The problem with RPGs

[weemadando]

I have often advocated for an RPG to be set in a very small environment, maybe a tavern that's full of guests stuck in a storm or at largest, a small village.

Focus all the efforts onto developing depth in all your characters and trying to make everyone as interwoven as possible in terms of making sure that decisions all have consequences, minor or major.

But then people would be all like: "It's not an RPG because it doesn't have a map screen."

[Stark]

'Isn't an RPG because I can't put more points into sword chops to defeat boss'.

I think its fair to say that one of the major weaknesses of Beth games is their overambitious scope.

[Starglider]

I have often advocated for an RPG to be set in a very small environment, maybe a tavern that's full of guests stuck in a storm or at largest, a small village.

Focus all the efforts onto developing depth in all your characters and trying to make everyone as interwoven as possible in terms of making sure that decisions all have consequences, minor or major.

But then people would be all like: "It's not an RPG because it doesn't have a map screen."

That works well for tabletop games (assuming decent GM and players). On a computer it just makes a 'interactive fiction' game or if the budget is higher 'interactive movie'. The fundamental technological limitation of CRPGs - restriction to canned plot and dialog - is masked in titles with lots of exploration and combat, because only a small fraction of your time is spent noticing plot/dialog restrictions. If you spent the whole game on that you'd get, at best, L.A. Noire without the shooting or driving. Quite a few games like this were made, particularly in the late 90s, but they were dropped because we found out that most players don't consider stringing together some movie clips via dialog choices to be much fun.

There have been intermittent efforts to do better than this, from Lure of the Temptress to Facade (with some failed attempts such as the original vision of Oblivion's 'Radiant AI'). Generally though mainstream publishers and game designers shut down any attempt to procedurally generate dialog or story, both because of the technical risk of trying to innovate and because they are locked into the egocentric Hollywood director 'I must fully dictate the player experience for it to be meaningful' model.

[weemadando]

I have often advocated for an RPG to be set in a very small environment, maybe a tavern that's full of guests stuck in a storm or at largest, a small village.

Focus all the efforts onto developing depth in all your characters and trying to make everyone as interwoven as possible in terms of making sure that decisions all have consequences, minor or major.

But then people would be all like: "It's not an RPG because it doesn't have a map screen."

That works well for tabletop games (assuming decent GM and players). On a computer it just makes a 'interactive fiction' game or if the budget is higher 'interactive movie'. The fundamental technological limitation of CRPGs - restriction to canned plot and dialog - is masked in titles with lots of exploration and combat, because only a small fraction of your time is spent noticing plot/dialog restrictions. If you spent the whole game on that you'd get, at best, L.A. Noire without the shooting or driving. Quite a few games like this were made, particularly in the late 90s, but they were dropped because we found out that most players don't consider stringing together some movie clips via dialog choices to be much fun.

There have been intermittent efforts to do better than this, from Lure of the Temptress to Facade (with some failed attempts such as the original vision of Oblivion's 'Radiant AI'). Generally though mainstream publishers and game designers shut down any attempt to procedurally generate dialog or story, both because of the technical risk of trying to innovate and because they are locked into the egocentric Hollywood director 'I must fully dictate the player experience for it to be meaningful' model.

The reason that it "doesn't work" in CRPGs is because people keep on with the same bullshit flawed design theory. It would take a radically different take on the genre, which is why building shitty Beth dialogue or dumbshit Bioware binary choices in wouldn't work. And the reason that I say keep it small scale is so that the time CAN be spent on doing this stuff right.

As for the "noticing dialogue restrictions" BS. What if this was a RPG where the plot played out in under an hour - say 45 minutes like an episode of TV. But where you still had quantities of dialogue like Mass Effect 2 (40k lines or something) so as to have such an amazing variety of dialogue that can happen in that time. And with levels of detail in the plot (not just "Goody two shoes" and "evil cunt") that let multiple play throughs be significantly different.

[Starglider]

What if this was a RPG where the plot played out in under an hour - say 45 minutes like an episode of TV. But where you still had quantities of dialogue like Mass Effect 2 (40k lines or something) so as to have such an amazing variety of dialogue that can happen in that time.

The basic reason that devs don't like to do much branching is that you have to pay to make content that not all players will see. If you branch massively, then the vast majority of content won't be seen on a playthrough. That drives up the budget to something ridiculous. No one is going to pay Mass Effect prices for a 45 minute game. You could just do it as text of course, which is the Interactive Fiction niche market of mostly free games. That's a slightly more advanced version of a 'choose your own adventure' book, but massive branching isn't popular even there. Writers don't like it because most of their work will be unappreciated by the average player, it's too hard for them to keep all the different plots and character attitudes in mind when writing it, and it turns out most players would rather have 8 hours of quality but fairly linear experience than a confused mess for 45 minutes with lots of choices.

And with levels of detail in the plot (not just "Goody two shoes" and "evil cunt") that let multiple play throughs be significantly different.

'Replay value' in the game industry is one small checkbox feature somewhere below graphics, audio, this year's trendy new mechanic etc. Only a small subset of gamers replay games and they seem mostly happy with one alternate path ('dark side' vs 'light side'). The industry mantra is that for longetivity you simply add online. Single-player experiences have been steadily pared down in many games anyway, and now you are talking about making massively more single player content that most players will never see?

No one is going to risk multi-million dollar budgets on that. Procedural is the only way this particular wish will be fulfilled.

[Stark]

You know skyrim just came out and is stupidly popular and profitable, right? Huge amounts of crafted content is still made; there's just the understanding that it'll be of low quality, broken etc, like the last 15 years of broken quest scripting. An effective procedural approach would help.

[Starglider]

You know skyrim just came out and is stupidly popular and profitable, right?

Sure but isn't the assumption there that most players will see most of the content? I haven't played it, but if it follows the Fallout 3 pattern then 80% of the assets are common across all plots/morality choices and a typical playthrough will experience more than half of the content. Filling out the world with a lot of second-tier content (side quests) for players who put a lot of time in is standard practice, but that's nothing like Weemadando's suggestion of having deeply branched main plots and only 10% of the content experienced on a typical playthrough.

[Stark]

Yeah, I doubt it takes any bold steps toward plot-unique content. Much of the content is literally cookie-cutter shit anyway and they still manage to fuck up the scripting anyway.

If the 'content' is procedural (even if its not quests or map area, just things the player gets rewarded for doing) you could get away with extremely variable experiences. There's not much market for non-traditional game designs, though.

[Starglider]

If the 'content' is procedural (even if its not quests or map area, just things the player gets rewarded for doing) you could get away with extremely variable experiences.

Yeah, that gives arts grad game designers butt-clench. When programmers had primary responsibility for game design there was a lot more willingness to solve problems with algorithms instead of just aping movies as much as possible. Done properly 'variable experiences' improves average player satisfaction, because different players want different things out of a game. Open-world already offers some of that, but in a passive way. Left 4 Dead's director AI does active optimisation, but only in a narrow context. The long-term dream is a suite of feedback systems that will actively optimise the whole player experience, even in open-world games.

To be fair though there's a big distinction between technology that's just a lot effort to implement and tech which doesn't actually exist yet. I complained about Id being lazy hacks with RAGE because all the technology has been developed (i.e. there are thousands of academic papers and technology demos) to do good procedural textures, models and levels, they just couldn't be bothered to implement it in their engine. Decent procedural dialog and plots are pretty blue-sky and I can't blame publishers for not risking AAA budgets on those ideas. Still, companies like EA should sink a lot more money into R&D of that stuff, compared to the usual graphics/physics engine stuff there's much more potential payoff in advancing games and keeping them fresh, interesting and competitive against other entertainment.

There's not much market for non-traditional game designs, though.

What's 'non-traditional'? As I said there's not much market for interactive fiction, it's a tiny niche. L.A. Noire is supposedly the most advanced 'RPG-ish' dialog system ever seen in an AAA title, is it mainstream? Rockstar did push the envelope on that one, but mostly with the motion capture aka 'we want our game to be even more like a movie than usual'. Open-world RPG with procedural dialog at least as good as Facade and procedural plots akin to a modernised scaled-up TALESPIN? There's no 'market' as such because no one has ever made one, but I'm pretty sure there's plenty of demand for one.

[Stark]

I mean that if you wanted to have a player experience that wasn't 'first/third person dude chops mans', you don't have much of a market. You could go Japan and make a totally dynamic bartender conversation simulator (or whatever) which would probably be easier than a conventional adventure game, but what's the point? I don't think you can move too far away from the 'level up, buy sword, kill boss' thing if you want to actually make money.

LA Noire's dialog system is entirely normal. The only thing special about it is the NPC animations, which are recorded from actors rather than being generated in response to player input.

I'm not sure why its so hard to make crafted narrative complexity anyway; surely software tools would make it trivial to keep track of and conceptualise. I get the impression most developers are doing it manually, by talking about and trying to remember what is what and who has the crown and shit.

[Starglider]

I mean that if you wanted to have a player experience that wasn't 'first/third person dude chops mans', you don't have much of a market.

The games industry has pioneered a lot of new genres over the last two decades. It's true that there has been more refinement and less brand-new stuff recently as we've picked the low-hanging fruit, but I think there's plenty of scope still to do new stuff. I mean, if we literally had the holodecks from Star Trek, are you saying everyone would just shoot/hack/slash all the time? Some people would do that, but I think TNG actually got it right there in that most adults would prefer story and character heavy experiences. If you accept that then the question is what are the incremental steps to get there and how do we make those incremental steps profitable. I'm not saying there are easy answers to this question, quite the opposite in fact, but I think the games industry has plenty of technical people willing and able to tackle it head on... if we could break the 'plot/dialog/cutscenes must be like a movie' thinking.

I'm not sure why its so hard to make crafted narrative complexity anyway; surely software tools would make it trivial to keep track of and conceptualise. I get the impression most developers are doing it manually, by talking about and trying to remember what is what and who has the crown and shit.

What kind of tool do you mean other than the dialog tree editor itself? You can annotate dialog and scripting nodes as much as you like with notes about what the characters are feeling and motivated by in this case, but it's always going to be harder jumping between lots of different versions of a story rather than thinking about one or two. TALESPIN was what we actually need; a procedural description of motivations, abilities, objects, interactions, dialog templates etc that lets the system generate something internally consistent on the fly. If you can't get it working reliably enough to ship in the game you could at least use it offline to automate the process. As I said before though, (in my experience) writers hate to give up control, both of the player experience to the software and of the narrative development process to programmers.

[Stark]

Developers (or funders) see the success of the 'xyz genre + thrilling story told in cutscenes or through quest scripting' concept, so its not surprising its so common. But the mechanics of 'RPGs' are so ingrained that they're even used to define the genre. If Skyrim was a huge world where you did something that wasn't walk around and chop, I seriously doubt it would be anywhere near as successful. Not saying that being a sheepherder or running an ecology would be any good.

At very least you'd think dialog trees would be able to autoscum to find glaring holes in scripting or contradictions in plot state, and yet we see that sort of thing in AAA titles. A part of the problem with multiple narrative paths or threads is I don't think most writers are really mentally equipped for it - they're writers, not player response creation coordinators, y'know? They want to tell a story, not set up a series of objects and motivations and let stories emerge or seriously give equal value to several branches. Look at the quotes back a page or two from Bioware writers; they're literally squealing fanboys themselves.

I remember you talked in another thread about the change from programmers as creators to movie-style creative staffs, and I think bringing that attitude to games really reduces the desire in creators to do anything like what is discussed here. And seriously, linear (or quasi-linear) games sell really well, so the pressure to change isn't going to come from above.

[Starglider]

If Skyrim was a huge world where you did something that wasn't walk around and chop, I seriously doubt it would be anywhere near as successful.

The most successful computer game ever is Farmville, with a complete absence of hack/slash. 'Casual gaming' is huge and quite different from Skyrim. The big trendy thing in the annoying game-fan blog space at the moment is 'building a game layer over real life', i.e. apparently people enjoy all kinds of trivial crap more if you put game mechanics on it. Some MMOs have suceeded with no combat at all (e.g. A Tale in the Desert). On the other hand, you've got both strong console-RPG expectations to deal with and the fact that we have lots of experience and technical capability to make satisfying hack/slash. Most other activities you'd either be developing from scratch, or would just be technologically impossible. Saints Row 2/3 for example pushed the boundaries of side-activity variety, but it's still obvious that they added things they could easily script into the engine with little new tech development.

So yeah, I would agree that right now it would be hard to make a successful main-stream game set in Tamriel with little/no combat, but I think that's a lack of adequate technology and design experience, I'm not clear if you think that or if you think people fundamentally just want to hack/slash. I think we're already seeing some incremental steps away from combat-focus in say GTA4 and SR3, where there are plenty of non-combat side activities that the average player does sink significant time in. But technical limitations mean those are background activities not even at the level of side quests, much less the main plot.

At very least you'd think dialog trees would be able to autoscum to find glaring holes in scripting or contradictions in plot state, and yet we see that sort of thing in AAA titles.

Unfortuantely the trend is to make dumbed down kiddie scripting systems that non-programmers can learn quickly and not cut themselves with, that don't even have the level of formal verification that real programming languages have, never mind something that will do global analysis (those kind of formal methods techniques are currently expensive, difficult and only really used for safety-critical systems). Publishers would rather spend money on hiring 100 untrained uncredited Vietnamese QA staff to play the game 10,000 times, rather than develop radical new tech. They see the former as lower risk.

A part of the problem with multiple narrative paths or threads is I don't think most writers are really mentally equipped for it - they're writers, not player response creation coordinators, y'know? They want to tell a story, not set up a series of objects and motivations and let stories emerge or seriously give equal value to several branches.

Yes, the later discipline doesn't exist yet. It's something of a chicken and egg problem with the technology and the experts. We were kind of overcoming that with AI rule systems in the 1980s - people were seriously talking about 'knowledge engineer' as a dedicated profession - but then the AI Winter hit and it all turned out to be horribly premature and was quietly forgotten. If I was trying to monetise my AI work in the games industry, rather than an industry that actually pays for technical excellence, then I'd probably be working on this stuff.

And seriously, linear (or quasi-linear) games sell really well, so the pressure to change isn't going to come from above.

Are you saying it's up to Indie Gaming to Save The Day? Looks like we're doomed to endless BioPlot rehashes and Call of Warfare Honor Modern Duty Medal XVII then

Seriously while a lot of indie gaming is (a) slavish immitation of mainstream (b) pointless obsessive retro-ism (c) minimum effort casual cash grabs, I would eventually expect Facade style tech demos to morph into something genuinely compelling and disruptive. Looking forward to it in fact.

[D.Turtle]

That's actually not at all analogous to what she said. She said she hates playing games, not that she hates games or writing for games.

What would work better as an analogy is a screenwriter or author saying that they hated watching films or reading novels. The latter would probably be more egregious but I can easily see a screenwriter go 'You know, I actually don't like watching movies. Writing them is fine but actually sitting in a dark room for a couple of hours doesn't strike me as enjoyable.'

Or a better example is how many actors are asked what they think of their performance, and they say something like 'actually I don't like watching myself on the screen so I can't really answer that question.'

Basically you're turning a statement and concluding the wrong thing from it. Writing for games does not equal to actually playing them, and I can see how someone who is writing for a game may not even be interested in the gameplay mechanics of it at all. Hell look at the specific complaints she makes about playing games:

I'm really terrible at so many things which most games use incessantly -- I have awful hand-eye coordination, I don't like tactics, I don't like fighting, I don't like keeping track of inventory, and I can't read a game map to save my life.

A lot of that has nothing to do with writing a plot and having great characters and meaningful motivations and themes and ideas and so on. It's 'hey I hate playing inventory tetris' and 'I don't know a thing about tactics beyond selecting all my dudes to hit on that one bad guy'.

I would somewhat agree with your objection, if they hadn't completely missed their goals. The writers are responsible for the world, what happens in it, the motivations of the people, etc. They may not have to like all the actual gameplay aspects of a particular game, but they do need to understand what makes a game of a specific genre good and interesting. If the take-away from RPGs is: "They're all about looting and killing, and I don't like that", then it is no wonder that they totally fail at writing for games. If you have a head writer whose reaction to "I killed this character in the previous game, why is she alive" is: "I'm the writer and I tell you what happens and what you think you may have done in a previous game is irrelevant" (paraphrased reaction to the fact that Leliana is in Dragon Age 2 even if you killed her in DA:O), is it any wonder they totally fail at providing any immersion into their game world, fail at providing any convincing reason for the actions you (or others) in the game world take? Dragon Age 2 is an excellent example of everything that is wrong with Bioware and cRPGs. And the writing is the worst part of it.

You can't blame the writer for stuff like Bioware's joke evil or retcons or whatever anyway, because they don't decide that sort of thing. They just write dialog or quests, and 'playing a lot of games' doesn't somehow make you better at either of those things.

And frankly linear games (like Halflife, or Metro or whatever) are so much easier to write for it isn't funny. Its only when you decide to give the player choice that you need to take a sophisticated and sensitive approach to acheiving that.

You can blame the head writer for that sort of thing, because he decides that kind of thing. You can blame the individual writers extremely badly written quests, dialog. Dragon Age 2 is a mostly linear game, because almost every single choice you take is rendered irrelevant, except for one or two different sentences here or there. If a writer views the entire gameplay as irrelevant or actual a hindrance for the actual story they want to tell, is it any wonder that they completely fail at providing the motivation for that game play? Its akin to a writer writing a Star Trek book, but hating all that Tech stuff, so he just writes a fantasy novel and at the very end says: "Oh it took place on the holodeck." The thing is that, yes, the vast majority of time in RPGs is spent running around, killing enemies, looting stuff, etc. But the writers' job is to provide the rationale for that. If they instead hate RPGs and all the stuff involved in them and actually just want to write a story (but can't cause they suck at it), then it shouldn't be a surprise if they completely fail at doing that, fail at building a convincing world, fail at writing convincing, consistent characters, etc. I mean, the main quest of Act 1 in Dragon Age 2 is "Gather 50 gold". And the reason for doing that, is that you want to go on an expedition into the Dark Roads (after fleeing the continent from a blight) in order to get more money. How is that anything else than bad writing? How is that anything else than the writers' fault?

I'm not saying that it's completely the writers' fault that a lot of RPGs suck, or are formulaic - but it is a large part of the problem.

[Stark]

It's not DA2 writers fault that it's a linear game with few branches; that's Bioware's MO. It's institutional. Some writer being an idiot fangirl doesn't make an RPG better or worse.

If you think some quest writer decided the basic structure of DA2, I'd like to know why. I'm not even going to get into the lols of what you don't like about it.

Equating FarmVille toa procedural narrative is a lol. Alarmingly, iOS capitalises FarmVille. THAT'S fucked.

[D.Turtle]

What I'm saying is that DA2 being a linear game with few branches is not what makes it a bad game/RPG. I'm saying that most of the writing of DA2 makes it a bad game.

This can be easily seen by the fact that there are tons of linear games with few (or no) branches that are very good games.

[weemadando]

There was a great interview with Guillermo Del Toro on Irrational Interviews where he made a comment along the lines of:

"When I'm playing a game, I'll always skip past cutscenes, because I'm playing the damn game and if you can't tell me the story in the game then I don't care."

[Starglider]

"When I'm playing a game, I'll always skip past cutscenes, because I'm playing the damn game and if you can't tell me the story in the game then I don't care."

What do you think of the in-game cutscenes that (for example) Half Life 2 has? Sections like Kliener's lab and the missile launch prep still have NPCs going through scripted dialog and animations. All you can do is change your viewpoint. It's more immersive, but still completely scripted and you don't even have the option to skip. But if you rule that out, what's left? Constant voice overs from characters on the radio / phone / sitting next to you in the car? Audio logs? Terminals? 'Environmental storytelling' aka suggestive level design?

[Thanas]

I actually love cutscenes when they are done right.

[Bakustra]

I think what should be taken from del Toro's comment is not "Oh, yeah, smart guy? Then what will you accept, huh? Do you wanna go back to DooM?", but rather, "Cutscenes should only be cutscenes if there's no better way to get your point across." Because, after all, there are plenty of ways to tell a story subtly and without interrupting play, and I am of the opinion that you should go for economy in play-interruption.

[Thanas]

This doesn't work for games that are known for their HQ cut scenes like Starcraft for example.

[Bakustra]

This doesn't work for games that are known for their HQ cut scenes like Starcraft for example.

Let's take a look at the original Starcraft cutscenes, though. I can't think of any way they could have gotten their meaning across better, what with the nature of the game rendering it impossible to effectively communicate the majority of them. Do you disagree?

[Thanas]

No, not really - but they still have a lot of cutscenes in the sequel as well. Maybe it is just typical for RTS games though.

[White Haven]

I've always found the anti-cutscene crowd...odd. Honestly? I love a good, well-done, dramatic cutscene. By the time in-game engines can match up to what current pre-rendered cutscenes can do, the state of pre-rendered cutscenes will be even better, and that's always been the progression. I don't even comprehend why people think that's a bad thing, it just doesn't compute.

[D.Turtle]

Yeah, I'd find a strict anti-cutscene stance weird too. I think there are tons of good possibilities of world-building, plot advancing, information dumps, etc. Its all a question of execution, though I think that cutscenes can become an easy crutch to use for world-buildingstory-telling, etc instead of building that into the gameplay itself.

One game that has a very cool way of doing that is Bastion (video) - basically the Narrator gives a running commentary of the game, providing the backstory, while you are playing.

Obviously you can't do that for many games, but I think it is an excellent example of integrating story-telling into the gameplay itself.