next »

[The_Flying_Dove]

Yes, I have played Fable, and I think that it is a step forward in moving RPGs away from just being all about combat and cut scenes. The good and evil choices were very black and white, instead of gray, but you could interact with people in other (funnier) ways. Marrying people was a more realistic, unique approach. Fable 3 looks promising as well. On the other hand, it's too bad that the second Fable game's story didn't fair so well. Other than that, the touch feature and playing as a king will prove to be interesting additions to an RPG.

Maybe having an RPG released that is along the lines of Harvest Moon, Fable, Animal Crossing, and the Sims, along with a form of grey morality (a la The Wither or Bioware's DAO) combined would be good.

[gimymblert]

Rune factory?

(RPG + harvest moon's social settings + pokemon's monster taming)

[iggie]

My favorite example for gameplay in cities is X-Com: Apocalypse

I appreciate the desire to make cities deeper in RPGs, but I still found the simple world map more satisfying in earlier XCom games to the complicated cityscape.

I've tried realistic cities in game engines - they cover many km and are very dull with mostly residential, and industrial areas, with a small core of commerce and offices. In a story-based RPG it is best if the player only has to deal with a refined slice of the city to follow the storyline.

Moving towards a more sim type game, if you leave story behind and let the city operate under AI there is more scope for larger complex towns. I auto-generate missions in my current game to allow the player to improve or defend the city's needs. You have to magnify the effect of the players actions to make it interesting - eg: destroying a hijack operation on imports of platinum makes the price of platinum armour halve or somesuch.

I think "Deep Towns" is a noble direction to head in - but very difficult to turn into a fun gameplay mechanic. Handwritten miniquests ala Fallout are going to be more fun.

[TeeGee]

VtM: Bloodlines was all set in a city. It broke it into small managable parts to allow for high level of detail and memorable locations. The city itself was GTA like - lots of faux-npcs wandering the streets. The whole thing happening at night also helped with lowering expectations of how much city life you should see for it to be believable.

The World Ends With You was another RPG with urban setting. It went further and made the city, its pulse, trends and mythology the central point of gameplay and storyline. It had lots of cool stuff like your actions affecting street fashion, monsters inspired by grafitti and plot making heavy use of urban legends. 

[bart_the_13th]

Have you played Fable? It strikes me as very similar to what you want. If so, what was your opinion of it?

Ah yes, Fable, how could I forget the game. I once slauhter the whole citizen in oakvale town(the village where you can do chicken kicking contest), buy their abandoned house and rent them, wicked.

[nihilocrat]

Daggerfall did a pretty good job with cities, except for the fact that in a given region they all looked amazingly identical. On the plus side, you could really actually go into just about every building, and the cities were actually a believable size.

Darklands cleverly turned city excursions into a largely text-based affair with some hazy backdrops. If you managed to get in a fight with some thugs at night, you were transported to some anonymous section of city arranged in a mazelike pattern. The advantage of both of these is that the player will imagine the city as big or as small as their mind wants it to be.

[nikki]

i'd look into procedurally generating a city,
this is a place to start :
http://ccl.northwestern.edu/papers/ProceduralCityMod.pdf

[starsrift]

I think the issue is more that RPG gameplay is largely centered around defeating monsters, whereas cities are supposed safe havens, free of monsters. It's more of a question of "how to make a city interesting that isn't just a waste of work and detail"? This leads to the question, "What would an adventurer want to do in a city? Buy stuff, sell stuff, talk to people". You can try to put monsters into a city, but then you kind of enter a siege situation (IE: NWN1) where the city just becomes another dungeon with a different background, and some extra NPCs scattered about, maybe in need of saving.

I think a better question is, why aren't dungeons more like cities where the monsters feel like it's a 'safe haven', free of sword-slinging dogooders?

[SirNiko]

RPGs are also about solving mysteries, exploring for treasure, or just looking for places you haven't yet seen. All of these things can be placed easily in towns without violating the feeling of safety and tranquility.

[Bree]

Zelda falls into the more actiony-RPG side, but I always thought it did towns well. It fulfills all the requirements- they're safe havens, they're memorable locations, and they have tons of little secrets. Plus, they usually have extra little quests you can partake in between the main story.

The example that came to my mind was Windfall Island- the first sleepy little town you visit in Wind Waker. You have your functional areas, shops and whatnot, but you can also play hide-and-seek with the little kids, play minigames, collect pigs, bid on items in an auction house, and find people to trade items with in that ridiculously complicated sidequest. It's probably worth nothing that it's also a really compact map, with lots of interactive things in close proximity. This is probably the best strategy for a developer trying to make a realistic-feeling place.

[The_Flying_Dove]

I guess that I was thinking more about a game like Facade, with much of a city to converse in. Game developers have used combat in their game design so many times, such that it may be the time to make better use of parsers, as voice recognition and AI grow more advanced.

In fact, complex, believable conversations are probably going to be the next big thing in a decade or so, when voice recognition will not be as expensive. EA and other companies have now begun going down that route. Soon, you will see sports games with much more varied dialogue coming from sports commentators. And maybe RPGs will allow players to say whatever they want, as opposed to having predetermined dialogue in a nonlinear game like Mass Effect 2 or Fable 2. Nailing the idea of fluid, varied voiced NPCs will not be easy, but that is a problem that game developers need to tackle in order to put forth more creativity into gaming.

[gimymblert]

The problem of conversation is not only one of technology or AI.
IT's also one of plain ergonomy and design.

1) Ergonomy in the sense of surface of action:

- Generally game interface does not exceed in control action that can be make in a game. Players are constrain to a few verbs and noun and everything flow from that.

- NLP or Voice recognition suppose a surface of action which is just TOO huge as it equate to langage. This mean that the game must hold every objects, interaction and verbs that reality can contain in a given langage, plus all permutation.

2) Design as Focus:
- The beauty of having restricted verbs is that, when all you can do is jump, you know every problem will revolve around that verb. With natural langage you blow that focus and it can lead to guess the right interaction, which already happen in IF.

- It put pressure on the designer side, they must cover the combinatorial explosion that langage lead into.

It's not something which going to happen, it's bad design and it's bad ergonomy. It also not something that can be resolve through brute force.

In order conversation evolve as an interface, we should find convenient way to abstract it, just like we have abstract motion. It should have clear boundaries and feel natural.

Today's game handle this through object as symbol or mediation, like in majora's mask where mask act like topic or in harvest moon where a blue feather is a proposal. Can't find any more natural and efficient way to handle this, except by adding extra interaction on object. Non verbal langage is also underuse in game.

[SirNiko]

As a specific example of what Neoshaman is saying, look at Scribblenauts, a game that does allow for the player to input a near infinite number of suggestions.

The player quickly finds that complex solutions do not work. You decide to tether a giraffe to a UFO to carry it to the zoo, only to discover the game isn't equipped to make that happen. You can't summon milk and a sun to make cheese. The player quickly finds the joy of infinite possibilities is broken because only a few possibilities actually work.

Even in the case of tabletop roleplaying, where you literally have a human being playing the role across from you, this frequently presents a problem. Given too much leeway, players get bored or confused because they have too many choices and no information about what choices they should make. They also run aground when they deviate too much from what the developer planned - that railroaded encounter with the guards on a narrow bridge is more likely to be fun than whatever combat logically follows the player's choices. In a game with infinite possibilities, Samus just nukes Zebes from orbit, Mario calls the military to help him take down Donkey Kong, or Cecil murders the false king in the opening cinema. Freedom turns out to make the game less fun, even when it seems logic dictates it should be more enjoyable.

There's a market for reality simulator games like you describe, Dove, but I think it's a very tiny niche and certainly not something that will change the face of all gaming. Facade is great for Facade, but that kind of dialogue wouldn't make a game like Final Fantasy any better, and probably would have made some parts of it much worse.

[The_Flying_Dove]

It may be a niche market right now, but that will change in the future. As I've said before, today's biggest game publishers have begun to make use of this technology, coming from a company called Phonetic Arts.

If what you say is true, and if games will never see conversations reach their full potential, leaving all story-driven games to mainly be about combat and nothing else, then David Braben, creator of the popular and highly influential Elite franchise is making a foolish choice by investing in this type of technology. In fact, his next game, The Outsider, is supposed to offer nonlinear narrative that is beyond anything you've ever imagined. Heck, he even claims that it will not need cut scenes to drive the conversations forward because it will be that good.

Maybe he is so full of himself, and maybe everyone else who sees potential in this as being the next big thing, perhaps even bigger than motion control technology is foolish, too. I don't know, but I definitely think that we need to rethink everything, in terms of game design and genres, and what better way to do it than through carrying on with this kind of technology?


Also, there has been a company that has been focusing on interactive drama for over 30 years now, and its work is pretty impressive. If you take at the following web site, you should be amazed as to what this technology is capable of doing for interactivity in games: http://www.idrama.com/idihome/index.htm

[gimymblert]

Well saying that leaving storytelling to combat is quite a shortcut with so many games out there that try to push the boundaries.

From simple to complex solution:

Phoenix wright is barely interactive and you have almost no choice (there is only one solution), but with clear goal and clear context mixed with great editing give you enough illusion. This game is all about FOCUS.

Animal crossing let you interact verbally through letter with neighbor, again it use clever boundaries (local social convention) to keep the illusion from shattering too much.

Majora's mask use object economy to interact with the world, putting a mask (or showing an object) is akin to asking a question or talking about a topic. It keep things focus and meaning change regarding context, that the same strategy used in harvest moon and tokimeki memorial.

The act from cercopia (http://www.cecropia.com/) tried a new way to handle conversation through non verbal communication (which is said to be 80% of the actual communication), and some dating sims include "emote" to use through a dialogue to change the perception of the Dialogue choice. Some like sakura wars take in account the reaction time to infer tone.

The sims have some indirect conversation system through motive selection, it's one of the biggest selling game. Fable? Oh yeah that one is crap regarding the subject at hand >.>

And IF have been pioneering for YEARS nlp, voice recognition is an extension of the input. If you follow the design of Emily SHORT on her blog, you will see they are YEARS ahead in term of gameplay of conversation, and yet struggle on the very same obstacles i have named (the interface not the structure). The idea of "Quips" to handle conversation is very powerful. That and you could see chatter bots also.

A lot have been done, so far it's still a problem of input regarding surface of actions. I hope someday someone would go through all those ideas and synthesize a correct conversation system, as abstract as we use for motion: We need the D-pad of conversation not true conversation.

EDIT:

How does Virtual Conversations® technology work?
Users conduct virtual dialogues with video subjects—real people—by speaking into a microphone attached to the computer.  The user asks a question and the character responds, and a face-to-face dialogue ensues.  The programs have an Intelligent Scrolling Prompts feature which provides relevant questions for the user to ask.

[The_Flying_Dove]

It's worth a try. And the only way to do that would be to, as I said before, rethink everything. Perhaps one should consider working for a company that is still advancing this technology, and with enough experience, one can later go into implementing it properly into a game. That is what I might consider for my future. Work with a company that is focused on creating voice-recognition, interactive conversational experiences, before actually working on a game that makes full use of it.

Studying interactivity in general is also a good idea. I've read Chris Crawford on Game Design, and I will probably go on to his book dealing with interactivity, along with any other authors that deal with that topic.

This is the only way that we will get anywhere within this medium. Building upon layer of layers with only minor innovation features that are within the scope of existent genres will not lead to as much of a meaningful, artistic game as we would hope for.

[gimymblert]

But the problem is not the technology as i have state! The problem does not rely on voice recognition but on what we can do with it. IMagine we have it. Now try to think about the problem state: How do you code common sense? It takes one's life to have it and yet!

And for me, you are caught in realism fallacy, the more real a thing the artistic it should be? Novel would never have a chance, written world is nothing like emotion, they are the Dpad of thought.

Voice recognition would be like wiimote or kinect, he would still rely on abstraction to limit it scope. No matter how precise they can capture movement, the meaning behind them is what is hard.
http://blog.emaki.net/2007/08/video-visual-grammar.html look at the second part of the talk you will understand what i mean.

[The Bag]

Sorry to bump this up (also sorry for the inevitable typos since I'm writing this in my phone), but city & world simulation is one of my favourite topics.  Origin used to have the slogan We Build Worlds and that's always stuck with me, there are games which are sets and nothing more, you can only ever see what the designers want you to see at that moment in time, e.g. God  of War but there are others that try to create a world for the player to be immersed it.

By night I may work on my own stuff but by day I'm a gameplay programmer for Lionhead and was heavily involved in the design and did most of the implementation for the village sim in Fable 2.  Our goal was not necessarily to create a realistic world, but to create a believable one.  NPCs have jobs and routines, we made everything as data driven as possible.  Our entity system is a component system so if a building has a shop component then it's a shop, the shop data determines what kind of shop it is and also contains things like dialogue so the code can be as general as possible.

No sim npcs are placed in the editor, they're all generated by our population functions - buildings specify how many people can live there and famlies are generated, business specify what type of employees are required and jobs are assigned to the npcs.  I'm theory because we already create the population of our world like this we could move towards procedurally generating the city itself, but at the moment that seems unlikely.

All our villages have routines defined for each job type, npcs eat, sleep, go to work, relax, etc.  But a big mistake we made was to spend a lot of time working on behaviours no one would see, we've got a large set of home behaviours which are all indoors.  Add to that shop and bar behaviours it meant most of our npcs were indoors and it made the world look very empty.  So we limited the amount of npcs who could be indoors during the day and scrambled to add some more out door spare time activities.  GTA is very good at doing little vingettes of life, we just didn't have the resources available when we found this was an issue.

GTA's an interesting sim, it creates a great illusion but it's ultimately a shallow one, e.g. I remember following a cop who went screaching up to a house in GTA IV, lept out the car, ran around the back of a house and stood there and did nothing more.  It comes apart if you start poking at it, but for the most part it meets their needs as the simulation is a backdrop, there's no real way of need to interact with them, they did a bit of it with GTA:SA, I'd like to see them doing more.

The world sim is a complex beast but it's one the player needs to feel they understand, maybe not fully but enough to know what the can really affect.  We have a lot of complex systems under the hood in Fable but we left the player in the dark, for example the economy.  Our ecomony is very complex, everything feeds into it, does the player understand it, no. Should the player understand it, yes.  That was a big failing on out part, plus the balance numbers got tweake. So every player was Scrooge McDuck.  Our other big failure was the hero was a zone of disturbance, npcs would stop whatever they were doing anytime the hero got anywhere near them (good use of 2 years of my life) to comment on the hero's socks or something. So players lost the effect of this living world we were trying to create and got hounded by npcs instead.

One thing I think we got right, and more games need to do likewise, is our crime system - only npcs who witnessed the crime or find you standing next to a body know about the crime (we have unsolved crimes which kicks off special dialogue), if you kill the only witness then knowledge of that crime dies with them.  Once the crime is reported all guards know about it though, it was just too confusing otherwise.

Sorry this has turned into a bit of a post-mortem, but if other people can learn from our mistakes the that's great.

I spend ages in other people's games trying to find the boundaries of their world sim, what will it let me do, what won't it.  RDR's sim for the most part was like GTA's, the little scripted vingettes helped but on the whole I was disappoint that my only real interaction beyond a tip of the hat was violence, with some odd limitations.  I can shoot someone dead, I can hogtie someone and drag them behind my horse, leave them on the railway tracks or leave them out for the cougars but I can't throw them in the water and drown them...

I'm rambling so I'll stop, I hope you don't mind me sharing some of that experience from my non-indie development.  World sim is something that's going to get bigger, there are still only a few places really doing anything most RPGs are content with their inhabitants stand in a single spot but I think that's going to change, we build worlds not sets.

[sodap]

now that was a great post

[agersant]

Very interesting indeed, especially since I can relate most of your words to my own experience playing Fable II. It's nice to see that you gave it a lot of thinking since then as I have high expectations for #3.

I hope you don't mind me sharing some of that experience from my non-indie development

I think I'll be speaking for everyone if I answer PLZ MOAR or something along those lines.