It’s come to that exciting (and terrifying) moment in the development of “Austen Translation” when we get to give the ending sequence a little love. Endings — in games and in literature — are tricky beasts since they need to check a lot of boxes. They need to feel earned, organic, satisfying to the player or reader, and a bit unexpected. And it’s in the balance between surprise and inevitability that the art of crafting a “good” ending lies. Too much surprise and it feels like it’s a conclusion out of left field. Too much inevitability and players see it coming a mile away and your suspense is shot.
Complicating the “Austen” ending problem is our aim to keep the number-crunching that’s going on under the hood, well, under the hood. We made a conscious design choice at the start of the project to downplay the spreadsheet aspects in favor of a more literary, narrative gameplay flow. But without showing players some obvious objective standard the game is using to judge winners and losers, we ran the risk of making the strategy elements feel like window dressing.
After a few iterations, we’ve decided to experiment with a hybrid approach which enthusiastically embraces both story and statistic, just not at the same time. At the end of each round, the player gets to see the “story” ending that happens to her heroine. This gives their story journey a sense of completion while maintaining a certain “literary” distance from the math that determined their result. By building out a pool of stock unsuccessful endings to draw from, we can offer players endings which seem fresh and integral to their stories each time they play.
Once the player’s individual story has come to a conclusion — either in glorious victory or punishing defeat — the round ends with a statistical table laying out how each of the players did and what each of the bachelors was looking for. An objective document of the round.
By placing it at the very end, outside of the narrative flow, we hope to provide players with feedback on their round without breaking their sense of a continuous narrative. Then, armed with some hard, cold facts about why they won, or didn’t, they can plunge back in to tackle new Austenian challenges and woo new Austenian bachelors.
We're excited to announce that "Austen Translation" just went live on Steam Greenlight (like, just minutes ago). To celebrate, we're posting the promotional trailer below. If you've enjoyed reading about our game design process, please head over to our greenlight page to give us a thumbs up of support. Thanks! http://steamcommunity.com/sharedfiles/filedetails/?id=927724334
We were at Pax East last month and I watched a young woman playing a lovely little story-based game. I couldn’t help but notice a phenomenon that I have seen frequently in the game world. She clicked through every bit of text. At the fastest possible pace and without appearing to make any effort at all to read anything. It was like the text was being perceived as getting in the way of her fun. And my writer’s heart bled a little with every click. Because someone somewhere (hopefully) had labored to put together copy to illuminate, enrich and expand her digital experience. And maybe it was successful text, and maybe it wasn’t, but that player would never know. That player is probably clicking away from this post even as you read this — too wordy! Get to the point!
The point is that “Austen Translation” is all about capturing the essence of a particular sort of literary experience, so text (to one degree or another) is not really optional. As a writer I am willing to die on this particular game-design hill. The question then becomes, in a literary strategy game, how do you make it advantageous to the player to actually read the modest copy that appears in the game? Ideally without punishing the fast-clickers (because this is a game and punishment is not fun). Can we layer on a mechanic that encourages players to engage with the story text? The answer may lie in the brutal pressure of the conventions of polite society.
Austen and the women she wrote about were often educated, artistic, witty, charming and excellent conversationalists, but not because these accomplishments were valued for their own merits. No one thought these women would go on to use their knowledge of philosophy to teach at Oxford. They were educated so that they could hold their own in a dinner party conversation. The aristocracy needed intellectual discourse and art and performance to connect people and demonstrate value. Perhaps this thematically appropriate social engine can be put to work in the service of meaningful text?
We’re currently looking at ways to weave details about the bachelors’ likes and dislikes — clues, essentially — into the game text. In that way, an attentive reader (our version of a good social listener) could see some advantage in making choices that show she is being attentive to her preferred bachelor. Will he like her more if she also claims to favor the breed of hunting dog he prefers? Will his heart warm to her just a little bit if she remembers his sister’s name? This approach also interests us because it allows the player to actually do the thing their avatar is doing in the game (listening and remembering and signaling that) which strengthens immersion and connection to the game world and goals. Stay tuned as we explore methods for implementing this idea without breaking the game...
At its most basic level, "Austen Translation" is a game about making choices — choices which affect how the other characters in the game perceive you. So, if you were guest at a swanky lawn party, you might choose to “bedazzle your croquet mallet” to get a +1 to your BEAUTY score. Or maybe “fail to warn the other spectators of an incoming polo ball” to boost your RUTHLESSNESS score. It's pretty easy to see the cause and effect in these simple examples.
But recently we’ve been experimenting with some more complex interactions between a player's character and the other women vying to marry the same well-off bachelors. These might include actions where you gain *and* an opponent loses. Or maybe you both lose but she loses more. In this way, we’re hoping to make the choices of the rivals — or other players in the multi-player version — more meaningful, since they might well have a direct impact on you. We also wanted to enhance the sense of direct competition between the women. For instance, at the archery club outing you might choose “You ‘mistake’ an inattentive rival for the target” which would give you a boost to RUTHLESSNESS and the rival of your choice a reduction in WIT.
What was particularly interesting to me was the conversation that followed when some less obvious relationships between the text and the resulting changes to the stats cropped up. Take the action option “Offer a rival the last slice of pie at the church picnic." The original slated result was -1 WIT for you and -1 RUTHLESSNESS for your rival, the reasoning being that you have be pretty dumb to give away the last slice of delicious pie, and that your rival would feel warmly toward you and more happy with the world in general if they just had more delicious pie (my partisan “pro-pie-bias” is showing here). But, not so fast! There is also a good case to be made that this would be a -1 RUTHLESSNESS for you, and a -1 BEAUTY for your rival event, the reasoning being that you will look like a softy by giving away the pie and your rival may well look a bit of a glutton. A case could also be made that your BEAUTY might suffer (what with all the crying you’re doing having had to give up pie) and your opponent’s WIT might suffer (since they were too dumb to figure out just how much you love pie before they took it from you).
As a writer, these sorts of “self-constructing narratives” are a fascinating mechanic. How much content do we need to give players before they start connecting their own story dots? Do players feel a stronger sense of connection and commitment when they feel like storytelling partners with the developers? And how can we structure the narrative we do provide to encourage this sort of imaginative investment in the action?
We found a similar human delight in connecting implied narrative dots when we were working on our first game, "A Matter of Murder." In playthroughs and livestreams it was pretty common for players to draw strong conclusions about relationships, motives and events base on the *flimsiest* of cues. Most often these story elements weren’t intentional or integral to “winning” the game, but players seemed to revel in these serendipitous discoveries. It’s a tribute to all the ways our brains are wired to see pattern and meaning and story in the world all around us. It’s a beautiful thing.
Writing compelling copy for a strategy game like “Austen Translation” is a daunting enterprise under the best of circumstances, but it’s doubly challenging when you’re writing to capture the essence of a beloved literary canon. It was relatively easy, however, to pick up some of the broad themes from her work, such as the role of women and her criticisms of marriage as an institution, since many of them are baked into the satirical premise of the game.
The conceit of the game is that the player is “writing” their own Jane Austen novel, so it seemed natural to adopt the conventions of classical literature, down to nomenclature and structure. Each action round is presented as a chapter with it’s own headline and subhead, as well as a blurb of “flavor text” to set the scene for the next phase of play.
We looked at the formal aspects of Austen’s work such as sentence structure, vocabulary choices and cadence. It was way too much fun plant contentual easter eggs in the text -- names, word choices and ideas which fans of her work would undoubtedly recognize. This serves the double purpose of grounding our game in the Austen universe and offering players the thrill of discovering the connections as they play.
A hallmark of her work was the seamless transition from the traditional omniscient narrative to text which gives the reader a window into what her characters are thinking and feeling. She is credited with pioneering this objective > internal > objective again formula. And this gives us, as game designers, a terrific opportunity to use this same mechanism to give players an additional level of strategic information they can act on in the next chapter.
It’s been a busy couple of weeks getting a working protoype of “Austen Translation” on its feet and preparing our submission for the 2017 Boston Festival of Independent Games. We’re excited to get the game into the hands of playtesters for feedback so we can begin the real work of making the game fly. Stay tuned.
This week we're bringing you a look at the cast of pre-generated rival characters who will populate the Austen Translation world. Players will have the chance to name and customize their own avatars, all built out of the same kit of parts used to create these NPCs. And yes, they look sweet now, but just wait til you get between them and the bank account -- I mean eligible bachelor -- they've set their cap for...
After the paper playthrough of “Austen Translation” (see our “TMI” post from November) we started putting together a playable, electronic version of the game we’ve been referring to as the "digital prototype." Unlike the traditional "vertical slice” method which strives to show a small section of a game in near-finished form to give a sense of final product, our “digital prototype” is intended to allow an alpha tester to play through an entire game from Chapter One to the closing Flurry of Marriage Proposals to give a sense of game flow.
With the digital prototype, we’re looking to explore big-picture questions of pacing, story structure, information delivery, and player agency. We’re less interested in having the animations appear polished or the transitions to be perfect, since this is intended to be a very rough-cut of the final full game experience. Some of the features we're planning to include in the finished release are also not included in this prototype, like avatar customization and multi-player, since those features don’t really affect the shape of a given playthrough at a fundamental level. In the interest of reducing variables (and preserving programmer sanity) the digital prototype also uses only a static subset of four events to build the plot, where the finished game will randomly pull and assemble from a pool of twenty-four.
The digital prototyping stage is also a chance for us to work out the mathematics of how the game evaluates the relationships and how the bachelors and the AI rivals make their decisions. You can see some of that numerical underpinning in the screen capture above. In the spirit of making “Austen” feel like a literary adventure and not an exercise in spreadsheet management (I’m looking at you Eve Online), we’re working to make this mathematical “backstage action" as invisible to our players as possible.
Once the digital prototype is in playable form, we’ll be anxious to get it into the hands of bunch of alpha testers so we can start getting some feedback to inform the next round of prototype. It’s an exciting time for us at W&M! We’re so glad you’re along for the ride.
Sometimes it takes a look to the left to get a clearer idea of where you might want to be going. In this case, the left in question was to Tim Conklin’s “Antihero” table at the recent Google Play Indie Games Festival. We were there with “A Matter of Murder” and Tim was demo’ing his Dickens-inspired title which features wonderful art in a lovely, gritty, cartoony, gas-lit London. Of particular note for this story -- the characters all had wildly disproportionate head sizes in comparison to their bodies. And, while the extreme contrast of “Antihero” gives it a strong feel that’s not appropriate for Austen, it did inspire our Art Department to take a second look at the more “classical” proportions we were originally planning on -- and to do a little experimenting.
Note how serious and adult the character with the smaller, original head size appears to be. She’s prim. She’s the babysitter. She probably flosses daily. But the character with the 200% head size? She’ll be the one cutting tags off of mattresses while she eats three-day-old sushi off her boyfriend’s motorcycle helmet. We like her more. We connect with her.
Now Disney and anime have long exploited our fondness for characters with big heads and especially big eyes. Cognitive psychologists would point to humans’ hard-wired reactions to infants (with their large heads relative to their bodies) as the root of this almost visceral connection we tend to feel more for characters like the one on the right. We looked at larger (300% and 400%) but larger turned out to be a game of diminishing returns, and after 200% began reducing the clothes and hands to a scale where they were losing too much detail to carry story. So we’re thinking 200% may be about the sweet spot for “Austen."
As an added plus, the larger head/eye size also focuses player attention on facial expression, which may help solve the problem we talked about in an earlier blog entry about communicating reactions.
Tim may have inspired us to start thinking about character proportions (thanks, Tim!), but I blame evolution for making us love the big-headed freaks-of-nature my old life drawing professors would be appalled by.
We sat down with a group of W&M family members to playtest a “paper” version of “Austen Translation” last week. Paper play testing of a video game may sound weird, but it's a great way to try out game mechanics, make gameplay adjustments in real time, and to get immediate player feedback before we sink a lot of time into developmental dead ends. And boy, did we learn a lot in a single evening! A little wine, a few hardcopy printouts and a slapped-together spreadsheet to crunch the game calculations on the fly and we were off to the races — or, in the case of one of the scenarios we played out, off to the polo matches! Here’s four of the many things we learned from that session:
(1) TMI is Deadly
We were experimenting with ways to give players information about how they were doing as the rounds progressed so they could make game decisions based on feedback. Some of the initial options turned out to be game-breakingly overpowered. One of the players got lucky and basically ended up with a complete road-map to victory. The types and the frequency of information delivery seemed to provide players with a nice arc, so the overall strategy seems viable. But we’re looking at the volume of information that the players get, aiming for delivering less guidance so we don’t accidentally give away the farm. There was also some discussion of how the amount of information in a particular round could be adjusted up or down to modulate difficulty level.
(2) “Austen Translation” is a Multi-player Game
While our first game, “A Matter of Murder,” is a distinctly single-player experience, our play testers responded enthusiastically to the player-on-player competition that emerged in the AT paper play through. The addition of “scandals” that players could use to handicap a rival, or to make a suitor less attractive, seemed to make the multiplayer experience come alive. And the whole idea of social intrigue and infighting inherent in the game’s theme just dovetails so well with a multiplayer experience. So we’re going to be looking at ways to expand and enrich the multiplayer AT experience right out of the box.
(3) Tis Better to Receive
Some of the information options available in the paper play through turned out not to be information at all. They were really opportunities to act on information or to create information, like being given the option to spread a rumor about a rival. Some players felt like too many of these opportunities robbed them of the chance to get data they could use. These opportunities also felt very different from the events that gave players information. There’s probably room for this kind of manipulation at some point in the game, but we’re looking at ways to make it optional, or perhaps to create a special phase or moment in the game narrative when players can choose to take this kind of action.
(4) Marc Should Never Be in Charge of the Excel Spreadsheet
Putting the art school grad in charge of managing the game outputs had predictable disastrous results in a couple of cases. We’ve all learned a lesson. Hopefully it will never happen again.
So with some concrete gameplay experience in the bank, we’re back to structuring the rounds and developing content. Our next steps will probably be a crude but functional electronic model of gameplay. Stay tuned.
Worthing and Moncrieff, LLC is an independent developer of video game stories founded in 2015.