94: Next-Generation Storytelling

"Not long ago, I had a conversation with Doug Church, secret master of gaming, where he said something like this: A story is constructed of sentences, strung together in a coherent, dramatically significant order. Game "sentences" are the actions available to and selected by a player. The more sentences we allow players to construct (in other words, the deeper the pool of options we offer), the cooler and more numerous the story possibilities will be. To that extent, a robust world and character simulation - both made possible by next-gen hardware - will allow us to tell a better story. But there's a hitch: all the graphics power of the new platforms."

Next-Generation Storytelling

I really liked Warren's articles on Next-Generation Storytelling.

To do justice to the game medium, what we're really talking about is reinterpreting what we mean by the word "story". Many times, when I hear about people talking about storytelling through games, I get the image of them trying to ape the accomplishments of books and movies. I don't think that's the best use of the game medium because, at least in theory, games are capable of so much more through the player's freewill.

Freewill can ruin a good story. How many of us, through the freewill that we express in our lives, end up with a good story? The good stories come afterwards through a retelling that edits out the drudgery and, perhaps, adds sparkle with a few judicious lies. (How many good story tellers are not liars?) The footage on the cutting room floor far outweighs the final edit. Games don't have the luxury of that much editing. That's why I think that it's important not to get too caught up in the storytelling paradigm. It's too easy to reduce games down to movies or books. It can be done, but I don't think that it's the best use of the game medium.

As much as I like what Warren has to say, I don't think that the word "story" is the best word to capture the creative vision needed to surpass the limits of books and movies. I prefer to think in terms of alternate world modeling, but maybe that's a byproduct of the type of playing and exploration I like to do.

...and, to be fair, Warren discussed world creation in an earlier article.

My self-appointed role is to promote text-to-speech as a technology useful for games, something I didn't see mentioned in the technology discussion. Current TTS quality is analagous to the quality of 3D graphics in 1980 (aka: Battlezone), but over time it will become more and more important for games. Today's reliance on recorded speech is similar to the early 1990's phase of interactive movies like Phantasmagoria, where character images were prerecorded as video.

I feel that TTS is critical for games with conversational NPCs. For more ramblings, see http://www.mxac.com.au/drt/NPCConversationWall.htm .

PS - I am creating a game that relies on TTS, and I'm certain that it won't be a mass-market success, partly because TTS isn't as good as professionally recorded speech... but TTS is infinitely cheaper and more flexible. The game is FAR from done, but you can try it out and get a feel for the uses and limits of TTS by downloading it from http://www.CircumReality.com .

Interestingly enough, Mike, my research relates to voice. Though it isn't directly on TTS, my algorithms could be used to increase the emotive expressiveness of TTS by modifying the perceived effort and breathiness in the voice.

I've done a little bit with breathiness. My algorithms store a voice as F0, a demiphone-resolution histogram of voiced energy, and another histogram for unvoiced. When the signal is generated, I use additive sine-wave synthesis. When there's unvoiced energy, I add in two other sine waves per harmonic and allow their frequencies and energies to vary randomly as a function of noise vs. voiced. Probably not up to what you're trying to do, but it works okay and doesn't have as much of a noise on/off effect as LPC. However, breathiness control is the least of my problems. :-)

At its best, current TTS sounds like a bored telephone operator. (Which is better than 10 years ago when it sounded like a "drunken swede".) I've read some papers where companies are trying to get "angry" and "happy" voices, but subtle emotion is a long way off. To some extent, transplanted prosody (see my web site for an explanation) solves this problem, but then you're back to pre-recording emotion-laden sentences.

Of course, breathiness is important for emotions, but seeing as my system often fails to produce a good F0, duration, and energy for a "bored operator voice", let alone "angry", "peaved", or "sexy", breathiness-control is low on my own priority list. (Although essential for singing TTS.)

My algorithms also manipulate the perceived vocal effort, which would relate to an angry sounding voice. I'm working with LPC but the concepts could be translated into the domain of sine-wave synthesis.

However, breathiness control is the least of my problems. :-)

I understand that making the voice sound breathy, angry, sad, or happy comes secondary to getting it to sound moderately natural in the first place. TTS has a long way to go. I just thought that it was interesting.

Best of luck with CircumReality.

Wow, I don't know what either of you (Mike and Nord) are talking about. But, hey, sounds really cool and keep up the good work ;-). But seriously, what I do understand of it sounds neat.

And as far as the article goes, it brought a number of nice conclusions to the idea brought up in this piece (as well as the past 3 articles).

About a month ago I wrote about narrative as a dimension within gaming and compared it to the addition of the Z-axis. Playing Prince of Persia: The Sands of Time (I buy most of my games used, so I'm sometimes behind the times), and hearing the voice over during the actual game got me thinking about how cutscenes are a terrible crutch for story telling in games. There article is here: http://www.prometheusperspective.com/features/march07/feature03240701p1.php though the site isn't really ready for the public just yet. (I try not to link externally on message boards, so I don't really know the etiquette. If its not proper for me to link to the site I'll remove it no problem)

I think it's really important that, as you say, developers start thinking more about Moby Dick and less about Star Wars when baking up stories. If games are going to be taken seriously then they need to be using serious sources for inspiration. There is nothing wrong with mindless action games, but if the whole lot of available games are mindless action, well, its hard to defend the medium. Gamers pounce on Ebert for saying that games aren't art, but the point he makes is decent: there just aren't enough examples that show why games should be considered art.

I like the direction that you're traveling. I think we're starting to outline the specific differences between games and other means of conveying a story, and I think there will be more clarity about the difference between written and digital literary techniques.

My algorithms also manipulate the perceived vocal effort, which would relate to an angry sounding voice.

That's more interesting. I just sent you some E-mail since I don't think people here would be too interested.

I enjoyed the article, especially since it brought out, in my mind, the reasons for my tastes in games. For the last couple of years I have migrated to online games as my staple. Its rare for me to play a single player game let alone play it till the conclusion. Most often I will buy the game and play the single player until I'm proficient enough to jump online. After reading Warrens article, though, I find myself wondering if its the barren story lines and daft npc behaviour that have caused me to do this.
Granted I'm mostly referring to FPS's and Strategic games such as counterstrike or command and conquer. Many of these games actually ship with absolutley no story. The only information you get is usally printed on a small card in the box or from a short movie prior to your playing. But these games also need the sort of freedom that Mr. Spectre spoke of.
I realise that Counterstrike is not based on story but on your ability to kill the other team. Period. You options include chosing a variety of weapons and where you want to go in a map. Thats pretty much it. The map never changes, the bomb has to be in the same place every time, and the hostages simply stand in a room like statues.
Wouldn't it be nice to have an environment that changes? Wouldn't it be nice if the hostages tried to escape? If you could set a house on fire to make the terrorists run out? To run into the 'market place' and worry about shooting at the terrorists because innocent woman and children are running past you frantically looking for cover. There is a morality and emotional depth in nearly all of those decisions without a single line of dialoge.
What happens if you're playing a terrorist and you shoot a random bystander. Do you feel bad? Probably not if, as it stands now, the bystander simply yells an obnoxious 'ow' but continues to stand in the same spot.
In the world of the FPS there are few games really pushing the envelope. The battlefield franchise alters their games in small ways without ever really pulling the trigger and trying something wild. In fact, I think the biggest progression from these games has come from the community. It took the 'indie' designer to create the team based counterstrike from Half Life. It took an indie devloper to make Desert Combat, whose popularity shrouded bf 1942. As far as these modders go, though, there are very few instances I find of an improved AI or storyline.
Ok, so I'm clearly diverging from the purpose of Warrens article so I'll stop there. I just feel that even when a developer chooses to negate story altogether they still need to remember that the player should be involved in the world even when there isn't a history behind it.

Warren said:

"...the thing I'd most like to see is more developers making games that offer players freedom to explore story spaces within constraints imposed by a dramatist. We'd let players off the rails a bit more. We wouldn't settle for offering tactical choices, challenging puzzles and movie-inspired cut scenes.

Instead, or in addition, we'd offer players opportunities to explore more freely and to delve deeper into the inner lives of their characters in ways that don't involve killing them. In other words, we'd offer players real choices, with real story and character consequences.

Great story games are only partly dependent on technology. They're hugely dependent on will and creativity - on the need to engage in dialogue with culture, problems and players. Games can be about something more than killing, fighting and puzzle solving."

We developed an Interactive Drama that does exactly that, it's called Masq and you can find it at www.alteraction.com (free). The challenge has been to promote something that does not conform to the pre-conceptions of the market, the publishers or the investors. Fortunately Masq was recently re-discovered by PCGamer UK and PCGameplay Benelux who wrote very positive reviews.

In addition working indie to achieve a better interactive story may force you into making decisions such as not allocating enough resources to the "graphics" so you can spend more time intertwining story and gameplay to create a better experience. The result is that Masq resembles an interactive comic or an old graphic adventure when in reality is very different. But try convincing the market or investors to look behind the make-up, mainly when all others are delivering great 3D graphics.

Mr. Specter says he thinks AI won't be able to keep up in gameworlds with the latest graphical technology, because it will be too much work for a developer to create this AI. I believe AI will be able to keep up, because of the recent trend of using middleware. When Specter wrote this in 2007 Havok was probably the most used middleware, but with systems like Euphoria (in GTA IV, Backbreaker, SW: The Force Unleashed), animations can be generated procedurally, creating a more believable world. I think there will be more companies that will focus on just AI/physical behaviour and redefine the way we perceive gameworlds.
This, combined with the succes of games like Heavy Rain and Dragon Age, make me believe the future will be great for non-lineair games.


Reply to Thread

Posting on this forum is disabled.