Quest Structure: Proactive vs. Reactive

[shihonage]

Norfleet : Eh. Have you been conditioned by Mass Effect? Because you assume a ton of linearity where there isn't necessarily much.

I have rarely, if ever, seen a game that offers more than token diverence from one, maybe two, possible branches. Even some of the Codex's most cherished games really diverges very little, contentwise. You have, at most, maybe a half-a-dozen possible ends, most of which diverge only slightly in terms of the main plot, where you did one sidequest differently, or resolved the end by saying different things, and maybe getting a different end battle as a result.

Did you rush all your characters in WoW to level cap as fast as possible?

If you did, then there's not much that will get through to you.

To me, the process is about the journey more than the destination. Because that's where you'll spend 99% of your time. Not masturbating to different states of the ending, but PLAYING THE GAME.

Fallout 1/2 opened up a great deal of potential in regards to a world that is a sum of flexible modules, and even if individually those modules offer limited outcome, the overarching effect becomes "dithered", as in, you start to see the forest instead of the trees. The overall feeling of freedom is there, because you have a variable set of interactions with variable outcomes, which manifest through player skills, dialogue, and multiple approaches to quests.

Realistically, as long as the game doesn't make you facepalm as you realize that NONE of your choices are something you would like to do or say, it is a grand fucking success. Because overwhelming majority of RPGs make me facepalm in this exact fashion, but Fallout did not.

Compare that to, say, the way a Paradox game can unfold completely differently if you reload from a mid-game save after reaching an ending: You will almost certainly make completely different choices, and not simply because you wanted to "See the alternate ending" or because some random element just didn't go your way. Something completely different will happen, and yet none of this will be "wrong", unlike RPGs, where failure to reach a given outcome is either because you chose specifically to do something different at a fixed branch point, or because you screwed something up (or unscrewed something up that you previously screwed up).

RTS can survive on mechanics alone - it has no obligation to give a meaningful written narrative to player actions. In an RPG, narrative is a mechanic that has to be interwoven closely to everything else. As long as we don't have AI that can imitate a live human dungeon master, there will be a multitude of limitations imposed on the game's mechanics so they play nicely with a dungeon master that CAN actually be coded to be functional.

Consider the one of Codex's favorites: Arcanum. There's a lot of different mini-branches, ways to resolve or not resolve a quest differently, and your decisions are ultimately acknowledged in the epilogue...but you're still essentially on a railroad: You go from A, to B, to C, deciding what to to do or not do along the way, and you cannot really deviate from the final outcomes. The final confrontation always unfolds in the same place, etc, and you never truly define your own goals. Each and every outcome is essentially a multiple-choice prompt predetermined in advance. You can pick from the set of options, but you cannot create any.

Can't speak of Arcanum, because the game turned me off from the very beginning and I never played past that. But Fallout formula is not a railroad. Your experience is always different, even if it ends up with just a couple of choices in the very end. And as I said before, obsessing about the ending alone is kind of stupid.

Yes, it is possible to use the same modular approach used for the rest of a Fallout game, and have a whole fuckton of endings taking place in different locations with different NPCs involved. It's not a limitation of AI, it's just a question of development resources and being able to maintain a strong narrative. And at the moment, it doesn't seem quite worth it, considering how every RPG these days fails horribly at learning lessons of Fallout - long before getting to the ending.

 

[Norfleet]

Did you rush all your characters in WoW to level cap as fast as possible?

I don't play WoW, but yes, that's sort of the goal in such games.

To me, the process is about the journey more than the destination. Because that's where you'll spend 99% of your time. Not masturbating to different states of the ending, but PLAYING THE GAME.

And yet, the level cap *IS* where you spend 99% of the time. Everything else is just a thing you haven't completed yet, an unrealized build in progress. To fixate on the process and not the goal is the mark of a slacker.

Fallout 1/2 opened up a great deal of potential in regards to a world that is a sum of flexible modules, and even if individually those modules offer limited outcome, the overarching effect becomes "dithered", as in, you start to see the forest instead of the trees. The overall feeling of freedom is there, because you have a variable set of interactions with variable outcomes, which manifest through player skills, dialogue, and multiple approaches to quests.

Fallout was...better than most.

RTS can survive on mechanics alone - it has no obligation to give a meaningful written narrative to player actions. In an RPG, narrative is a mechanic that has to be interwoven closely to everything else.

Indeed, it seems to be a limitation of the genre. This is why you won't see an improvement outside of those hybrid strategy-RPGs.

 

[Castanova]

Furthermore, if your game's story is to have any unique aspects to it, like NPCs with unique dialogue trees, quests that aren't randomly generated, etc, then you have a lot more work to do. Writing in a proactive game evironment like this can be a lo harder. Specially if you want the resulting story to have a good "timing" to it. Instead of writing linearly, you need to think of your story as springing from the game structure instead. Not to mention that you have to consider how different things the player can do can change how things go. I don't think this is impossible to do at all, at least to a much greater extent than has been done in computer games up to now. But still, given how much work it would require, and how little most gamers seem to care about this aspect, it isn't all that surprising that most games simply ignore that.

I think perhaps there is an non-ideal middle ground available if someone were to try to take a lower-cost stab at a concept like this. Basically, you model NPCs as part of the simulation including their motivations/reactions/personality/etc., but you simply skip the concept of dialog. Dialog could be instead described narratively. Like, instead of showing what the character actually says to you, the game would simply describe what they said. ("Bob looks scared. He tells you about a problem with werewolves in a nearby cave.") Then you could handle the problem however you see fit and you don't need to write specific dialog outcomes.

You clear the cave. "Bob looks relaxed."
You ignore the werewolves and you come back a few days later and Bob is dead.
You ally with the werewolves and he sees you with them. "Bob is scared of you." and maybe he even just runs away from you on sight.
You ally with the werewolves, which he witnesses, and they kill his wife while you're not even around. "Bob wants revenge for his wife. He attacks you with a pitchfork!"
You feed the starving werewolves with Bob's crops. "Bob is upset. Bob has no food supply." then later "Bob is hungry." then later Bob walks to town to get more food. If he has no money, he ends up trying to steal food and guards kill him. You may or may not witness this happen.
Your entries to the dialog would be topics like in Wizardry and perhaps there's a system to link topic nouns and certain verbs to communicate with NPCs.

Things like that. Quite similar to some of the things Dwarf Fortress does except you focus much more on simulating NPC behavior and you use it in a game with a clear problem to solve. You would basically need to model each NPC's wants/needs/mood/relationship to other NPCs and then have a dynamic way of describing that to the player.

 

[shihonage]

And yet, the level cap *IS* where you spend 99% of the time. Everything else is just a thing you haven't completed yet, an unrealized build in progress. To fixate on the process and not the goal is the mark of a slacker.

1) When it comes to the subject at hand, this is completely and objectively false - nobody plays games like Fallout/Arcanum/Planescape to rush through the game content as fast as possible and then spend 99% of their time at the "endgame" (which doesn't really exist). So your complaints about finite end-states still hold little value considering that the majority of player's time is spent elsewhere.

2) When it comes to MMOs, a lot of people play this way, rushing through, and it is their choice, however it is not the only way, nor does it strike me as particularly fulfilling. I play(ed) MMOs to enjoy the content and enjoy the feeling of progress, customization and power my character experiences.

You are quick to use derisive language, one which coincidentally gives hint of your attitude toward games. "Slacker", for instance, means that you see the leveling process in MMOs as necessary "work", instead of a game - something fun that can be paced and enjoyed in cooperation with other players.

This is usually the mindset of someone who has not had enough exposure to what constitutes "work" in the real-world, and, needless to say, our demands from entertainment are too different to bother prolonging this exchange.

At some point you may come to realize that life itself is not about the goal, but about the process. If you are not here NOW, having fun NOW, then you are not really alive.

 

[SCO]

Things like that. Quite similar to some of the things Dwarf Fortress does except you focus much more on simulating NPC behavior and you use it in a game with a clear problem to solve. You would basically need to model each NPC's wants/needs/mood/relationship to other NPCs and then have a dynamic way of describing that to the player.

Would be a cool sandbox game.

 

[Esterhaze]

The problem is the quest based model of presenting challenges to the player. Even with branching quests, you are never going to get any sort of "emergence", and very little feeling of proactivity. But your goal is not out of the reach of modern AI: I've argued before that strategy games have been doing proactivity well for years. It's all about designing a world with factions that can be strengthened or weakened, and keeping track of the right numbers. These numbers are going to be things like faction wealth, territory, prestige: different measures of health, just as they are in your run of the mill strategy game. The reason why combat is the center of so many rpgs is because it easily provides this sort of proactivity: you can choose different tactics, based on your character skills and the situation, to achieve a pretty straight forward goal of reducing the enemy's health points. I see no reason why this couldn't be implemented on a factional, and thus narrative level in an rpg. This may put a novelesque story out of reach, but who cares, rpg stories suck anyway. Better to focus on settings, themes, scenarios. Characters too: nobody will ever accuse JA2 of having flat characters. If you want some narrative determinism, take another page from strategy games: use triggered events, modifiers to individual factions, and design in a way that there are only a few possible trends and macro outcomes to the game, such that npcs can reliably comment.

 

[sea]

It is interesting to consider the implication of journal systems on player sense of autonomy. Most games feature journals for organizational purposes - players do not play all games endlessly without breaks. They will leave for a few days, or weeks, or months, and come back and expect to be able to resume play more or less at the same point they were before. This is a reasonable demand from an audience and developers are reasonable to accommodate it through journal systems. Yet the very nature of a journal system as a means of organizing and documenting tasks feels like tells the player what there is to do and how to do it.

There is a way around this by allowing the player to write his or her own journal entries instead. However, this is not convenient at all, which is also a bad thing to have in any game (especially one which already has lots of reading, numbers to tinker with, menus to thumb through, etc.), and the last thing designers want is for people to feel like they have to do bookkeeping for a game. Even worse is if the player forgets to write down what to do, leaves the game, comes back and then has no idea what to do - they'll feel their time is being wasted just to get back on track.

There is also the simple fact that some players like journals that are very detailed, and want every quest to be documented, because they see it as a checklist - a handy guide of "what can be done in this game." I admit sometimes I have this same feeling, and I often want to drive myself to "100% finish" a game, which is very often motivated by the presence of a journal with well-documented quests. So in that sense there is actually an advantage, if the game you are making is not necessarily based so heavily around freeform play and exploration to begin with.

In the end I think that the solution proposed by shihonage is very fair: provide quests to the player in a way that feels proactive rather than reactive, and have sufficient quantity of gameplay that is free-form and not fully documented to the player. Side-quests which are not stored in the journal are a great example of this. Let the main story beats be displayed there for all to see, but save the sense of exploration and autonomy in everything else. So long as you don't have the player being asked to do X or Y for no good reason other than the plot says so, and better yet, have the player initiate and "control" the action in the story, you shouldn't really run into major problems with all this.

 

[J1M]

The problem is the quest based model of presenting challenges to the player. Even with branching quests, you are never going to get any sort of "emergence", and very little feeling of proactivity. But your goal is not out of the reach of modern AI: I've argued before that strategy games have been doing proactivity well for years. It's all about designing a world with factions that can be strengthened or weakened, and keeping track of the right numbers. These numbers are going to be things like faction wealth, territory, prestige: different measures of health, just as they are in your run of the mill strategy game. The reason why combat is the center of so many rpgs is because it easily provides this sort of proactivity: you can choose different tactics, based on your character skills and the situation, to achieve a pretty straight forward goal of reducing the enemy's health points. I see no reason why this couldn't be implemented on a factional, and thus narrative level in an rpg. This may put a novelesque story out of reach, but who cares, rpg stories suck anyway. Better to focus on settings, themes, scenarios. Characters too: nobody will ever accuse JA2 of having flat characters. If you want some narrative determinism, take another page from strategy games: use triggered events, modifiers to individual factions, and design in a way that there are only a few possible trends and macro outcomes to the game, such that npcs can reliably comment.

One of the common problems with the early attempts at this approach which is especially highlighted by Civ III, IV, and V in comparison to Alpha Centauri is that you need to design the factions to be radically different from each other in order to have the player view this sort of emergence as meaningful.

 

[Lorica]

Esterhaze: Exactly!

sea: I suppose I've always felt that poorly implemented journal systems are somewhat ugly. Nothing can make a game feel like you're nothing but an inefficient replacement for a tidy bit of script than a journal system that looks like a list of binary variables. Keeping your own notes on bits of lore--discoveries, pending questions, a note on how the fictional universe works, vague instructions about where you should explore next--is a pretty special part of playing cRPGs that you don't really get in other games. Convenience be damned, I'm one of those sorry fucks that likes keeping a notebook on my desk and poring over it in a spare moment when I'm not playing the game.

Personal feelings aside, journal systems kind of train you to ignore things that don't pop up in your journal, don't they?

The best sort for a cRPG, I think, is something that just collects all the things you've read or heard in game without simplifying it.

 

[Norfleet]

You are quick to use derisive language, one which coincidentally gives hint of your attitude toward games. "Slacker", for instance, means that you see the leveling process in MMOs as necessary "work", instead of a game

Of course it's work. You have a goal, that you're attempting to reach. Everything else, therefore, is an obstacle to be destroyed. This is work.

- something fun that can be paced and enjoyed in cooperation with other players.

Why would I enjoy things? I don't enjoy things unless I'm paid to enjoy them. Enjoying things for FREE is a bad and fundmentally unprofitable behavior.

This is usually the mindset of someone who has not had enough exposure to what constitutes "work" in the real-world, and, needless to say, our demands from entertainment are too different to bother prolonging this exchange.

Hardly. I have this exact same approach to work. It was very effective, and also very frightening, which is why I am now retired. My work is both respected and feared, and people fear it enough that they're willing to pay me not to do it at all. In my line of work, I solve problems. I solve problems very effectively, problems that other people do not want me to solve, and they're willing to pay me not to solve any more problems.

At some point you may come to realize that life itself is not about the goal, but about the process. If you are not here NOW, having fun NOW, then you are not really alive.

No, life is a problem to be solved. That is what I was paid to do. And having fun?

 

[sea]