Letting Go: A Florence Postmortem

[Music] my name is Ken Wong and I'm the creative director and founder of a little development studio called mountains our debut game Florence came out on Valentine's Day last year I was the game's designer and I also worked on a lot of the odd and story I'm excited to be here today to share with you some of the things that I learned from working in this game and there's lots of spoilers coming up so I heard that you've all played the game I have a lot to talk about so we're probably not going to get time for questions but I'm gonna stick around in like the breakout alcove later on so if you have if you want to follow up with me just hang around and we'll go out there and please remember to fill out your feedback surveys afterwards so the plan for today first of all I want to talk about the setup of this project like how it came to be next I'd like to talk about the evolution of the game idea has it changed quite a lot throughout the

development then we're going to look at some of the highlights of the narrative design and the interaction design and I finally like to wrap up with sharing a couple of the other takeaways and lessons but first of all let's talk about how this project came to be so five years ago I had the great privilege of working on Monument Valley as the lead designer for us two games Monument Valley is kind of an unconventional mobile game in many ways at ninety minutes it's a very short game it looks like a puzzle game but it's actually really easy and it's more about the journey than the reward at the end it's a quiet and a relaxing experience and from top to bottom we designed this game not for gamers but for people that don't usually play games and the game success gave me confidence that mobile players want high quality experiences it gave me the confidence in myself as a designer

and my ability to make bold design decisions I decided that the next challenge that I wanted to take on was starting my own studio so three years ago from London to Melbourne and started mountains this is our producer kameena myself programmer Sam and lead program attorney because of the success of Monument Valley I was able to secure a partnership with Annapurna and at the time their new interactive division had not yet released any games but the team included former members of Sony Santa Monica who published Journey her her command more when I meant with them I felt like we shared the same goals to make beautiful unique games so we signed up and I faced a lot of challenges when I was starting the studio but fortunately funding and publishing were sort of taken care of from the start in

my opinion many of the aspects of how we made game set us to worked really well for a certain type of game and I wanted to carry these over to my new studio probably the most important of those is lots of iteration we want to make innovative games and often that means taking a couple of tries to get something right our planning includes lots of room for failure because failure helps us gain the knowledge and understanding about the thing that we're making to support this process we do a lot of user testing we're not especially rigorous about it just friends and other game developers who come to visit the studio but we don't usually tell these play testers anything about the game and we don't tell them how to play they should be able to work all of this out from the game itself as they would as a regular customer and lastly we think that short games are okay like it's okay to make

shorter experiences Monument Valley taught me that there are many busy people out there that would rather have an experience that's short but higher-quality throughout I had hoped to keep productions short and tight at mountains actually wrote in to our employee handbook that mountains would ship a game roughly every 12 months and this is how Florence actually turned out 22 months the first six months could be described as a period of ideation we went through several different prototypes just to figure out what game we wanted to make it then took a further nine months before we could play through this this game this experience from start to finish and to learn how to make a good chapter of Florence and then finally it took a took about seven months of revisions and polish to bring the game to launch when I started

mountains I didn't have in my mind a particular game that I wanted to make now of course over the years I built up a long list of game ideas that I you know would love to make one day but I didn't know which one of these would be the great fit for this team with this budget at this time I think having a clear vision for what you want to make can be a great benefit but I also think it's valid to find the idea organically through a process of experimentation and discussion and that was the case with Florence instead of a game I had a starting point an area for investigation and what that was was manipulation of 3d objects on a touchscreen I feel like a handful of games have actually pulled this off including the room and shadow Matic but I wanted to see if we could find our own solution to this particular challenge

our first prototype to explore this was called parade parade could probably have become a decent game but I realized I wanted more than just solid mechanics and good level design I wanted to make an emotional impact so our next prototype included a more human element human heads the idea was to put human heads together like 3d jigsaw puzzles or like Rubik's cubes I thought that this mechanic would allow us to explore identity and psychology I thought it was pretty cool but once you put one human head together doing more heads just kind of felt like more of the same I still thought that the idea of 3d jigsaw puzzles had potential so I came up with a list of five things that we could use three 3d puzzles for and what the team really honed in on was a game about a relationship between two people if I were to write out what the

concept of the game was at this point it would be a story of a relationship told through 3d jigsaw puzzles we would reuse the structure of Monument Valley a linear series of chapters and each chapter would be a variation on a 3d jigsaw puzzle as a metaphor for the high and low moments in a relationship it had taken us six months just to reach this point and throughout this process Annapurna seemed to be patient thanks sector they know that you need to break a lot of eggs in order to make a hit indie omelet but the next nine months were a real uphill battle I slowly realized that my initial exploration direction of 3d manipulation was just not going to work out it turns out it's really awkward to manipulate 3d objects on a touchscreen and the third dimension actually wasn't adding

anything to the storytelling so we switched to 2d puzzles they were clearer and more straightforward but our concept of the game was becoming simpler and simpler and I was wondering is this what Annapurna had signed up for we also really struggled with how to make a good chapter of Florence it felt like for every four ideas we have three wouldn't work we'd spend two or three weeks creating like a chapter prototype only to find out that the emotion or the story just didn't come across sometimes we would have to throw away all of that card and all and start again with a different approach as the game designer and the founder of this young studio it was really disheartening work everyone around me was really supportive but I couldn't help but feel like my team and my publisher they were trusting in me and I was leading them down the path to disaster but gradually we found

mechanics that did work a few chapters started to feel really strong like they would actually make it into the final game more and more levels started to not be jigsaw puzzles at all and we opened ourselves to using whatever mechanic was needed in order to tell that story B we also let go of the idea of each chapter just being one mechanic instead each chapter became like a container that could contain one mechanic or several interspersed with non-interactive image panels the game kind of began to resemble a cross between WarioWare and an interactive comic book and I began to use storyboards like these to plan the chapters you could now describe the game as the story of a relationship told through different touch screen mechanics

in July of 2017 15 months since we started the company we could finally play through the game from start to finish in a very rough form it wasn't good yet many of the emotional beats just didn't land and the ending just didn't feel right but I now had a complete first draft that worked well enough that I could reflect upon it and start improving and I was at this point I was confident that some version of this game would actually ship let's rewind a little bit and talk about narrative I personally don't enjoy most narrative games I don't have anything against them they're just not my cup of tea but sometimes not being a fan of something can be an advantage game designer Zack gay she does not like word games so he made a word game that he felt addressed some of the gate the some of the genres

shortcomings the result is spelt our which is really really good so sometimes I think a skeptic or an outsider is able to find a different way of doing things perhaps an approach that the genre enthusiasts missed so when I talk about narrative in Florence I'm really coming from the perspective of someone who doesn't enjoy most narrative games but who loves other forms of narrative like film and podcasts and all forms of visual storytelling one of the first tasks to do with narrative was to cast our two main characters my first instinct was to try and make them as blank as possible maybe never give them names and their particular character traits and make them ethnically ambiguous I thought at the time that making them as generic as possible would sort of make them as relatable as possible to the widest audience I can't remember how we got the

idea in our head to go in the opposite direction to make these characters very specific perhaps we found it difficult to write a story for characters with no personalities or maybe we just realized that creating an average person is just impossible I found guidance on this topic in an interview with Constance worried about her television show fresh off the bone she talked about the impossibility of representing all Asian Americans in just one sure the pressure only existed because there was only one show depicting the Asian American experience but what they could do was tell an authentic story about one particular asian-american family by making these fictional characters specific detailed flawed deep characters they honored the complex and the varied lives of real Asian

Americans so as the primary writer I decided to fill the characters with details from my own experiences and those of my friends and family establishing our characters ethnic backgrounds helped anchor their identities and inform their relationship like me Florence was born in Australia to ethnically Chinese parents who emigrated from Malaysia her surname yer is my mother's maiden name like many of my female Asian friends Florence has a sometimes tense relationship with her mother who she only speaks to in Cantonese Krish on the other hand is Indian Australian I love the idea of a dark-skinned hairy creative guy being our object of desire Krishna's family is based on the family of a Sri Lankan friend from high school and his surname comes from one of our producers at Annapurna my South Asian

friends all told me that Chris would probably be very close to his family so Krish introduces Florence to his family early on in the relationship while Florence keeps it keeps the whole relationship a secret from her mother our couple happen to be two people of color of different faiths and they're the children of immigrants this is such a mundane ordinary thing in Australia which is where we made this game and where the game is set believe it or not you know we exist we have lives and we fall in love it's so normal and yet this representation is meaningful because we're not we're still not used to seeing people of color and people have different faiths as the central characters in the stories that we tell we received so many tweets and emails from people who played Florence who felt

themselves represented in a video game for the first time before I was a game designer I was an artist illustrator and art director telling stories with pitches is my native language that background really shaped how I design games and you can see that influence on Monument Valley and also in Florence both games use almost no words but make you make heavy use of visual composition color and music it took us a long time to find the Florence art style here's a couple of experiments that happened along the way in the one on the left I'm definitely trying too hard to come up with a really unique art style for this game the other two are a bit more subtle but I still don't think they quite set the right turn for the story I'm actually not very good at drawing the same thing over and over again but

especially characters I've never drawn a comic longer than a single page so it wasn't immediately obvious to me that we should borrow from the language of comic books but once we went down that direction it totally made sense I took visual cues from slice-of-life graphic novels and zine x' these would often feature incredibly personal narratives and would often be printed with limited colors the resulting art style is simple and ordinary and it's kind of rough around the edges but it brings a sense of vulnerability and authenticity these two mock-up posters were some of the first images that I drew in what became the final art style when I drew the poster on the riot I was like that's it that's the hero image of the game no other game looks like that and I would love to see that as the banner on the app store one day another

consideration with the art style was that many people would be playing this game on their phones a traditional comic page with like five or six panels on one page that would feel quite cluttered in this context and on this size of screen so our comic strip sequences are almost all single with an they scroll up and down more like a website or like a social media feed than a comic book and this lets people focus on one image at a time music of course plays a huge part in the story and we were very fortunate to have Kevin Peng Kanaan as our composer I could spend this whole talk on all the amazing things that his score does but it's probably best that Kevin give that talk another time today I just want to touch on one aspect of the music and that's that we realized that the two main

instruments could act as the voices for the characters we knew quite early on that Krish would play an instrument which ended up being a cello one of Kevin's first assignments was to write critius theme which is the music that you hear when Florence comes across him in the street you only hear the cello when Krish is around or when Florence is thinking of Krish Florence also has a theme which actually comes from Kevin's audition track for this project it's played on piano sorry the piano often serves as the voice for Florence when it came time to write music for the characters first argument I asked Kevin to just write a cello and a piano having an argument and it just worked perfectly a few more a few months before launch we had two curveballs to do the narrative that we had to deal with the first was a

conversation that we had about gender and sexuality Annapurna asked us if we would consider allowing the player to play as either a male or female version of the main character and the play would also be able to pick the gender of the love interest only the art assets would change and the story would otherwise remain completely identical creating two different sets of assets for these two characters would extend production so Annapurna was prepared to increase the budget and I was a bit shocked at home I didn't expect a publisher to ever propose increasing the scope and budget of our game internally at mountains we had discussed the topic of gender at the start of the project arriving at a female protagonist and a male love interest was the result of many different considerations for example as

the primary writer I wanted to write a straight couple because those are the experiences that I know firsthand we liked having a female point of view character to add to the underrepresented field of female protagonists in games and we liked the idea of a male person of color as the love interest as they're also underrepresented and it helped us avoid the sharp of the manic context the Manic Pixie dream girl but Annapurna's suggestion asked us to revisit all of these decisions I think there's definitely a game to be made in the same mold as Florence but with gender options is that a better game does it tell a better story I don't know every member of our team cares about representation but as the designer I'm not sure if if making this change late in the game would have been the best way to achieve that in the end I made the call to stick

with the path that we were on and only a few months away from completing I had really come to know these two characters and I really wanted to see their story through to the end the second late development was the rewriting of the ending since the beginning of development the story had always been about two people coming together breaking apart and then 30 years later meeting again by chance and we even prototype this ending I wrote this version of the story when I was recovering from a really difficult breakup but over the ensuing year of development I had gotten to know Florence really well she felt like a real person to me and I really wanted the best for her slowly I understood what all of my teammates and collaborators have been pushing for this all the time Florence's story shouldn't end with a

chance reunion with Krish and it shouldn't end with her meeting someone else someone you either all this time I had told Florence to let go but then I always ended up leading her back to Krish it turns out I was the one who needed to let go to let go of this ending that had been so personal to me that I once thought was the reason that I wanted to tell a story at all this part of Florence's life needed to end as it began with Florence by herself but because of the time that she shared with this really amazing person she became more confident more independent and more fulfilled so how do we say goodbye to Krish I really thought that was important at the last time that we saw him he was packing up his things on a rainy day and he left his toothbrush behind the worst so I wrote this tiny

little goodbye as florence is clearing out her office ready for the next stage in her life she finds an old photo of Krish Florence smiles it's a happy memory we hear teller for the final time she puts the fighter in the box with the rest of her images the rest of her things from her past and we close the lid so let's move on to the final way in which florence tells story interaction design in general I think there are some ideas of what a videogame is that could be summarized as traditionalists I think which traditionalists might say that game mechanics should all be about challenge and skill that games should be about choice and agency and the game should last a few hours long that doesn't come from any particular manifesto or anything that's just my impression

but I think that a lot of what being a creative and an artist and a game designer is about is questioning assumptions and searching for new perspectives some of the insights that informed Florence are that game mechanics can be instead used to evoke a feeling or an idea that we can use whatever game mechanic we need in order to tell the story and that games can be whatever length they need to be many of the games that have helped me question what a game is and and what effect game mechanics have on us are include passage papers please go on home and everything on this project I was particularly inspired by the vignette games of Jenny Garcia and Nina Freeman Jenny's game consume me is coming out soon and remember if you think it looks a bit like Florence it was her work that

inspired ours and not the other way around so let's take a look at how some of the mechanics of Florence work none of the mechanics are what you would call deep from a systemic point of view but that's not the point in vignette games mechanics are often about evoking a feeling or an idea or perhaps bringing up similar memories that the player has and thereby creating an empathetic connection with the character a great example is the toothbrush interaction it's such a simple action but it's a so many things at once the interactions not difficult at all but it's repetitive enough to suggest in monotonous work the presence of a progress bar signifies gamification as if Florence is living life for the sake of arbitrary completion the fact that of all the moments in Florence's day were featuring her brushing her teeth say something about how exciting her life is at this

point many of the interactions in the game are simplifications of what we do in real life perhaps even the things that we do on mobile devices so the game is using the players familiarity with mobile interfaces to tell a story while also creating a connection between the player and Florence other interactions in Florence are more abstract or are activities that are metaphors for what's going on emotionally with the characters one example of that is the chapter called fragments the player has to tear up an image of Krrish piece-by-piece but it's a metaphor for Florence losing fish from her life this is where the touchscreen can be so powerful the player is actually having to tear Krrish apart with their bare hands in this chapter we wanted to convey that

Florence is really nervous for her first date with Krrish forming a speech bubble out of puzzle pieces requires just enough brain power and just enough time to make the player feel like it's taking effort for Florence to think of the right words to say to Krrish the resulting speech bubble itself is a colored shape with no words in it and this suggests to the player that what's important is not what was said but what it took to say it Krrish on the other hand responds pretty quick as if he's relaxed and confident as Florence and Krrish go on more dates the puzzles get easier and the player applying the metaphor understands that Florence is becoming more comfortable around this person and the words are flowing easier we found that revisiting mechanics was a great way to show that the characters or the relationship had undergone change here we see a different take on the speech bubble mechanic when Florence and Krrish are having their first big argument we're still creating speech

bubbles out of pieces but the shape of the pieces and the way the characters talk over each other produce a very different feeling the final two chapters of the game are sort of a best-of compilation of all the mechanics from the rest of the game it's as if Florence and the player are using everything that they learned along the way in order to complete this phase of Florence's life finally I'd like to talk about some of the takeaways I have from working on this project I'm very proud of what our little game accomplished the success of Florence was enough for mountains to secure new projects so our little studio will continue to exist and to grow and for that I'm really really grateful but despite all of the awards that it's won and all the wonderful reviews and word-of-mouth thank you all Florence only did sort of bouquet it made its money back and then some we could afford to raise our salaries a little and do a little sponsorship but

we're not at the point yet where we can completely fund our next project by ourselves it makes me wonder how premium games fare when they don't win Apple Design Awards or when they don't get constant featuring on the App Store I think if your business plan is based around premium mobile you're up for a really tough challenge I don't know if that's any more challenging than standing out in the free to play crowd some industry commentators would say that game paths style subscriptions will be the next big thing but we'll have to wait and see who that benefits a couple of stats iOS makes up 88% of sales and Android 12% but note that Florence is not yet out on Android and China speaking of China China is our biggest market on iOS by far over 40 percent that's almost double the next biggest country by sales which is the u.s. so China is a massive and important market just a couple of anecdotal player

patterns very briefly we didn't expect that couples who play this game together that we saw that happen with our demo and then so that continued one with the full experience we heard of people that played Florence just after a breakup and that sounds really painful to me but we heard that it we heard that it helps the healing process some parents told us that their kids played Florence which is not something that I expected like I thought that it would be a nostalgic look back on dating I didn't really anticipate the effect of people you know playing Florence before they had started having relationships many people told us that they played Florence on the plane or in public transport and they usually shared this because the game made them cry and I couldn't help but notice that a lot of these stories about crying came from men and I think maybe this would surprise some people because they'd think that Florence was sort of a game

more targeted towards women but I think the signifies that men are actually more than open to more than open to having emotional experiences in games and sharing that oh yeah men cry too almost at a time this is not Florence this is a gays life a free web game made by a huang hala 27 year old developer from China a games life borrows heavily from Florence's art style and mechanics but it tells its own story that of a man discovering his sexuality against the backdrop of Chinese family and social dynamics based on Wong's own experiences unlike Florence it introduces the element of choice and makes heavy use of text it's only available in Chinese and has been played more than 1.6 million times this is the greatest impact that I could have hoped for for someone with less privilege and resources than us to be be able to be able to benefit from our work and use it to tell their own story I don't feel particularly possessive

over any of the techniques or insights that I've discussed today so if there's anything in Florence that you feel is useful for you and your work please feel free to borrow adapt or steal right now games could use a wider variety of stories from a wider range of creators and we couldn't be happier if Florence helps people achieve that thank you very much [Applause]