Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Quick Links
Random Quest Path Generators
Lucioon
Member UncommonPosts: 819
Continuing from my last Post about Quests in MMO, most of the views of the forum is that Quests are needed, but at their current form, its tedious and boring.
So What do people think of Random Quest Generators?
Instead of chains of quests that needed to be completed in order, each player chooses 1 quest that they wanted to do, depending on choice during the quest, they get randomly generated follow up quests.
Lets use examples:
1) 4 player group, finishes Quest A
2) each player gets an different B quest option
3) They decided to continue the path down Player 1's Quest
4) The quest continues and one of the players decided to go to bed.
5) They found another player around the neighborhood, and the other player can join right in and get participation credit.
6) Each person gets their own rewards. And can veer off course anytime they wanted, but once an decision has been made, it can't be unmade.
And when completed, different quests and different events will unfold for your character, that you can pick up more players to help you accomplish them without penalizing them.
so basically, moving decisions from Raids and Instances into the world and have those choice matter. If its an enviromental and world impact, then have all roads lead to the same end, but like many said, its the Journey that matters. Therefore, having many journey paths creates uniquness and makes your character more alive.
Does anyone agree?
Life is a Maze, so make sure you bring your GPS incase you get lost in it.
Comments
I don't think quests actually matter, unless it is connected to a deep story. Otherwise, it is just an excuse to kill lots of mobs. So it boils down to if the combat is fun.
The kill 10 rat quests are boring not because it is "killing 10 rats", it is because the combat is not challenging, or does not feel powerful (hacking rat one by one is pretty mundane).
People play the same quests in Diablo 1,2 and 3 month after month becuase it is fun to wipe out horde of mobs, and combat can be very challenging with very different combat style for different classes.
So the fix is easy. Just make combat fun. The excuse to kill does not really matter that much.
Long ago I wrote a dungeon/maze generator that was based from a random plot generator. While it "worked", it doesn't look anything like a quality hand crafted quest. Too artificial.
Epic Music: https://www.youtube.com/watch?v=vAigCvelkhQ&list=PLo9FRw1AkDuQLEz7Gvvaz3ideB2NpFtT1
https://archive.org/details/softwarelibrary_msdos?&sort=-downloads&page=1
Kyleran: "Now there's the real trick, learning to accept and enjoy a game for what it offers rather than pass on what might be a great playing experience because it lacks a few features you prefer."
John Henry Newman: "A man would do nothing if he waited until he could do it so well that no one could find fault."
FreddyNoNose: "A good game needs no defense; a bad game has no defense." "Easily digested content is just as easily forgotten."
LacedOpium: "So the question that begs to be asked is, if you are not interested in the game mechanics that define the MMORPG genre, then why are you playing an MMORPG?"
Great. So how do you implement it? Infinite amounts of content that is actually good is the holy grail of MMORPG design. Randomly generated content that is actually good and has sufficient amounts of randomness to seem like a wide variety would solve it. But if doing X would solve a holy grail problem, then doing X is probably hard.
That's not to say that it shouldn't be tried. Sometimes holy grail problems do get solved. But it's a long, long way from "It would be nice if someone made random quests that were good" to "I can do it and here are the details of exactly how". The latter is what we need, not the former.
It's isn't a silver bullet.
Epic Music: https://www.youtube.com/watch?v=vAigCvelkhQ&list=PLo9FRw1AkDuQLEz7Gvvaz3ideB2NpFtT1
https://archive.org/details/softwarelibrary_msdos?&sort=-downloads&page=1
Kyleran: "Now there's the real trick, learning to accept and enjoy a game for what it offers rather than pass on what might be a great playing experience because it lacks a few features you prefer."
John Henry Newman: "A man would do nothing if he waited until he could do it so well that no one could find fault."
FreddyNoNose: "A good game needs no defense; a bad game has no defense." "Easily digested content is just as easily forgotten."
LacedOpium: "So the question that begs to be asked is, if you are not interested in the game mechanics that define the MMORPG genre, then why are you playing an MMORPG?"
You got to start somewhere right
Although I don't have the programming skills, while those with skills are learning how to program, I was busy playing games. Which kinda sucks in the long run, but I think there are hundreds if not thousands of Long time Gamers that could have the one idea that sparked that one idea in an Programmer that helps them implement that idea.
I know Many are done with quests, but like what many stated in my previous post, its an necessary Evil. Otherwise there won't be enough activities to do.
I am also an avid Anime watcher, I remember when I was young, many Anime has Weekly tournaments, Monthly Tournaments that is open to everyone. The winner wins the entrance fees put up by everyone, there was also underground gamblings as well as official bids.
Then you got tournaments in every town, that you get an invitation to the biggest and toughest tournament in the main city. The journey to those places, the underhand of dark politics.
Why aren't they implemented in MMORPGs, they feel so alive and MMORPG like.
Life is a Maze, so make sure you bring your GPS incase you get lost in it.
If you want to have randomly generated quests, then the key question is how to randomly generate those quests. If you have a bunch of custom quests and pick one at random (which I don't think is what you meant), then it's going to get repetitive much more quickly than if you let players do all of the quests by traditional methods, as you won't do that many quests before you start seeing repeats.
And if the quests themselves are created randomly, so that millions or billions of quests are possible, then the key question is how to generate those quests without them being dreadfully boring, highly repetitive, severely buggy, or (most likely) all of the above. That's not a simple matter of programming. If you've got the aptitude, it only takes a few weeks to pick up a computer language. Rather, it's going to take some serious probability skills--which most game programmers don't have--to have any meaningful chance of pulling it off.
If you've got a way to randomly generate a large number of widely varied quests that are actually good, then doing anything you could conceivably want to do with them is trivial by comparison. (By the way, such quests will not have quest text, unless you're satisfied with text about on par with what spambots can produce.) And what would a game programmer in that situation do? What he wanted to do--and not necessarily what you want him to do.
The issue is that everyone and his neighbor's dog has ideas. A lot of people even have good ideas. Game programmers have ideas, too. Ideas that they'd like to implement, no less--but nowhere near enough time to implement them all. And they tend to like their own ideas better than your ideas, which is why they're their ideas.
Programming a game is mostly a matter of taking vague ideas in your head and making them absolutely precise, with perhaps a relatively minor side of optimizing your algorithms to run fast on slow hardware. Learning a programming language is just a matter of learning to write down your ideas in a form that computers can understand. But if you can write down your ideas at all with the necessary precision, putting it in a form that computers can understand is usually pretty easy. The problem is filling in the details of vague ideas to make them precise.
Back to general quest generators. In order to pull it off you need to give NPCs a set of parameters, say 20 or 30. These include personal information: the region they are from, town or city they are from, thier status, sex, race, etc. Also included are modes of interaction: small talk, story, quest give, etc. These seem pretty standard but if you add them to a unified quest database you can get random results based on how many parameters were set up to begin with.
Example:
You just finished killing rats (couldn't help myself) for Farmer Joe McBob. One of 20 random quests he has for players that talk to him based on the time of year, time of day and what faction you have within his territory. He thanks you and tells you:
"My cousin Tery McBob is a senator in Quezart. He may find your skills useful. Thanks again."
This was random based on:
Relation to NPC. It chose "family" from neighbor, townsfolk, family, etc. since it chose "family" every NPC with the last name "McBob" was considered and once it randomly chose Tery McBob his information, from the database, was used in the text. It just has to be set up that way.
What keeps a system like this interesting is that not only does it send you all over the game world, potentially, it also has chances to give you deeper story quests. Once you get to Tery McBob he may give you the simple task of delivering a letter to a secret love, depending on the random roll, but after trust you enough to give you a story quest where you are his right hand man exacting justice on a rogue villiage.
Are you relying on having 20 quests custom written for Joe McBob, of which a player will only actually see one? Because if so, that's going to be enormously expensive--and if you have the budget for it, why not just let players do all 20?
If you're relying on Mad Libs style quest text where most of the text is filled in and a few words are substituted here and there, then the quest text is going to be really repetitive. Anarchy Online ran into that problem. The second time you see the same template with a few words filled in, it's going to feel repetitive. The twentieth time, it's going to feel flagrantly broken. Furthermore, unlike some things in game programming, writing custom quest text manually is something that everyone and his neighbor's dog can do.
Quest text is extremely hard to randomize without it being stupid. Unlike a lot of other things one might try to randomize, there's a huge, obvious commercial use for random text: spambots. If you can find a way to get bots to write "random" text touting your product that looks decent, you can circumvent any spam filter and may even fool forum mods. But even with huge commercial incentives to create it, spambots have awful grammar.
This is a pipe dream.
It is enormously expensive, hard and you'll never achieve the same quality or feel as you would with a custom quest. Furthermore, you cannot convey a coherent and overarching storyline with generated quests, or it would be even further harder and more expensive. The template will start to feel repetitive at some point. At worst, players may exploit it for example cherry picking. Making changes to the different variables may be easy, but changes to the structure and template could be problematic.
I skate to where the puck is going to be, not where it has been -Wayne Gretzky
I don't feel that randomization alone will work well - it's not just the unknown that I crave, its the hints of structure, that sense that there is in fact something to be learned by exploring these tasks. If you have a purely random quest chain, I suspect I will quickly notice that it is random and arbitrary and the illusion that my actions matter any more than they did with static quests will be dispelled. So my preference is to start with the world, unfolding like some slow-motion strategy game, with quests spawning as hints as to what actions are currently available which will influence the state of the game.
What I was proposing above was not a replacement of the entire quest system but an addition to it for avid questers. IMO you don't need thousands of single, prewritten quests in a game but rather hundreds of packaged ones that feel like actual quests. Some games have thousands of quests, each with start, finish and sometimes arch text. Say you would have had 3,000 quests so you total pages of text would be 6,000. Imagine if you cut the total in half, 1,500 quests, 3000 pages and spent the other half writing variations to put in a database.
What quests would they provide? It isn't preset at all. The algorythm could start with kill, gather, talk to. Once that rolls, say "kill" then it selects town or province, "province", then pulls at random a monster in the whole province. If it had selected town then it would have picked a monster listed in the database that had a tag for that town. The system pulls up "Bay Wolf" from the data base, near the northern coast of the continent, and decides if you see someone else in the nearest town or get a reward when you return to the town where the quest giver is.
Finally, and this is where Quiz makes a good point, the actual text. If you could manage a paragraph I think that would suffice for variation in a "standard" quest. Again, this system is not intended for every quest as the marque long quests would have there own dialog.
The start quest text would be dependent on what rolls happened. The man asking you to kill Bay Wolves would have have say two responses to every seperate roll. The final combnation would be random but not too generic. If you kept coming back to the same person you would get similar responses but since the system could send you near or far you probably will not meet the same people on a regular basis unless they are part of story arcs.
For example:
1st roll, Kill
2nd roll, Province
3rd roll, Bay Wolf
* checks for closest town and rolls for NPC
The NPC would have a set of starter sentences:
"Could you help me..."
"Can I ask a favor..."
"You know what?..."
"Would you help and old out?..."
From there it would be a matter of inserting information that was rolled on and adding variant bridges to get one paragraph. The idea is that if you are sent off to random places and not on the same quest the guy next to you did then the "story" wouldn't have to be spelled out from A to Z, you would be living the story. Would it be hard to implement? Absolutlely but if done from the ground up it would seem to provide what the OP is talking about. Is my idea water right? No but if it was I would be writing it myself to copywrite it and not posting it on a MMORPG forum
Although I don't have the programming skills, while those with skills are learning how to program, I was busy playing games. Which kinda sucks in the long run, but I think there are hundreds if not thousands of Long time Gamers that could have the one idea that sparked that one idea in an Programmer that helps them implement that idea.
I know Many are done with quests, but like what many stated in my previous post, its an necessary Evil. Otherwise there won't be enough activities to do.
I am also an avid Anime watcher, I remember when I was young, many Anime has Weekly tournaments, Monthly Tournaments that is open to everyone. The winner wins the entrance fees put up by everyone, there was also underground gamblings as well as official bids.
Then you got tournaments in every town, that you get an invitation to the biggest and toughest tournament in the main city. The journey to those places, the underhand of dark politics.
Why aren't they implemented in MMORPGs, they feel so alive and MMORPG like.
Programming a system like this isn't the hard part. Designing this system so that it works on paper is the hard part. Well, I don't know if it's the hard part, but it's the missing part. It's almost better if you're not a programmer because you're not going to bother stopping in the middle of what you're doing to think about the database backend or the objects you need to write to handle what you're doing.
So, step one is make it work on paper. Write it out and walk through several scenarios pretending to be a player, but instead of a computer providing responses, provide the responses yourself, on paper. Think of it like a pen-n-paper RPG session.
If you can make it work on paper, it can probably be written. If you can make it work on paper, then you can probably diagram the process so that someone who is a programmer can write it.
I can not remember winning or losing a single debate on the internet.
First, I think the word you're looking for is "marquee", not "marque", unless you mean that the game needs to be full-loot pvp where you can steal stuff from merchants of other nations. Which you might, as there are some fans of that around here.
If you're only going to do this for the really formulaic quests, and willing to accept really generic sounding quest text, then it could work. You could do this with kill ten rats type of quests, for example, where the "ten" and the "rats" are rolled at random, with the latter coming from mobs in the area. But you wouldn't be able to write an NPC's impassioned plea of how desperately he needs for you to kill those ten rats, without it sounding really stupid if it rolls something other than rats, or having to basically write custom quest text for every possible mob. But if you're satisfied with quest text along the lines of "Please kill ten rats and then return to me. You can find them northeast of here. I'll give you 300 experience and 42 gold.", then it's not hard to write custom quest text for every quest without needing a formula.
It wouldn't work so well with a lot of "go talk to so and so" quests, as the purpose there is to send you to the next quest hub. Try to randomize that and you send the player to the wrong quest hub.
If you can write something out and fill in all of the details on paper, then yeah, you can probably code it. Coding it might be impractical for performance reasons if on paper, you need several nested loops (as in, check every combination of a, b, c, and d). But creating the code from the formulas is typically pretty easy.
The problem is filling in all of the details. Writing source code forces you to fill in all of the details and explain exactly what it needs to do under every possible circumstance, in order to get the program to compile and run without crashing. If you're only writing it down on paper, then it's easy to gloss over vast swaths of complexity and not realize that you're skipping nearly all of the work. What you need is for someone who is hostile to you and trying to break your system to be unable to come up with circumstances in which he can find results that are obviously broken, but still completely consistent with what you've written down.
IMHO, most developers are looking at this incorrectly. It isn't about the quests needing to be random, but that the quests need to be somewhat player defined. I think blizzard was on the right track with monster power play in D3, but the Path of Exile developers got it much more perfect with their customizable dungeon scrolls.
True. You have to remove the human decision making and move that into the process itself. Which is the whole point, right? Taking something that works in your head, and then making the computers do it. Glossing over things is going to happen with or without programming experience, but you have to start with something.
That starting point is going to be writing it down, and walking through the process yourself, pretending to be a player. That really is just step one. Maybe Step 0.1. Then you have to talk your friends into walking through it and then start bringing in people who have some idea of what the software would need to do in order to accomplish what you want to do.
Then you go back and redo the whole thing, probably more than once. :-)
** edit **
At the very least, you'll have an idea if what you want to do is worth the effort.
I can not remember winning or losing a single debate on the internet.
Good random quest design would be even more difficult than good random level design. Its an area known as procedurally generated content. Its simple in concept, but insanely difficult to actually design and code. Good content is one of the most expensive parts of MMO's. If someone eventually discovers how to procedurally generate good MMO content, I suspect they would be RICH in relatively short order.
Look at this as just an example of it as applied to level design.
http://indiegames.com/2012/11/5_tips_for_using_procedurally-.html
This is an overview of the approach used in many different fields.
http://en.wikipedia.org/wiki/Procedural_generation
I didn't clarify this earlier but my idea of generation involved every, or close to, NPC in the game. This would mean that no matter who you were sent to or talked to you would get some sort of kill/collect/talk at random.
With the unknown comes a sense of mystery and even though the quests themselves may be generic with short to the point dialog the fact that the whole experience is random would be more useful and intriguing to me than the same prewritten quests the guy next to me is doing.
But even monster power in D3 is limited to 10 levels, and i bet each is calibrated carefully. And taht is only 10, a very small number of options.
I highly doubt if a system can generate thousands, or even millions of combinations of quests can guarantee each combination make sense. Either it will feel generic, and occasionally have something that is weird, or you have to hand craft the quests.
And my question is whether this is worth the effort at all. Look at D3 .. the same set of quests are played over and over again. Players probably only cares about the actual story the first time through .. after that, it is just an excuse to kill stuff. And Blizz cleverly focused on making killing stuff fun.
I played WoW up until WotLK, played RoM for 2 years and now Rift.
I am F2P player. I support games when I feel they deserve my money and I want the items enough.
I don't troll, and I don't take kindly to trolls.
Occasionally having something that is weird isn't so bad. If 1 random quest in a thousand is completely broken, you're still doing better on that count than most games where every quest is hand-crafted. If one in a hundred is completely broken, that's completely acceptable so long as you can drop it and reroll and most of the others are pretty good. It's better to have 80 good quests, 19 boring, and 1 completely broken than to have 100 boring quests that all work right. The problem is if it's 99 boring and one broken, or if half of the quests are completely broken rather than only 1%.
That is the problem.
You don't know if the random quests are going to be boring or good. What is a good quest anyway? Let me take a extreme example. Let say you have 100 mobs, and 100 location of containers.
I can have a quest like .. kill xxx mob to get key, go to location yyy to open the container (box, chest, cabinet .. filli n differnet good sounding names) to get quest item.
There are 100x100 .. 10000 combinations .. but it feel generic and boring. So a large number of combinations don't guarantee interesting stuff.
Now you would want to make the interaction between parts more complex ... use a more sophisticated procedure .. may be some if-then branching .. but at the end of the day, they are just combinations.
If i am going to get "kill x random mob" to obtain "y random items" to "satisy random NPC" .. i may as well have them handcrafted. At least the handcraft version will sound more coherent, and some quest text that may make sense.
Food for thought: would a game of chess make a good questline generator?
Imagine you take a zone map and divide it up into 64 areas (it doesn't have to actually be an 8x8 grid, but logically we'll be treating it that way). You have 16 NPCs on each of two sides that will correspond to the various chess pieces.
A game of chess now plays out but each move is actually the next quest being offered. A move of a friendly piece into an empty square is represented by a quest vs some local wildlife (with the NPC then moving to that location). A move of a friendly piece into an enemy square represents a quest to kill members of the opposing faction.
During the opposing side's move, you have similar quests, but they are defensive (escorting wounded NPCs out of harm's way or recovering goods from enemy camps.
One version of the zone would have the server playing chess against itself (or the imported list of moves from an already-played game). Another version might allow you a choice of "moves" (go help different NPCs) so that while you are questing, you are simultanously playing a game of chess, attempting to corner the enemy boss before the enemy corners your boss.
( the idea I'm trying to probe is whether even repetitive quests can become more meaningful if players know they are actually part of a wider, slower-moving supergame rather than just an arbitrary story )