Quote:

Original post by onyxflame Maybe I don't WANT to know exactly what I'm supposed to do from the beginning. What if someone tells me to find the foo, but doesn't tell me how, because he's really hoping I'll die in the attempt? What if he tells me to find the foo, but then someone else tells me the foo is vital to the survival of my civilization? What if the foo is really a dirty sock instead of a foo, and he was just testing my loyalty by telling me to get it? Maybe the foo isn't the point of the quest after all, or maybe the foo is really figuring out who to trust and who not to.

Thanks, these are intriguing variations on missions. It gives me the idea that missions goals might be fuzzy, contradictory or contested by enemies of your employer (or you) while you're doing them, which could make for interesting mid-mission hijinks from time to time. What's cool about this idea is that a reputation system that makes you weigh one thing against another would add so much to gameplay: Fail the mission and get an in with an exclusive faction, for instance, or keep raising your rep with a faction that may be on its way out-- could be an interesting choice.

You might find these pages helpful if you find yourself short of ideas for plot templates and/or variations for each: The 36 Plots and The Big List of RPG Plots.

I've played Imperium Galactica II some. It's a space strategy, conquer the galaxy deal. You can play as one of three major races, beginning with one planet. The AI is pretty shabby, the diplomacy slightly below average and it has many faults, but it really shines in one area, and that is telling a story.

It works out like this, that although the big goal is pretty much to whipe every other race from the galaxy, it all begins with a small planet where events starts to happen. For example, playing as the solarians, there is one mandatory eventchain, that of a cult searching for four lost crystals. Throughout the beginning of the game they cause a lot of trouble with riots and what not until they in midgame has grown important and calm enough to reason with, and have you search for the crystals too. There are also two large eventchains, of which one is always choosen. The first one is the pirate campaign, in which pirates attack you, and after you capture their first planet early on, you gain a lot of taxbase aswell as an extra shipresearch centre. The other is the Cha'kra campaign, in which an old mysterious race attacks you early on and whipes out half yourfleet, but you gain the blueprints to a cheap advanced ship after beating them. Which one of these storylines you get influence the way the game is played tremendously, and has followups until the late midgame where both threats are dealt with. There are also numerous short episodes, terrorists threaten bioterrorism, vulcanoes errupts, traders sell faulty radiationsuits etc, which prompts you to take certain actions.

What I sit here, enjoying a game like I havn't since master of magic, ufo xcom, and the rest of the classics, is ofcourse an expansion of the funfactor. More semi-random and intertwined missions building up an epic saga. One of the hero units is even named Sheldon, perhaps hinting towards their inspiration.

In an amalgam of previous ideas, I would suggest that based on the major players of the world, large storylines would be randomized. These major storylines' buildingblocks would be events and missions which are randomized as well, with parties derived from the larger one. I don't know if something like this would make the game cease to be openended, since completing it would mean a 'win', but I'd pay you money to play it :).

In the case of the RPG where the personality of the player can't be assumed (high lord of the galaxy wannabe), to chose the largest of the storylines would be preferable, so a dwarven traden don't get a long save-the-forest tiriade.. Also, developing that idea, perhaps to have semi-long storylines from the environment could be made secondary to a gamespanning personal storyline for the character?

I'm just thinking off the top of my head here though. Since this is a school project, I would suggest you check out the AI technique of using Agents. That's like tracking the wishes of local kings, popes, bishops and thievesguilds leaders and have them want things, and ergo create quests, which can be fullfilled by either the player or perhaps computercontrolled adventuregroups. I think however those a bit easier can get the autogenerated feel to them. More realistic, but 8 bountyhunt missions in a row might not make a good story.

/Rosecriox

I think the point you need to take care to avoid is not that quests might get repeated, but that seemingly unique quests will repeat.
It is one thing to select between 10 bounty hunting quests which will (apart from names, locations and enemies involved) be more or less the same every time, and another to destroy the death star three times in a row.

(why? because one is an immersion breaker and the other isn´t)

For memorable quests I´d rather put in the extra effort and script a bit more, or create a few more semi-unique locations for your adventures. Most players these days have fairly short attention spans, gameplay time isn´t really that much of an issue. Better to cut 5 or 10 hours from the gameplay total and not bore your players with repeating quests.

On the other hand, if your game mechanics are diverse enough you can try and put together a story free infinigame first (there were a lot of RPGs like that in the nineties where you could just play, wander through the landscape, do random stuff and not really get involved with the main storyline at all).
Then add your scripted encounters and plot and you should have a winner.

For putting together concrete missions just start up a list...

objectives / means / situations. then see how many unique gameplay situations you have and package those into randomised scenarios.
(also keep in mind: if one quest is long, say 4 hours or more, then you won´t need that many repeats during the course of a game)

Everyone's on the ball here, so I've got very little to contribute, and even that has been stated more or less, but I felt it deserved a spotlight, so here it is:

Whatever the player does, good or bad, ramifications should be clear.

The more consequence there is in the game, be it a certain kind of people really liking you or hating you depending on what you do, the cooler the quest-aspect of the game will be. Fallout was awesome for this... even if they weren't, the possibilities seemed endless. I played thru as a goodie two shoes, and the armies supported me, I got into the creepy Brotherhood, and bad guys trembled or shot at me on sight.

I went as a psycho-killer and everyone feared me, but super-bad guys gave me free weapons and were really giving. Bounty hunters came after my ass though...and they sucked!

Point is, do whatever you feel is reasonable to make all actions have a reaction in your game. Hope this helps.

Sorry i missed this post when it first came around.

Quote:

How will you feel, then, if you encounter the same quest again for a different NPC applying to a different situation? What's cool about this idea is that a reputation system that makes you weigh one thing against another would add so much to gameplay: Fail the mission and get an in with an exclusive faction, for instance, or keep raising your rep with a faction that may be on its way out-- could be an interesting choice.

Doesn't Galaxies handle missions this way? They're all pretty much the same context - Kill X, or retrieve item Y, or protect Z. Again its the context that makes the missions feel different. Most seem to be set up to combine exploration and faction standing with kill points and loot, depending on what the mission is. I've refused to take missions because I already had a precarious position with, say, Gungans and really don't enjoy getting ambushed by them at every turn whenever I'm out in the swamps of Naboo.

The game may not be the best in the world, but admittedly I did like their mission structure. I knew, in my heart, I was doing the same thing over and over (and travelling between worlds to get there), but because of the descriptions of the missions and the number of ambushes you could encounter along the way I found myself using stealth tactics as much, or more, than pure firepower.

Quote:

Original post by onyxflame I think if you're going to make random quests, they should be REALLY random.

Random doesn't really mean anything unless you specify some sort of distribution to the randomness. I mean, what's the difference between random and really random? Please do note that I'm not trying to flame you here, as I really want to support this idea, but to extend it the probabilities of different scenarios should be defined. The probability of generating a common, every-day situation (e.g. an outlaw is menacing the countryside) should be much greater than something completely out of the ordinary (e.g. the Order of the Obsessive-Compulsive Necromancers have summoned a Lord of Oblivion to enforce the installation of the Windows Operating System on every machine; *gasp* the horror... [lol]). You could also enforce some certain, very unlikely quests to be generated every now and then (to keep the player happy).

The actual structure of the quests should allow the player some tactical thinking and a freedom of choice. For instance, if Jack the Honourable knows that Archibald the Malicious has the Dreaded Staff of Ultraviolence hidden somewhere, and Jack thinks this staff should be destroyed for the good of mankind, the player should be allowed to achieve this in a number of ways. If the only way to accomplish this is to follow Archibald through the forest to a hidden cave where you can ultimately burn the staff, the replayability value of the quest is very low. If the player's character was able to beat Archibald into bloody pulp and to force him to eat the staff, the quest would still be solved, as Jack just wanted to rid the world of the staff. The quest should, therefore, be of the form "quest is solved if the staff is (provably) destroyed", not "PC must follow Archibald and then burn the staff". As Wysardry said:

Quote:

Original post by Wysardry The Foozle wouldn't have to be killed: it could be bribed, have a relative kidnapped, be stolen from, politically assassinated (have its reputation destroyed), be imprisoned or exiled, have its memory wiped or otherwise have its "teeth pulled". The same thing applies to the Foo: the Foo can be a single object, multiple objects, a person or information and the requirements for getting it can also vary (it's hidden, lost, owned by someone else, broken, dangerous in itself, a prize in a competition...).

I'm not sure whether Wysardry was giving examples of varying the objectives in the quests or examples of making the objectives more general, but I would like to emphasize the fact that most NPCs wouldn't really care how you achieved the goal as long as you achieved it. Thus the quest's completion trigger should really be the thing the original questmonger wanted; if the questmonger wanted someone dead, then killing (in one way or another) is the only option. If the questmonger wanted someone "out of the way", you have options, as killing is just one way of a whole plethora. Do note, however, that the actual in-game wording might not be exact. The questmonger could talk about killing someone even if he just wanted him out of the way. Or the quest description could even be something completely different, e.g. if a highwayman has stolen a peasant's fortune, the peasant will probably be in a very angry mood and tell the PC he wants the highwayman dead, while he would be completely happy if he just got his money back. Of course, this difference of the quest description and the actual quest should be somehow noted in the internal game logic as well, as it could have major implications.

However, coming back to the staff scenario, if the player had the choice and chose to kill Archibald in the process and this was not what Jack had in mind, either Archibald's survival should have been a part of the quest description as "Archibald must not die" (in cases where it really is important) or then it should not be a part of the quest system but rather the reputation system (in cases where Archibald's survival was not vital, such as if Jack respected all living things despite their flaws). Killing Archibald could have a huge impact on the PC's reputation and thus Jack might not give the reward for the quest to the PC anyway, not because the quest wasn't solved (since it was), but because Jack just doesn't like the PC any more.

I'm all for randomly generated quests. Personally I'd rather play a game with adequate random quests than a game with phenomenal fixed quests, as replayability means a great deal to me. Games like Neverwinter Nights and Morrowind are great once, but even with a different type of character they vary too little on consecutive replays.

Just please don't include random quests involving a nobleman asking to rescue pigs and hens that have become lost in the woods (an astonishing distance of four meters from the aforementioned nobleman!) as can be seen in Sacred... [lol]

And as for the original question, while many would consider clearly similar quests boring repetition (and ultimately it would be, unless the situations were immersive enough), a few times wouldn't hurt as the player would feel recognizing the similar setting and knowing what to do rewarding. The problem with unique quests is that after the quest you will know how to handle similar situations, but you never need to use that information (as the quest was unique); you wouldn't get the psychological reward of feeling competent next time in a similar situation.

By "really random" I just meant that rather than having a choice of 10 fully detailed quests that can pop up at any given time, you should have 10 different quest-piece positions which each can contain 10 quest pieces, giving at least 100 possible quests depending on how you implement putting them together. You could even make 2 separate quest generating programs, one for normal everyday quests and one for weird far out stuff, and pick between them semi-randomly (the variable might be affected by how long it's been since you did a big weird quest, for instance). I didn't mean that every quest should involve getting sucked into a toilet and killing living piles of crap in order to gain the Helm of Incontinence, just that there should be a *lot* of possible combinations so one game wouldn't be able to contain them all.

Quote:

Original post by onyxflame By "really random" I just meant that rather than having a choice of 10 fully detailed quests that can pop up at any given time, you should have 10 different quest-piece positions which each can contain 10 quest pieces, giving at least 100 possible quests depending on how you implement putting them together.

Well, yes, I did realise that this is what you meant, and I wholeheartedly agree that this is a great idea, but it just seems a bit silly to have something that is random and something that is even more random. I just find it somewhat annoying when people start talking about randomness without even mentioning the distributions at all. After all, that is what we are talking about; when generating random quests, it is the distribution (or the parameters of the distribution) that we need to define in order to make the quests plausible. But then again it's easy to degenerate an issue like this into a semantics argument and it's not really important anyway, so let's forget it...

Quote:

Original post by onyxflame You could even make 2 separate quest generating programs, one for normal everyday quests and one for weird far out stuff, and pick between them semi-randomly (the variable might be affected by how long it's been since you did a big weird quest, for instance). I didn't mean that every quest should involve getting sucked into a toilet and killing living piles of crap in order to gain the Helm of Incontinence, just that there should be a *lot* of possible combinations so one game wouldn't be able to contain them all.

Why not give each concept (the objectives of the quests) likelihood values and base the probability of each quest on that? For instance, you could have that a quest involving an outlaw has an likelihood of (say) 999999 and a quest involving the Abysmal Parallel Murderer of Vr'graalh has a likelihood of (say) 1. Now, in all the quests where one parameter has to be either of these two, once in a million times it will be the murderer.

Quote:

Original post by Grim Random doesn't really mean anything unless you specify some sort of distribution to the randomness.

Good point Grim because I see you're talking directly about player's experience as they move through the game world. Gunfights are expected, for instance, in frontier territory, but not in urban settings (necessarily). That you encounter some gunfight every other day on Planet Laredo doesn't really cheapen the value of this as a random encounter because the distribution leads you to expect it.

Quote:

You could also enforce some certain, very unlikely quests to be generated every now and then (to keep the player happy).

Yes, I think it's vital that you get interrupts to the pattern to throw you off, especially quests that look like one thing but actually are another.

Quote:

The actual structure of the quests should allow the player some tactical thinking and a freedom of choice.

In an way open to many approaches, yes. Easier said than done, but agreed. To some extent you have to expose the quest resolution testing logic to the player in an in-game fashion so that they can be very clear about what will and will not solve the quest. This will allow them to distinguish between destroying something and just getting it out of the picture temporarily.

Quote:

I'm not sure whether Wysardry was giving examples of varying the objectives in the quests or examples of making the objectives more general, but I would like to emphasize the fact that most NPCs wouldn't really care how you achieved the goal as long as you achieved it. Thus the quest's completion trigger should really be the thing the original questmonger wanted;

What really appears to be necessary here is an object which exists in the game world based on many different stats. The most simple paradigm is the enemy, which persists as long as there are hit points and does a bunch of things to lower HP in the hero or his allies. The gameplay for doing this resolves into a kind of grammar involving a subject and object, and this gives you a certain amount of flexibility with combat (does that make any sense?)

This grammar expands and becomes unweildy and even contradictory in terms of gameplay or context the bigger it gets. An example of this was a cheat where you could kill an ally in Ultima and then use their body as dead weight on a floor lever, then resurrect them, simply because the grammar included dead bodies as weight but not killing of allies.

Quote:

Do note, however, that the actual in-game wording might not be exact.

Yes, a bit of vagueness may either be refreshing or it may tempt players down the easy path. For instance, if you're walking around in Power Armor and get a choice between kidnapping and killing a guy, you may find it easy to do the latter until the former has a huge payoff (or the latter a drawback).

Quote:

I'm all for randomly generated quests. Personally I'd rather play a game with adequate random quests than a game with phenomenal fixed quests, as replayability means a great deal to me. Games like Neverwinter Nights and Morrowind are great once, but even with a different type of character they vary too little on consecutive replays.

We're in the same boat here. I want self-sustaining gameplay that allows you to ignore the plot, or even better, self-sustaining gameplay that is so hooked into the characters and environment that it IS the plot.

Quote:

Just please don't include random quests involving a nobleman asking to rescue pigs and hens that have become lost in the woods (an astonishing distance of four meters from the aforementioned nobleman!) as can be seen in Sacred... [lol]

[grin] Reminds me of the early Daggerfall quests to rob your own house.

Quote:

And as for the original question, while many would consider clearly similar quests boring repetition (and ultimately it would be, unless the situations were immersive enough), a few times wouldn't hurt as the player would feel recognizing the similar setting and knowing what to do rewarding.

I think the context drives everything here, as was mentioned about Galaxies. What I imagine is an encounter generator based on your reputation, territory and affiliations such that you eventually know you're the cause of what's happening to you in the game. That to me is the essence of choice, which is something we all crave more of in game.

From an RPG perspective I want players to start caring about improving their encounter stats, which support other goals in the game (skilling up or equiping or alliances). I've noticed it is possible to do repetitive things if the gameplay varies (every fight, sneak or trade not the same) and the context is building up to something (will this faction win? what is at the heart of the mystery, etc.)

Like life, it only sucks if it is empty and meaningless.