Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[ductape]

casserol, Virtz, and pilgrimboy get the gist of what I am after here. If you read that article I post on the first page, and follow along the points they made, I think that starts to lay the foundation to explain whats been tossing around my head lately.

I have been giving Ren'Py a good hard look lately and some of the games that come out of that. Personally, I am not a fan of most of what is created there but there are some interesting tools available in one of the frameworks that goes with Ren'Py, the dating sim framework. In a nutshell, it serves up events based upon variables that are set by the way the player answers dialogue trees. Sure it's not all the way there for what we need, but it outlines a basic framework for doing at least part of the job.

Not every option on a tree send you down a completely unique branch, but it can set some variables that affect the game in all sorts of interesting ways. For an example based on conversations between NPC and player, the player may have a number of options of things to say that range from anger, rebuking, flattery, etc. The option that player picks sets some variables indicating the NPCs opinion of the player and the response given might be dynamic to some dehree, drawing up the values of variables set previously and currently, doing the math, and grabbing snips of text that build up the response.

In the action scenes and dungeon crawling portions, we can present the player with rich options according to their equipment carried and player skills. Monsters would react differently if they have heard the player party battling their way through the dungeon and be prepared for a fight, laying in ambush, this alone is something we just dont see as players metagame their way through the dungeon assuming that each room is a tidy little package with hallways to heal up in between.

There could be dynamic events going on, there might be another party of adventurers who have similar goals, you might meet them in the dungeon, what will you do?

We have a gradient of design options here. On one end we have a fully written branching storyline which focuses on delivering a cohesive and believable narrative. On the other end we have something like a sim, with actors pursuing their motivations and the story is emergent. While the sim end of the spectrum provides us with a broad range of player freedom, it lacks the control the authors need to deliver a good story. I feel that somewhere in the middle might be a sweet spot.

Take for example, that other party of adventurers whom we run into the dungeon. We dont need to model out specifically where they are at any given turn in the game, we can plop them down in an encounter that is rich with juicy story potential. The details of the encounter would draw from previous decisions the player made while in town or other factors.

I dont presume that with current technology we could create the dynamic experience a human DM can create, but I do think we can create a rich experience that comes CLOSER than anything else out there. Even just a few baby steps in that direction would be ground breaking, I am not aware of anyone who has even attempted it.

[pilgrimboy]

I have always liked the idea of competing heroes in a game. Fables acted like that would happen. If you don't take a quest, others will. The quest isn't just going to wait on you. Then the other heroes could die, succeed and prosper, etc. There was a game that did this a little bit, but I can't remember the name of it. It had different parties going out to do the same quests as you. But was more of a Diablo II clone, and I didn't like it all that much.

But the idea would be a Skyrim game with a world that is progressing and changing whether or not you're involved. Some other hero will get the credit. If you spend your time going after the thieves guild quests, someone else is going after the Mage guild questline. Etc.

[Trif]

I don't find linear scripts for games all that enthralling actually. If I wanted to watch a movie, a movie does that better than a game, any day. And it doesn't kick sand in my face for not being "good enough" to finish the movie :/

Could you post examples of some games that have good non-linear stories? I honestly can't think of a game that has a good storyline, but isn't, at most, 5 linear paths. Or do you mean that all video game storytelling to date is bad, and sandbox games are the only enthralling ones (a belief that I can certainly entertain)?

Try Blue Lacuna. It features a lot of non-linear exploration, and the storyline changes depending on your earlier choices. There are some events that happen every time you play, but the way the story progresses really depends on your play style. Plus it has the best character development and interaction with an NPC I've ever seen in a game.

It's quite a long game (18-20 hours I think), but it's definitely worth the time spent, and the closest I know to a truly interactive story.

[Zangi]

Wow, Renpy has a heck of a lot of dating sims.

@pilgrimboy, ah.. I have that game I think.  The AI are idiots and just end up murdering each other... every time.  Honestly, I rather they go questing and defend the town.

Blue Lacuna hit me with a pretentious hipster vibe when I first opened the page.

[jhxmt]

I have always liked the idea of competing heroes in a game. Fables acted like that would happen. If you don't take a quest, others will. The quest isn't just going to wait on you. Then the other heroes could die, succeed and prosper, etc. There was a game that did this a little bit, but I can't remember the name of it. It had different parties going out to do the same quests as you. But was more of a Diablo II clone, and I didn't like it all that much.

But the idea would be a Skyrim game with a world that is progressing and changing whether or not you're involved. Some other hero will get the credit. If you spend your time going after the thieves guild quests, someone else is going after the Mage guild questline. Etc.

If you haven't played them previously (and if they're your type of game), you might want to check out the Stalker franchise:

"...for the first in the franchise 'Shadow of Chernobyl' it was stated that the AI had to be impeded because it could have completed the game by itself, characters were non-scripted and could develop even when not in contact with the player..."

  - from http://all-things-linux.blogspot.co.uk/2012/09/stalker-clear-sky-late-verdict.html

It sounds like a lot of what's being discussed in this thread revolves around the idea of improved artificial intelligence and artificial life in games.  Some games have made good use of this (such as Stalker and, obviously, DF to some extent!), but the problem is that you run into the same irreducible complexity idea.  Yes, you can have a lot of runtime variables being modified by the player's actions (and NPC actions), but at the end of the day the player has to see the outcomes of these actions or they're not going to 'get' the point of the game.

At which point, you have two options:
  1 - write a large number of detailed, in depth responses to a large number of possible 'game states' (i.e. "if player has been this horrible to this many people that this NPC knows, then say something like this.  If player has been horrible to slightly fewer people that this NPC knows, but is wearing something that looks amusing, then say this" etc.)  This requires a large amount of unique content creation - which is not something that can currently be done automatically without feeling incredibly stilted and artificial.
  2 - write fewer or less detailed responses.  This runs the risk of feeling repetitive or, again, stilted and artificial, but less content generation makes it more achievable.
  3 - write a detailed railroaded storyline, and make the sandbox outside that storyline slightly less detailed.  This is the case for games like Freelancer, for example, which had (what I thought was) a very well-presented story, but if you wandered off into the sandbox for too long the repetitive nature of umpteen different people saying to you, "We don't own this station, but we have an understanding with the people who do" would drive you space-crazy.
  4 - make it online-only and hire a human GM. 

Is it possible to do a highly branching game, to the extent that the player won't realise that they're branches rather than pure sandbox?  Yes, if you're willing to do a lot of content generation (and I don't mean the typical 'fancy graphics and voice acting' content here, either, I mean actual gameplay content, written dialogue, anticipated reactions, et cetera - you know, all that stuff a good GM has to plan for in a tabletop session).

Is it possible to do such a game without putting in a lot of content generation?  I don't think so - to even get a program to start generating content smart and subtle enough to make a player think it's hand-written, you'd need to write such a complex program that you might as well just write the content in the first place! 

Still, even achieving a good version of options 1 or 3 in the above list would probably elevate the game above most of the 'mainstream' RPGs out there at present.  DF actually seems to be progressing (slowly) in that direction, which is fantastic.

[Vattic]

If you haven't played them previously (and if they're your type of game), you might want to check out the Stalker franchise:

"...for the first in the franchise 'Shadow of Chernobyl' it was stated that the AI had to be impeded because it could have completed the game by itself, characters were non-scripted and could develop even when not in contact with the player..."

  - from http://all-things-linux.blogspot.co.uk/2012/09/stalker-clear-sky-late-verdict.html

I've only played the first and not finished it but I never noticed any of this in SoC. You'd find the same guys in the same areas doing the same things. Like that group at the start always waiting to raid the same house, waiting for you to prompt them.

[jhxmt]

If you haven't played them previously (and if they're your type of game), you might want to check out the Stalker franchise:

"...for the first in the franchise 'Shadow of Chernobyl' it was stated that the AI had to be impeded because it could have completed the game by itself, characters were non-scripted and could develop even when not in contact with the player..."

  - from http://all-things-linux.blogspot.co.uk/2012/09/stalker-clear-sky-late-verdict.html

I've only played the first and not finished it but I never noticed any of this in SoC. You'd find the same guys in the same areas doing the same things. Like that group at the start always waiting to raid the same house, waiting for you to prompt them.

Yeah, I believe that group is one of the, ah, "impeded" groups, because they need to be there as part of the story (to introduce the player to the world).  Allegedly, prior to the AI being cut back, initial testers complained that the game would end prematurely...because some of the AI stalkers ended up pursuing and completing the main storyline missions.  Not sure how true or accurate that story is, but I like to think it's true. 

[LoSboccacc]

I only ever saw branching RPG done right in Morrowind, which had the right balance between choices and still an entertaining story.

The problem in the end is that you have to win, and the game has to make its point about ending the story. Morrowind was so aimless that you could just get lost and lose interest and getting bored of it well before reaching the final; in the point of view of narrators, having only half the story entertaining is a failure.

This forces all games to have one set of ending convergence. What it means is that you have three/four endings for the same story arc (think fallout) so of the 1000 decision you made trough the game you soon realize that only three/four actually made the difference.

But there is a far more important problem in the end, which is, all those game can hardly be called 'role playing game' because you more or less always have to choose between the role of the one that saves the world, kills the baddie and gets the lady/gold.

Hardly roleplaying, if your choice between a rogue and a paladin is that you stab the orcs instead of charging at them.

My question is, is the average player ready for an open world sandbox? Many people are just that bad at setting their goals, if they don't have a big arrow pointing at the baddies they will probably get lost, bored and put the game away never to be played anymore. It is not just about telling a story and managing the branches/state, its about motivating the player to reach their goal.

So, how a sandboxy game could push the player forward to action? (think game dev story, that has the kind of pace that always pushes you for 'five minutes more!')
 


addendum: interesting post about the branching of stories in Jade(now adom 2):
http://www.ancientdomainsofmystery.com/2011/11/first-complex-story-arc-king-of-orcs.html

[Seraphim342]

Love the ideas jhxmt posted, and they're already implemented to an extent in some games.  Mount and Blade comes to mind, with the NPCs having basic goals like "sack towns" and "take territory" where they'll cooperate, form armies, etc.  I think the best example of this so far, though, is the X series, particularly X3: Albion Prelude.  They added in a system where NPCs form minor corporations.  An NPC trader will trade goods, and if he becomes successful enough he can start a corporation, hire additional ships, buy defensive drones to protect himself from pirates, et cetera.  The game's engine also simulates factories and such going out of business by deleting stations whose products are rarely bought.  One of the common complaints with the game is that the galactic economy tends to crash crash if left to its own devices without player intervention, but it turns out it actually had to be designed that way because if the game started with a completely stable economy it was way too hard for the player to enter the market, so they had to deliberately leave holes where the player could step in.  The other problem was that as the galaxy developed and tens of thousands of ships flew around with their own goals, more and more processing power was needed to simulate things the player couldn't even see (referred to as the Out Of System or OOS game), even with an extremely simplified, turn-based system handling these interactions.  This because especially problematic when player assests started becoming more prevalent, because OOS interactions outside the system the player is in and not "visible" by any of his assets were calculated every 30 seconds, but ones within range of his assets and therefore visible on the galaxy map were calculated every 5 to be more real-time.  When you control vast swaths of space full of thousands of ships the difference was huge.  It ends up being like DF in the fact that the end-point of any game is usually lag because of the massive amount of resources needed to simulate everything as the game expands.  Hardware's the main thing holding back artificial life for this reason, as you can get hundreds of thousands of scripts piling up and executing constantly. 

As far as Stalker goes, from what I've read about it the NPCs run on an objectives and inventory system similar to the player.  It had to be dumbed down a lot to make the game enjoyable, though.  For example, NPCs originally would have had finite ammo and would have had to scrounge for it or loot corpses, but that led to major ammo shortages throughout the game, and increasing the supply to a sustainable level would have made things too easy for the player, so they gave the NPCs the standard invisible ammo.  They also swapped out their weapons for better ones, but this also led to problems with NPCs scooping up all the high-end weapons and running around with them, so it was cut as well.  You can still see remnants of it by tossing an AK for an NPC who only has a pistol or other secondary weapon.  He'll pick it up and fire it but run out of ammo for it unless it's a submachine gun that fires the same caliber as the pistol he had, of which he has an infinite supply because of the changes.  Then there was the most famous instance of an NPC given the goals the player is given during the main story and completing the game.  They had to tone down the goals system as well because if one faction got a slight upper hand they'd usually exploit it ruthlessly and take out pretty much everyone else, so it basically boiled down to "Are you friends with the Duty?  Cool, you never have to fight anything.  Enjoy your leisurely stroll through the Zone. Oh, you're not friends with Duty?  Get ready to fight Duty for the entire game."  Even with the dumbing down you'd still see it a lot in Clear Sky where the Bandits or Freedom would take over everywhere they were permitted to take over before you even got there. 

The other thing to be aware of is that computers don't think like people, and a lot of times you have to put a lot of restriction on how they go about accomplishing goals to keep them from behaving in completely unforseen ways.  Sometimes these are awesome, like the NPC beating Stalker, but sometimes they're a major WTF? moment. 

[ductape]

The examples we are talking about now, Mount & Blade, Stalker, X series are not really on target I think for true narrative. They are more of the sandbox nature where story is emergent, up to the player to imagine the narrative bits rather than the game actually spelling it out.

Emergent story is awesome, but to get it to be truly coherent from dynamic actors who are moving around in a sandbox sim would be an amazing feat of master coding. Again I think somewhere in between emergent story/sandbox and straight up narrative from story branching would be interesting and something I havent seen.

For the main plot and sub-plots, one possible approach would be to have blocks of smaller story pieces that can be arranged in a variety of ways to make up the whole. Within any block, we can be rich with narrative and provide numerous branches within it. As the player interacts with the elements within the story block, variables are set as needed. The magic happens when we start to mix and match the blocks that the player actually moves through during any given playthrough.

Likewise, some of the blocks can lay in the other end of the spectrum and be a completely dynamic sanbox/sim. The goings on within that sim will affect the overall availability of story blocks and elements within.

I am also mulling over the endings and obviously this can be improved. Instead of having a set number of endings, such as the "good" ending and the "bad " ending we could have endings that tally up the goings on within the entire game and present an ending that is based on many factors. For example, in a roguelike you are often presented with a page that has your score and that page often includes many other stats such as how many of X moster was killed. The RPG ending might learn something from this. If we can somehow take the elements of a score page and instead present the player with story and words, we might have something very cool. Exactly HOW happily ever after does everyone live?

[Mech#4]

Is this all like how, in Mount and Blade for example, you can take over a castle to become a baron, but there's no real point to talking to that peasant because they have nothing interesting to say?
I know that I usually won't do some things in games like Morrowind or Oblivion because I can be reasonably sure the game won't react to it in the expected manner. Does killing Vivec have any impact on the church, does leaving the oblivion portals result in other towns being destroyed etc.

[Seraphim342]

I know that Stalker, Mount and Blade, and the X Series aren't really shining examples of magnificent narrative.  Was just using them as an example of emergent, non-linear gameplay =p

Overall, I feel like the best solution would be to have the majority of the "meat" of the game be based on an emergent system, with pre-written story events occurring when certain conditions are met.  Probably not the best way to explain it, but I'll try anyway.  For example, let's take the Civil War arc of Skyrim.  Basically, it's on rails with you taking the same forts and cities in the same order, and each encounter's a point A to point B experience.  Sure, every once in a while you'll see a random Stormcloak patrol and fighting an Imperial patrol, but that's about the extent of it. 

What would have been much better is to have them fighting back and forth with or without your input, taking and losing territory.  Have the war affect how NPCs behave, and even to an effect how the scripted quests play out.  For example, higher taxes and curfews in an Imperial-occupied Stormcloak town and vice-versa, more bandit activity in areas of heavy fighting, even have armies set up camps or occupy random points of interest like the random abandoned forts all over the map.  If you're doing missions for the Thieves Guild in Riften and it's occupied, the characters talk about the damned Imperials making their jobs harder, and having different targets and objectives for missions, or change how they play out. 

If actually you join the war, you get much broader goals.  Instead of just "single-handedly slaughter the entire enemy army at these locations," you affect the course of the war in less direct ways.  Have a high blacksmithing skill?  Forge some weapons and armor for the soldiers at a certain location to help them push back the enemy.  Ambush enemy patrols to reduce their overall strength.  Assassinate enemy commanders. Escort dynamically-generated supply caravans.  Bribe or eliminate bandit groups.  Any of that would have been better than "walk into this fort and kill all the bastards to claim it for the Empire."  Instead of "oh, you brought back that crown thing, congrats you're a Centurion,"  have the ranks mean something. "Congratulations on your promotion to Centurion.  You're now in command of 3rd Cohort." Then you can command your men to attack various locations.

A lot of this would have been really easy to implement with the radiant AI and story systems already in place, probably even easier than making some of the missions that were actually in the storyline.  Overall the whole civil war arc, which in terms of the story should have been HUGE, felt unfinished and barely worth a sidenote. 

[Virtz]

You forget that the Radian AI is a load of bull that never actually did what they advertised pre-release, and the things it does in-game look really ridiculous. Like anything you try to make epic with their engine ultimately ends up looking really stupid. The most you can get is like 10-20 guys running in a completely disorganized and comical fashion. It just doesn't work.

[Neonivek]

You forget that the Radian AI is a load of bull that never actually did what they advertised pre-release, and the things it does in-game look really ridiculous. Like anything you try to make epic with their engine ultimately ends up looking really stupid. The most you can get is like 10-20 guys running in a completely disorganized and comical fashion. It just doesn't work.

The problem with Radian AI is that it doesn't work well WITH ITSELF!

[Seraphim342]

Yeah, it eventually worked out to "If I walk past a chopping block I'll probably use it whether or not I'm a woodcutter," but the concept was there.  Just was never finished, like a lot of the game.