive exploiting methods such as kiting ,luring ,perching etc. – Nothing stellar needed here either .Either cheap methods  ( no collision detection for NPC a-la SB) , either a bit more advanced dynamic spawning ,either  decent pathfinding (heard city of heroes has that) . Dynamic spawning would be imho needed to discourage chaotics/enemies raids on highly secure areas  - some range around guard barrack/fortresses/watch posts will automatically spawn law enforcement units near KoSed/chaotic/Enemy PC. – Patrol routes would be imho too complicated to actually implement.


E.g. depending on the “security” level of the area ( influenced by the amount of player bound to lifestone , amount of guard structures) there will be probability of guards spawning near the invader (enemy or chaotic player) .  For each fixed amount of time of being in guarded territory  the probability increases. Balance should be made so the most secure areas practically guarantee quick death (quick escape ) for little hostile groups or individual griefers. – Such as nation capital serving as newbie area will have  increased security with practically zero chance for successful intrusion save for huge enemy army. While small frontier areas with  lonely guard shack may be very well inadequate for assuring safe travel of NPC trade caravan or cloudless journey of Joe The newbie.

Anyways the goal is to pose a challenge to PC – if the guards are intended by design to guard area from small raids they should be able to do it in 90% of cases. Of course intelligent balancing is necessary –we do not want to transform it to players vs guards game .


 There could be also other very important function for guards – reinforcing your own PC armies with NPC guards by controlling guards directly trough tactical combat interface (imagine a RTS like warcraft/savage style interface ,where you could assign targets to individual guards/ groups , set way points, patrol paths, give attack/heal/nuke orders etc. etc.)  . That will allow to small guilds to effectively defend their resources/cities even if their actual human resources are limited. –That would be also an edge a lawful guild would have over a chaotic bunch of mobsters . – Even one lawful citizen with good tactic skill could be very well able to  route roaming band of PC outlaws  with help of the garrison of NPC guards he has control over.

Another option for player controlled NPC guard would be NPC mercenaries – to help in wars and armed conflicts for a steep price.



NPC grunts -  economy driving engine and vendors. They will do all routine job like crafting/resource gathering/selling stuff/etc. Players will have fun part of giving orders to those NPC (where ,what ,how much, for how long ) , NPC will do routine part (e.g. player design  items he wants, stats he wants  ,puts the order to his NPC blacksmith , NPC blacksmith crafts it for XX hours depending on the quality of item – so  there wont be overflow of ultra uber items as there is time factor involved and there wont be boring click fest for 10000 times till you get needed item from random generator, - btw SB vendors were essentially timed random generators with  clicking ,as there was no way to make item  you wanted) . Of course good interface for controlling those NPC is essential ,so controlling  them wont become SBs carpal syndrome fest ,when you have to spend hours just clicking

 We can regulate how easily it is to get resources /items by regulating our NPC –we do not kill our players  with boredom of camping the  4 hours spawn which has 1% chance of drop this “rare” item… While at the same time  no amount  of catassing would skew the game balance :)  Rareness and uniqueness is NPC /resources dependent. We can easily regulate it by making resources harder to get, harder /longer to craft by NPC.


NPCs will also be one of the main factor in territory control balancing .In dark/light empire balancing you can make guards increasingly stronger closer the bind stone is to empire Capital . Similar could be done about resources : for example unclaimed mine at frontier would have initial dwarf guards hostile to both empire by default , and after one guild takes control over it there is no additional benefits to controlling side .While for example mine close to regional center in Dark   Empire will have undead guard /powerful lich boss as guards  non-hostile to dark empire guilds but hostile to any light sided guild . Even if light sided empire guild manages to seize control over this mine it could be forced to clear lich spawn periodically to just maintain control ,while dark empire seizing control over this mine will have additional benefits of undead/lich protection against rival light side guilds and  maybe even increased mining output (with help of undead miners) . –By balancing such factors can make very hard for one empire win over another(which would be the point  of balance :) ) . And even if one empire manages to do that regain lost territory would be easier.

And of course with NPC being the main infrastructure you can have “real lore” ,enforced by NPCs –  if we don’t want orcs buddy- buddy with elves its very easy to make racial elves NPCs hostile to orcs by design .- Overcoming this limitation ( by mules and such) would a deterrent enough to promote roleplay and make players actually aware of such things .

[Dren]

I agree that NPCs could fill in many "gaps" that are in MMOG's today.  SB has done some of that right, but didn't go far enough with it.  Your example of random generator of items is a case in point.  Why can't you just tell the blacksmith you are paying to make that "Hammer of the Gods" that he made just a second ago again?  That doesn't make sense.

If they want to make items like that rare, then the resources and time needed to first get a blacksmith to that level of skill and then to actually create it should be tweaked accordingly.  Making it random was just annoying to me and seemed like just another way to extend my subscription because my final days came down to logging in a couple times a day to recycle all the town vendors (people who ran towns will know what I mean.)

The various guards for towns was a great feature too, but just implemented poorly.  The AI was just not there, and may not be there yet (technology gap.)

To go further, I'd like to see some of what you talk about by actually using NPC's for attacks or even mercenary services.  In all the talk about MMOG's not allowing us to feel like the hero, wouldn't being in charge of a small army of NPCs give at least some of that to us?  While we still could be defeated, we were in command of more than just ourselves in the process.

I always thought a skill or attribute could be assigned to your character that would affect your "command" over multiple NPCs.  This would affect the quality and skill level of NPC you could attract and the quantity.  If you want to play an outlaw, you will be basically limited to small groups of bandits that can make quick yet damaging attacts upon a larger group be it a town or an army.

If you want to be a general, then most of your time will be spent growing an army and putting your skills to work on strategies, etc.  If you are the loner type, then you can do the normal MMOG thing and just worry about your own character advancement.  Your role in the bigger picture would be perhaps a captain within a player general's army or something.  Or you would just be considered a Champion upon the battlefield (i.e. Warcraft.)

Good thoughts Dark.

[Roac]

On the subject of player law-enforcement:

One thing you have to have, obviously, are laws to enforce.  This needs to be a game system; it needs to be something clearly defined, with a clear method of creation, modification, removal, and enforcement.  If I walk into a wooded area with 30 men, declare myself king, and make a law that anyone not a part of my guild is to be killed on sight for trespassing - that isn't much of a law as we understand it.  It's right by might, obviously, but I'm less a king and more a warlord, whatever I may call myself (or people call me).  Law requires some sort of legitimacy, permacy, and process.

Ok, so instead of the above, I get elected/take over/rise to the rank of a ruler.  I can now set laws for my city/region/country, through some method - either by personal diction, council electoral process, whatever.  Black box it - buttons get pushed, and now murder is illegal, and punnishable by a heavy fine.  In this system, what are law enforcement types to do?  More importantly, what parts of law enforcement are fun?

One thing that is certainly not fun is random patrols.  No one wants to wander the woods around the king's land at 3AM - at every 3AM - to see if anyone is poaching the king's animals.  It's a tedious chore, so scratch it - or at least, allow it but do not in any way require it or greatly reward it (ie, "patrol" career paths, or "patrol" missions unless for newbies, etc).  What can be fun are investigations - that is, figuring out whodunnit after the fact, or tracking down a known fugitive.  If you know there is a "crime in progress", that might be fun too (someone is killing the king's deer, but they may have left...), being a mix of both investigative and enforcement.

For the grunt parts, NPCs should most definately be looked to to fill gaps.  They could "raise the alarm", attempt enforcement (but likely be poor at it - but that's ok), etc.  Their limits should be at the point where the job becomes fun.  Even if an NPC could be programmed to do investigations - don't let them.  That sort of thing is fun to some people, so let them handle it.  Don't spend dev time locking up fun tasks.

[Daeven]

In terms of AI, NPC’s in MMOG’s should be modled toward emergent behavior based on object level attribute interaction (al la the sims) as opposed to extensively modeled stimulus / response paradigms as seen in bots. Remember, these are server artifacts. It doesn’t matter (to a certain degree) how much horsepower you need to devote to AI treads. There are both hardware and software based solutions to interconnect processing nodes to significantly increase your processing power in both parallel and asynchronous modes.

/techspeak

Short answer: sounds good.

Honestly though, one of my favorite parts of the fantasy genre was creating strongholds, building economic infrastructure, raising an army, and then dealing with the political brouhaha surrounding said events. IMO, it makes far more sense for regularly scheduled caravans of goods to actually move scarce good from point A to point B with caravan guards and all. That simple scenario provides multiple avenues of adventure in a dynamic manner right there.

IMO, the best ways to make MMOG’s dynamic instead of pre-scripted would be to implement scarcity, economies of scale, and make all base level mobiles model emergent behavior based on their surrounding objects. If you want to get really crazy, include meta level environment AI for a particular ‘race’ or a ‘societal grouping’ that modulates overall behavior based on race level interactions. Lots of humans killing Orcs? Well that should piss off the Orcs in general, and increase their aggressive nature specifically toward humans and their settlements…

[Mr_PeaCH]

On the topic of resource gathering/availability and NPCs:  This is an area that FFXI pulled off rather nicely I thought.  I like the setup with the 3 main starting cities and how players, by their activity in slaying monsters in the various zones of the world, vied for economic superiority of those zones.  Specifically, if the players from your city were 'in control' of a zone by virtue of having slayed more beasties than players from other cities, NPC vendors in your city could offer goods from that zone (which were, in the game sense, more or less unique to that zone and needed for various PC crafts, etc.)  So the players are involved, but only indirectly... no person has to necesarily do all the harvesting (although there were special roles that could be filled by PCs).  IMO, it was a good solution.  Players actions 'mattered' somewhat; your hometown was better off if you and your group and others like you went out to the hinterlands to level up.

Daeven, I have been formulating a premise for a MMORPG world, the centerpiece of which will be PvP centered around the routes of trade (caravans) between both NPC and PC cities.  I'll be eager to have you comment on it when I post it up.  As your previous post alludes to; this is the sort of idea which for me has its roots in being a DM for so long in PnP D&D in the past... I'm eager to see if/how such an idea might work in the MMO sense.

[Roac]

That's what I'd like to see too, Daeven.  I could see whole groups of players mobilized just around caravan transports.  Merchants who sort out what the trade routes will be (for both trade profit and to avoid bandits), to heros-for-hire who ride with the caravan incognito, to bandits who want to pillage them for profit.  Emergent behavior requires a level of complexity in the game system like that, but once you get it, you open up multiple paths for players to branch out and explore or find their niche.

[Fargull]

Hmm.

Like the topic.  I am fully for a greater NPC presense in the mundane of the MMORPG world.  I can only see a couple of caviats to this endeavor, one I think the common world as we know it would need to be expanded greatly to allow for PC territory and such.  Some basic fundamentals such as resourse control and macro economics would need to be taken seriously.  I think the resources are available, but it would be moving to a true multi-tiered server structure.  I am betting a lot of the performance with COH has been the back end lack of a large DB to keep track of the virtual economy.  Foot print size today makes the worlds seem way overly cramped.  Shadowbane has too small a footprint.  I am thinking like Harn Sized worlds with a few NPC controlled cities and a lot of open range.

Thoughts?

[Dren]

I too like the caravan feature.  Cities could base their income heavily on succesful caravan trips to other player or npc cities.  The caravan would be scheduled up by the leader/s and the guild/city/etc. would provide protection for an actual slow moving caravan from their city to the next point.  The farther away the delivery is, the higher the payback in terms of gold or resources.

Thieves or competing cities could take that opportunity to try and raid or destroy this caravan to either just make a quick buck or cripple their opposing army financially.  I see this as a great alternative to the current SB tact of totally destroying a city.  Instead, just bring it to its knees by choking off their supplies all in real time.

The leader/s could make the decision to make lots of smaller caravan trips (automatically) and just bet on only 1/2 making it but still turning a profit.  This option could allow players to choose to follow the caravans periodically and get some pvp action in at the same time.  Or, schedule really big caravans that are heavily guarded by players and, thus, would have a greater chance of making it to the destination.

That would make for some nice fun, plus be quite real since you would probably come across roving bands of npc caravans quite often.  You would then have to decide "Do I chance it and try to steal from this caravan or not?"  "What nation does this caravan belong to?  I don't want to anger the wrong people."  I can see a lot of play from this.

Good ideas people.

[Dren]

Oh, along the lines of investigation and such, when npc's of whatever type are murdered, there could be a featuer in the game much like UO's forensic evaluation (except this time it would actually be functional.)  This skill would point the way towards what nation did the deed and as the skill is increased could pinpoint the culprit/s better and better.

That way, murders of caravan npcs could be noted and retribution could be decided upon by the leader/s.  This would work for guards, merchants, mercenaries, etc.  This could be used on the carts for the caravans too so that thieves could be sniffed out.  This was another UO feature at one point that allowed a person to tell who looted the corpse, which was fun to do sometimes.  It certainly started some major fighting.

[Roac]

Toss in a few more bits too.  Assume that any caravan must be scheduled somewhat in advance - couple hours or so, minimum.  If they're recurring, that's ok too (so you can have one go out every half hour or whatever, to keep from putting all your eggs in one basket).  You'll want the schedule secret, but NPCs must know about it.  Now tie in the mission seed thread.

NPCs will be wanting supplies of their own - spawn off other caravans automatically that are not directly related to city/outpost health.  If they are constantly destroyed, the cost of transport goes up (limit in caravan supplies), but so do the bounties on the bandits.  Thieves, spies, sabateurs, or whatever class/skill that does something similar, might be able to sniff out the schedule if they look around (possibly having to do missions in the process to pay off the NPCs - hurry up and get them done before the caravan sets out!).  

If there is trade between two cities, I'd expect both cities to get pissed off at you if you ganked the caravan.  What if you could disguise the nationality of the caravan - would the disguised city get disgruntled (I mean, it wasn't their caravan, but you thought it was when you killed the guards)?  Hey, more mission potential in getting people to fake nationality.

[Daeven]

And don't forget the possibilities for stronghold owners in terms of protection a guilds. One of the problems that faced the medieval Holy Roman Empire was that every petty Ritterbrau, potentate and guy with a dam on the Elbe, Order or whatever levied a tax.

Goods will follow the path of least resistance. A Stronghold that provides protection in a certain area for minimal fees will garner more trade volume than a true robber-barons area of influence. Further, more goods can easily move on waterways, making rivers (and buildable canals) valuable strategic resources.

All leading to more reasons for player conflict.

[Alluvian]

Didn't get a chance to read it all yet, but I will.  I just want to point out quick that next publish SWG is putting in maintenance bots that pay your harvestor maintenance and can do harvestor collection runs for you.  They are putting in all sorts of cool bots actually.   Medical bots, Trap bots, (still need the skills to use them and still need to by filled with actual traps/medicine), survey bots (will go from where you realease them to the largest concentration of a particular item they can find).

Sounds like some of what you are driving at.  SWG has some good ideas, and is adding more, but it does not all gel into a fun game.  Too bad.  I do hope other devs keep an eye on it to add some of the good stuff and add it to an actual fun game.

[Mr_PeaCH]

Try this out with me; follow this:

World Map has a number of large, NPC cities strung about in a great circle.

All cities require various resources to flourish; NPC vendors cannot offer players nice things if not supplied and so forth.

All resources are not distributed equally; a city rich in some things, poor in others must use what it has to trade for what it lacks.

Players belonging to a city have a vested interest in maintaining control of their surrounding lands; some resources they may supply personally but simply wresting control of lands from monsters ensures a supply to the NPCs.

Cities automatically generate caravans between themselves travelling around the great circle and stopping at each town in turn.  The outer ring of this world is relatively tame so the NPC caravans are safe; but slow and expensive for the Players to use... which they might do in order to offer resources and finished goods (found, bought or crafted) to other markets.

Now enter the high level game for the players; as guilds, high-level players can finally wrest control of old interior caravan routes and set themselves up in a stronghold at the crossroads.

Player strongholds can generate resources of their own (and perhaps some extremely rare stuff) by exerting their control over their surroundings; these players/guilds no longer 'belong' to an NPC city; they are independant and tied to their stronghold.

Strongholds can also levy taxation on caravans coming through their zone of control; they will be wise to price themselves competitively against the NPC ones however and thus encourage more trade.  But an abundance of strongholds on the same route could drive the price too high and choke off the trade.  Competition between strongholds for routes will be high.

Coupled with this, players will want to prey on caravans that use a different route; others will want to protect their route.

***********

I know there's a lot of check and balance issues; but I think there's a start to be made with this.  Principally, around the notion of resources and trading and territorial control.

[Dark_MadMax]

On the subject of player law-enforcement:

One thing you have to have, obviously, are laws to enforce. This needs to be a game system; it needs to be something clearly defined, with a clear method of creation, modification, removal, and enforcement. If I walk into a wooded area with 30 men, declare myself king, and make a law that anyone not a part of my guild is to be killed on sight for trespassing - that isn't much of a law as we understand it. It's right by might, obviously, but I'm less a king and more a warlord, whatever I may call myself (or people call me). Law requires some sort of legitimacy, permacy, and process.

Honestly though, one of my favorite parts of the fantasy genre was creating strongholds, building economic infrastructure, raising an army, and then dealing with the political brouhaha surrounding said events. IMO, it makes far more sense for regularly scheduled caravans of goods to actually move scarce good from point A to point B with caravan guards and all. That simple scenario provides multiple avenues of adventure in a dynamic manner right there.

 I designed such political , economical and pvp system. Promoiting real poltical ,economical and war competition . Real war of thrones . Real accountability .   But  before posting it I'd like to have some feedback on ideas suppoting it . NPC would be one of  the supporting elements of my system .

 Also I would like to stress the importance of details - while for example SB nominally had NPCs supporting it (guards,vendors) actual implementation was far from being adequate. -Partially  imho by desing (designer didnt have clear desing goals ) ,partially due to poor implementaiton (bugs ,incompletness).

[ajax34i]

Game becomes more of an RTS, though.

It'd have to be a PvP world too.  There's no more PvE/PvP distinction, as I can send my NPC's to farm your fields, attack your NPC's, and so on.  Or I can PvE your NPC's, and thus hurt you.  

What you end up with is kings with NPC armies attacking other kings with NPC armies.  Fun, in a novel way, but many many polygons to display for what is essentially a 1v1 fight.


Also, the whole NPC-control system will become the point of the game (RTS again).  What you're suggesting is complicated; devs would have to spend a large portion of their development period implementing it right.  Yet you seem to suggest that the whole control-the-NPC interface is just to get boring tasks out of the way, so that players can focus on fun stuff.  Well, the devs won't have much time left to actually code the fun stuff, whatever it is, so the NPC-control system becomes the focus of the game.

In the end, the game is a mining/patrolling/farming spreadsheet that you "play" via the NPC's you control, with a lot of time required to set up and maintain.


Even the interface may force an RTS game.  At some point you'll need to control multiple NPC's fast: as a newbie you may interact only with your blacksmith, but as a clan leader you'll need to control your guards, shopkeepers, farmers, all at once.  RTS's already have good interfaces for controlling multiple NPC's, and it would be dumb not to take one such interface and apply it to the game, thus making it easy to control but giving it an RTS feel.

[Dren]

Indirectly, being more of an RTS is exactly what we are saying I believe.  You can mix direct PvP and RTS into one game.  It is being done by several games today (i.e. Savage.)  The interface and the development of this could be complicated, but we are mostly brainstorming the general ideas here, not actually developing the game.  

I'd expect there to be some issues with these concepts in development, because they are somewhat new to MMOGs.  Anything new is tougher.  I'm not discussing ideas to make it easier on developers.  I'm discussing ideas to make it better for gamers.

You talk as though RTS would be bad.  I consider it good as long as it is mixed in evenly with direct skill related PvP action as well.  Also, mixing the components of tradeskills, player driven markets, etc. (MMOG elements.)

[Dark_MadMax]

Game becomes more of an RTS, though.

More or less of RTS is the question of balancing. - IF you have more players than NPCs in combat in no way its RTS :) .RTS interface will just  faciliate commanders job and make real tactics possible (see my post about tactical MMORPG interface)

It'd have to be a PvP world too.  There's no more PvE/PvP distinction, as I can send my NPC's to farm your fields, attack your NPC's, and so on.  Or I can PvE your NPC's, and thus hurt you.  

 Of course it will be :)

What you end up with is kings with NPC armies attacking other kings with NPC armies.  Fun, in a novel way, but many many polygons to display for what is essentially a 1v1 fight.

 No. In my sytem we will end with real kings ,with their general and warlords,with their heroes and soldiers - all player characters . NPC being serfs and cannonball meat- ther roles PC do not like to take :) .

Also, the whole NPC-control system will become the point of the game (RTS again).  What you're suggesting is complicated; devs would have to spend a large portion of their development period implementing it right.  Yet you seem to suggest that the whole control-the-NPC interface is just to get boring tasks out of the way, so that players can focus on fun stuff.  Well, the devs won't have much time left to actually code the fun stuff, whatever it is, so the NPC-control system becomes the focus of the game.

 Well while it poses some additonal development difficulties. It wont be in any way overcomlicated. - Such interface is a trivial task ,NPCs AI and infrastracture is  more complicated thing to implement. But by doing that we also save hundreds of hours for developing "content" -as  the system as a  whole will be awesome content generator itself.

 "Code the fun stuff" -  I wonder what you mean under it , as coding itself game mechanics itself is usually a trivial task compared to the rest of the game . I have seen much more complex things coded in amateur games than this .

 The main coding burden for game devs are graphics egine ,net engine(server-client) , database....  NPC AI , scripting engine for quests and mobs, etc ,etc. UI is coded anyways and its far from being most complex  thing in a MMORPG ,especially if you design UI right way ( having in mind that you will need it to be versatile) .

In the end, the game is a mining/patrolling/farming spreadsheet that you "play" via the NPC's you control, with a lot of time required to set up and maintain.

 You didnt see my game design yet .How you making such conclusions? Players control the NPC.  But that doesnt mean that everything in game will be able to be done by NPC. NPC are essentiallly another type of resources and assets .They wont " win   game" for you. -As being successfull in MMOPRG PVP game first and foremost will mean to control other players.

Even the interface may force an RTS game.  At some point you'll need to control multiple NPC's fast: as a newbie you may interact only with your blacksmith, but as a clan leader you'll need to control your guards, shopkeepers, farmers, all at once.  RTS's already have good interfaces for controlling multiple NPC's, and it would be dumb not to take one such interface and apply it to the game, thus making it easy to control but giving it an RTS feel.

RTS feel is when you build buildings , mass  produce units and waste those units . -All  in a timeframe of 30 minutes.

     When there are real players under command ,units take time to produce, you have  territory  ,player owned assets  ,player run economy its becomes simulation of  feudal warring world , not a mere RTS.

Its like saying world war 2 was an RTS.

[Roac]

What you end up with is kings with NPC armies attacking other kings with NPC armies. Fun, in a novel way, but many many polygons to display for what is essentially a 1v1 fight.

Not at all.  There are far fewer kings than peasants (soldiers, mercenaries, merchants, bandits, sheriffs, assassins...).  Some of these others ought to have control over, or at least INFLUENCE over, NPCs as well.  If a merchant guild (group, whatever) moves into an area and starts operating, I'd expect to see NPCs naturally spring up to start trade routes.  Roac the Merchant doesn't control them directly, but the merchanty behavior of he and his dozen pals causes them to start doing their thing.  Roacc the Bandit might move in with his pals, and start clearing out some of the NPC traders for profit - which might affect profitability for Roac the Merchant.  The King, seeing a negative impact on tax (he may own a bridge, tax the income of the merchants, whatever), decides to have some of his [NPC] soldiers move into the area and patrol more regularly.

And all of THAT assumes that there are players around who actually want to buy what the merchants are selling, to mine resources (whether UO style, SWG style, or something else) for the merchants, to craft stuff for the merchants (again, regardless of what system is in place), and so forth.  

One thing I generally don't like is how much power and authority "the guy at the top" has in games.  The GL can toss everyone in the guild out, instantly, without notice (or just dissolve the guild).  The King can sell off or burn down buildings.  The Emperor can empty the nation coffers, pocket them in his private bank, and retire to a small tropical island somewhere.  It is utterly unbalancing to grant this amount of power to a player, as well as unrealistic (yes, kings DID sometimes empty nation coffers, but the threat of death to angry mobs or a ticked patriotic chef tended to limit that somewhat).  Transfer of power, and limits of power, are neccessary.

Beyond that, what we're talking about wanting isn't a RTS game - it's a MMOG/RTS hybrid of sorts.  Soon as you toss in other players, you need something for all these other players to do.  Being the top dog shouldn't be the only fun avenue, or grant all the neat toys.  Soldiers, mercenaries, bandits, assassins, sheriffs, guards, treasurers, and whatever else also need a role.  How about civic officers?  Treasurer, General, Guard Captain, etc.  Give these roles certain amounts of exclusive power too.

What if a king's only authority was to slot people into these roles.  Any account can only occupy one role at a time (if you want to buy a half dozen accounts to fill a city, go for it), although the King could also occupy one other role (Kings shouldn't get bored).  Getting rid of an appointment is difficult - maybe requiring a majority vote of the city officers (King counts as two votes?).  

Ok, so now you have the Treasurer who sets taxes, slots money for other departments, and sorts out disputes, the Guard Captain who organizes city (and surrounding area) defenses, the General who organizes troop movements, a Justiciar who sets laws (yes, sets - law by prescedent! :P ), such as how high a fine there is for murder (up to whatever cap the game has), whether elves are viewed as "tresspassing" if seen (we Humans don't like elves...), and whatever else.

Something like that, anyway, to mechanically break up responsibility (because players won't do it themselves, or find a need to).  Each of these top dogs should have a big influence on NPC movements - guards route trade, generals send NPCs off to war, or to battle the orc infestation, treasurers affect the flow of money which in turn affect NPC thieves/bandits/etc activity location and frequency.  The NPCs, in turn, have motivations, which kick off missions for players, who can in turn affect (on a more minor, but more direct way) how the NPCs operate.

[Daeven]

What you end up with is kings with NPC armies attacking other kings with NPC armies.  Fun, in a novel way, but many many polygons to display for what is essentially a 1v1 fight.

Because, you know, Mideival: Total War is such a boring, one - dimensional game and whatnot.

...

[CassandraR]

I been working on this idea in my head for a great long while. I set it up so there would be two main types of players. Lords and Heros. Lords would work more on a marcomanagement scale, having alot of RTS like options open to them. Build cities, strongholds, forts. Mass-produce weapons, armours, trade goods, and setup wagon routes and do politics with other cities. Heros would be more like traditional MMOG characters. They could do quests, and slay the big beasties and do heroic style crafting of objects of dreadful power. Heros could work under Lords as just about any role that was named above if they liked or they could go solo. A single lord could run a small sized city but it would take several lieunitants under him to run bigger ones successfully.

For a metagame or end game I planned on letting the characters eventually ascend to be demi-gods and wrestle over control over the world and its people's minds and hearts. Lords and Heros could pledge to one of the demi-gods and they could receive aid depending on what favor the demi-god was. They couldn't directly effect the world except for rare times when the veil between the worlds was thin enough or they collected enough power, but they could found religions and people that wanted to be become paladins and cleric type characters would have to pledge to one of them. Demi-gods could also war with others on their own plane of existence over ancient artifacts and sacred sites and go on epic quests and adventures.

There would be various races open to people, from the normal fantasy ones to less used ones. Each with some special rules and cultural differents that made building and living in their societies different. 'Lesser' npc races would also wander around the map, making camps and eventually small outposts and towns if left alone and would not respawn if killed. They could sack near by towns and raid wagon routes and generally make a naunce of themselves. Big boss type monsters could spawn randomly out in the world. Like a necromancer raiding your graveyards at night until he has enough bodies to raid a town with his skeleton hordes or build his dark tower to continue his studies.

[Bluey]

I have thought about the issue of either spawning commodities or mining primary resources and considered the value of NPC involvement.

I had thought about some proportional relationship between player activity and NPC activity. The concept being to force the players in doing certain tasks to enable primary resource production, the hard slog as it were but magnifying it by NPC effort. The players activities would control the amount of resource available but possibly reduce to some ectent the uninteresting tasks of mining etc with more rewarding tasks. The problem of player work to drive an economy can be got around with large efficiencies but it might be more rewarding and less open to artificial manipulation if the NPC's carried some of that burden in a less efficient economy.

My background in this is from the 3 space based MMORPG's EVE, Earth and Beyond and Jumpgate. All of which had some aspects to their economies I liked but much I did not. EVE places too much emphasis on the hard slog of mining for instance and all could do with some NPC interaction to manage the economy as the thread starter discussed.

[Nopeadope]

Mu's thoughts on food may prove useful.  His treatise/rant is good readings anyway.

[Mi_Tes]

Mythica was to have many uses for NPC's.  Besides worshiping you, they would help you with making weapons, building things, and helping you on quests and other combat.  This was one of the many reasons I was looking forward to Mythica.  I think there is much room to be innovative with the use of NPC's to make the game more immersive, free your time up to do the fun stuff, and give more flexability to when and where you quest.  I definately would love to give the timesink of crafting to NPC's or get help with combat/quests if I couldn't find friends on or didn't have time to wait on others.

[Tairnyn]

Very nice thread. I think 'intelligent' NPCs could be the key to making these online worlds really seem alive. Right now many games tend to limit the landmass to avoid having huge tracts of abandoned real estate. As a result, you really don't have room for countless NPCs wandering about carrying on with tasks, as you would in, say.. the real world.

I feel that a game starting out as a community of NPCs and slowly folding in the players could really promote the idea that you've joined an existing world that has a life of its own, not merely created from scratch, balanced on the roleplay that you yourself have created around it. A small group of anti-roleplayers can't tear down your house of RP cards because the mechanisms would be in place to support your power in the region.

Consider a system in which a city's tasks are scheduled by (PC and NPC) 'captains', and players are given the opportunity to assume responsibility for certain tasks based on rank in that city. A new player to the area may get hired for simple mercenary work, working their way up through the ranks to become head of the city guard.

Once they attain an administrative position they are given (virtual) city resources which they can distribute to certain tasks. They may assign a 100g bounty to locating a thief that has been terrorizing the townsfolk or 200g and extra rank for escorting a particularly important caravan. (Perhaps the city's trade advisor offered the guard advisor resources to protect it for him, not trusting his own sub-standard NPC guards) The head of a city would be responsible for distributing resources to different departments based on their own perception of priority.

In some ways, NPC actors *should* be imperfect. There should be a motivation for a player to take that responsibility away from the NPCs and reap the rewards of success. The price for a service rendered may increase because no one seems interested in spending that much time for the meager reward, or decrease in a bidding war.

I envision a stranger entering a town and approaching any one of the NPC (or PC) representatives for the different departments. The trade department may have dynamic crafting and caravan tasks available, while the city NPC constable can offer you a list of short-term protection and enforcement duties. Any of these activities will occur on schedule, done by the NPCs, but playes have the option of assuming responsibility for that task and doing it themselves for the reward.. or penalty for failure.

In this way, the dynamics that occur between different cities and regions can continue to flow regardless of player involvement, but such involvement enriches the town and offers many opportunities for meaningful roleplay and reward. The player has the option of claiming a city as his home and spending all his time working on his reputation there to someday become the leader, or being a wanderer doing simple mercenary work for anyone with enough cash.

The net result is a network of cities that, when controlled by players, could evolve from humble hamlets to hotbeds of activity (or better yet, thriving cities, a result of resources put towards building) through player effort and skill.

Granted, there are some holes in this idea, such as people making alts to assume responsibilities for an enemy city and going out of their way to fail. It was just an idea I thought I'd share, and this seemed to be the place to do it. My apologies if it seems a bit disjointed.