Note: The following article contains major spoilers for What Remains of Edith Finch.
Ian Dallas is the creative director at indie studio Giant Sparrow and oversaw the development of The Unfinished Swan and What Remains of Edith Finch. Recently, we discussed the inspirations behind and production of What Remains of Edith Finch.
Gamer_152: In terms of art style and play, Giant Sparrow's first two titles seem about as far apart as two games could be. Were there any stylistic choices or lessons that you carried forwards from The Unfinished Swan onto What Remains of Edith Finch?
Ian Dallas: No, I think the different styles were just a reflection of the different worlds, and to a large extent, the different artists who were involved. We had an illustrator, a 2D artist, who led development on The Unfinished Swan and then on Edith Finch. Actually, our lead artist came from Call of Duty, and he was really good at building giant lists of things to make in a military-like fashion, just burning through those lists and getting the art team to do that. So we played to his strengths and created a world that had much more realistic lighting and tons and tons of little details because that's what the artists gravitated to and what they were good at. So, in addition to just conceptually these worlds being different, and one being in a fantastical space and one being in a more grounded reality, I think a lot of it was actually just the artists who we had on those projects.
G_152: It must have been a bit of a jump from Call to Duty to something like Edith Finch.
ID: Yes and I don't think that he ever fully landed exactly. It was an interesting time. The lead artist for Edith Finch, Brandon [Martynowicz], came on a little bit late in development and was kind of difficult to get up to speed on everything. So, I'm not sure that he ever 100% understood what we were making, but we were able to play to his strengths and the things that he enjoyed and did well in terms of creating lots of rich detail in the world. I think the game looks as good as it does because we were able to connect with what motivated him.
G_152: You did mention starting with the world there and the world defining The Unfinished Swan and What Remains of Edith Finch. Is that where all of this came from? You didn't start with the gameplay; you just had this idea of a world and wanted to move forward from there?
ID: No, not at all. We don't start from a world, we don't start from a story, we start from a feeling, and on Edith Finch, it was the feeling of the sublime and experiences that I, personally, had scuba diving that have evoked the feeling of something that is beautiful but also unsettling. And then we would just start creating prototypes that evoked that, and the house was one of those things that emerged as a surprising conduit for sublime feelings in its scale, and also in the density of clutter approaching more organic experiences that people have had that just evoke that sense of wonder.
G_152: I understand that the underwater diving prototype was how the game existed for the first three or four months, but you then pivoted to this idea of the house at the end of that period. It seems like most studios are locked into their chosen concept by that point. What allowed Giant Sparrow to make that change so drastically that far in?
ID: Well, again, because we're starting from a feeling, there's a lot of roads that lead to that destination, and I think in a lot of studios you're starting from "Okay, we're making a cooperative deathmatch capture the flag-based experience", and so you have made a number of decisions upfront that either inevitably result in certain kinds of experiences, or you're dealing in a space that players have such firm expectations about that it is difficult to go off in different directions for developers and for players. But because we are quite a bit looser early on, all we want to do is evoke this sense, and we don't really know where that's going to come from, how much is going to be audio, how much is going to be context, just like a few words that set the mood, and how much of it is gameplay, and all these factors. So, we're just throwing stuff together to see, and often we're wrong, because we're going after experiences that are unusual in the real world and rare or non-existent in games, so there's a necessary element of experimentation that is there.
G_152: Was it from starting with those underwater sections that you ended up with sequences like Baby Gregory in the bathtub or going out into the sea when the tide is down to see the old house?
ID: I think the most direct child of the early underwater prototypes was Molly's story. So, the seaweed, for example, when you're swimming along as a shark, that seaweed and that terrain was something that we built for the scuba diving prototypes, basically, although a lot of the early prototypes we'd actually build in Unity instead of Unreal because Unity was a lot easier to work with out of the box. Eventually, as we became more familiar with Unreal, and as we built more tools, it was easier to prototype there, but I think it was useful to be working in a completely different space. I think there's a tendency to try to move your prototypes, to take that, and then turn it into something you can ship, and it's kind of like building a balsa wood framework of a house as a model and then trying to use that balsa wood to make your real house. It's like "This is a bad idea".
So, aesthetically, a lot of the things we just started over from scratch once we understood if and what we were going to do, so there wasn't a lot of direct transfer from prototypes to shipping game experiences which is a good thing, generally, but in the case of Molly's story, there were some very direct elements there that translated. We also were interested in flashlight gameplay early on, and Edith had a cell phone that was going to be a big part of her character, and we discovered over time that that was not something that made a whole lot of sense and felt good. For the Edith section where you're exploring the house, there was already plenty to occupy the player without having another thing to interact with and make the house even more confusing to navigate, like a flashlight, which was something we had in the early scuba diving prototype that worked well there.
G_152: I know during development you were dropped by your initial publisher, Sony, but Annapurna picked you up and they facilitated work on the game that wouldn't otherwise have been possible. Do you think Sony and perhaps console manufacturers as a whole are starting to opt out of indie publishing?
ID: It's really hard to say. I think we did not have much visibility into why decisions were made. Sony is a very large company, and the only people that we ever really interacted with were super nice and very understanding, particularly Shuhei [Yoshida], who has superhuman levels of comprehension and kindness, and every interaction we had with him was fantastic, and I'm sure there were very good reasons outlined in spreadsheets we never saw for why it didn't make sense to continue development of the game at Sony. But they were very generous in helping to find a home for the game, and without their help, even when they decided it was no longer a good fit for them, if they hadn't been super helpful, we would not have been able to find another place to bring the game.
G_152: So it was a case of them giving you the resources to put out something that Annapurna then saw or were able to pick up, and without that early stage Annapurna wouldn't have seen it and picked it up?
ID: Yeah, yeah, absolutely, and we're very grateful to have had the opportunity to work with two publishers who were generally very supportive, and from a developer standpoint, bizarrely so. I don't quite understand why, at any point, a giant for-profit company would have an interest in the sort of experiences that we make, but I'm happy that that has been the case.
G_152: Getting back into the game itself, this is a game that feels very knowing of family conflict and death. Was the development team drawing from their own experiences when putting this together?
ID: A bit. The character of Edie is modelled off of a similar character in One Hundred Years of Solitude, named Ursula, who's the matriarch in that book, but then also significantly based on my own grandmother in some aspects of her character, but there's not a whole lot of direct [characters], other than that one that comes to mind, from my own life, that made their way into the [game], but some of those dynamics are things that echo experiences that I've had. My own family was so placid for the most part, just well-meaning people who were doing their best and not too outsized, everyone survived their childhood, so it wasn't like a literal move there, but certainly, when it came to the props and things in the house, that was something we were all drawing from. Things that we felt echoed suburban family experience like having a tennis ball in the garage that hangs down on a string so that when you pull your car in you know when to stop. I think that's something that I was happy to find out our family was not the only one that had a tennis ball like that hanging, that that was just a relatively common fixture for people.
G_152: So was it a case where the inspiration was more literary than personal?
ID: Yeah. I think the challenge we had was evoking a sense of sublime which is a rare experience in real-life and also a pretty rare experience in art, so there wasn't a whole lot to draw from but weird fiction, this genre of literature from the 20s with people like H.P. Lovecraft, as an example, was really helpful, so there was a fair amount of short stories and One Hundred Years of Solitude, [Jorge Luis] Borges, the films of Luis Buñuel, there were a few templates for how to create the sense of the sublime and awesome mixed with surreal elements. So those were the models, but none of those except One Hundred Years of Solitude are really about families, so we were taking aspects that we found in references, and then grafting them onto something that was coming a little bit more from our lives and experience.
G_152: I don't know if it was a direct influence, but I did read one article which made a connection between Canterbury Tales and What Remains of Edith Finch.
ID: Definitely not an influence. Structurally, there are some similarities, I suppose, in terms of frame story and tales within tales, but my interest in literature really stops at Shakespeare and doesn't pick up again until Homer. It's quite a gap. I tried reading the Eddas last month actually, and it's pretty impenetrable. The Middle English period, I would like to have more familiarity with. I hear good things about Faerie Queene and other bits there. Actually, I take that back; I really like Beowulf. That's a story that I listen to somewhat regularly. But no, Canterbury Tales was not a direct influence, though it could have been if I was more open to Middle English.
G_152: I guess there's something a tiny bit [Homeric] about [Lewis's] segment, with the king in the fish plant.
ID: Yeah, I think there's a little bit of Homer in everything I do. I'd say Homer and Shakespeare are the two biggest aesthetic influences. I think my sense of the epic and grandeur mixed with mundane aspects of daily life is something straight out of Homer.
G_152: Are there areas of Edith Finch which you feel draw from Shakespeare?
ID: I'm sure there are; I can't think of any specifics. I think, for me, one of the things when I think about what I love about Shakespeare, and what I find missing in contemporary stories, is the treatment of villains. In Shakespeare, all of the villains are usually the most interesting characters in the stories, and with the exception of a few people like Iago that are a little moustache-twirling, for the most part, like in Macbeth, I would say is the prime example, we really identify a lot more closely with the more classically antagonistic forces in that world, and for me, that's a lot truer to reality. That I don't think there are people who are evil, that it's enough for everyone to be trying to do good, but to have different ideas of what that good is for there to be problems in this world, and stories nowadays generally have forces of antagonism that are very convenient that allow the heroes to be white hats fighting the black hats.
There's no overlap there, you know, all the Nazis were bad people; that would never happen in England, or anywhere else. I think that stories that don't let the heroes off the hook as easily, that suggest that these destructive elements are things that we all carry inside of us, and have to learn, in order to not fall prey to them, I think is something that Edith Finch is grounded on. That there's no villains in a Disney sense, there's no evil; it's just misunderstandings and people pushing things to extremes that create difficulty.
G_152: The game is also walking a very fine line between being a series of vignettes and a single overarching story. Were you ever worried about ending up too far on one side of that line or the other?
ID: Yeah, that was one of the biggest challenges. Probably the biggest difficulty that I, personally, ran into was trying to find a satisfying balance and conclusion to all of that which is made more difficult by the gnawing question of "Do you even need any kind of framing device here?". I think The Ballad of Buster Scruggs, the Coen brothers movie that came out a month ago, is a good example of an anthology story, or an anthology series, that doesn't actually have any common elements in terms of characters or plot pieces. I think there's a deeper level of shared thematic concerns and aesthetics that run through the stories, but in our case, we looked at things like The Twilight Zone, I think a pretty good model for doing an anthology series, and that succeeds in not feeling random. Which, if you didn't have the initial introductions from Rod Serling and the music that's consistent between every episode, then they would just feel like a jumble, and even then, the connections are very tenuous, but because it's only a twenty-minute, thirty-minute episode, that's sufficient. We felt like for a two to three-hour game experience that, unfortunately, was not going to be sufficient, partly because anthology series have almost never been done in games before.
The player expectation is that they are absorbing something that is a cohesive whole, so having a strong throughline was something that we felt, as players ourselves, was necessary, even though I suspect there would have been a way to make a game without it. It just would have been a very different experience, and then we also found because many of the stories ask a lot of players emotionally, that it was helpful to have moments of low tension between those stories to give players time to recover and to vary the emotional pacing, probably just so we can ratchet it up again. You know, we're just interested in jumping into these intense moments, but you can't sit there and eat a bowl of cinnamon; you need something to spread that out. So, that was partly also why we ended up with something that balanced out this overall frame of story with the more intense gameplay-oriented vignettes.
G_152: So was it a case of making some of those vignettes and then working backwards to a message or narrative that tied them together or were you having the narrative propel a lot of the vignettes?
ID: No, the vignettes were in the driver's seat. Again, our goal was to evoke a sense of sublime and the vignettes is where that magic happened, if ever, and so it was quite hard enough to create those sublime moments through interactive gameplay on its own without having to worry about basing it off of things that would be convenient. We started with a general framework of experiences from 1900 on. Things that involve children and families and worlds that we were familiar with, so I think we were building the vignettes from ingredients that, for the most part, we understood, and were in this common pool, so hopefully, they feel like they're a little bit cohesive, and then once we had stories that were playable, and that we could wrap our heads around, then we went in and made things cross over a little bit more. For example, having Calvin's story take place on the swing that you can see as Edith and that you can also see as Molly earlier on, were things that emerged somewhat organically. Without having to work too hard, we'd blended these things together to feel more like a cohesive whole, but they did not start out that way. They were just trying, from a raw gameplay experience, to do something, and then once that was working, we tried to, while the clay is soft, manoeuvre it into or massage it into a desired shape.
G_152: So was it a case that you had all of the vignettes before you worked out the narrative that wrapped around them?
ID: We had everything coming together over the whole four year development period, but I would say the Edith story didn't solidify until the last couple of months. We tried to do that work ahead of time. It's very scary to have huge chunks of the game that aren't working, but we did the best we could and then we looked at it and went "Ah yeah, this is kinda shitty, but it's the best we can do", and then as everything else settled down, and as we had a chance to spend more time with it and steep in those flavours, we had a better sense of what that overarching narrative with Edith actually needed to do, and what questions players would have after certain stories, and where they would be emotionally, and we could write a little bit more to those concerns and questions. So we had stories early on, but they were just terrible, and it wasn't until the end that they evolved pretty drastically to fit the shape that we found that we needed.
G_152: That's amazing to me, as a player, because going in, it feels like it was built from the ground up. It feels like it strings together so very carefully that I didn't imagine that the little vignette sections could have been made completely separate from the main storyline.
ID: Well thanks. It's really all working backwards which has a similar effect really of "It feels that all these things are sutured together so nicely, as if they were custom-made", but it turns out to just be a lot easier to make the wrapping around it than it is to make the nuggets of gameplay, and that, again, comes out of the challenge of gameplay [that's] so steep. That it's very, very hard to change gameplay drastically without having all kinds of knock-on effects. So we started with the thing that was the most difficult to make, and the most difficult to change, and then all the story and art that sits on top of that experience was much more malleable. So, there was a lot of work trying to get it to feel cohesive, but I think it would have been an even more daunting challenge if we had started from that cohesion and left holes for "Okay, we'll have satisfying gameplay that will fit inside this hole". That's something that's beyond my skill level.
G_152: There's something that makes sense about developing it backwards because the game is also told roughly in that way. You start with the older members of the family, and you work your way back until you get to Edith's story.
ID: Yeah, although funnily enough, in terms of development chronology, Molly's story was the very first one we ever did, and that's the first one that you experience as a player, so it's not quite as simple [as that]. But I think it helps when dealing with something that is stranger than you, yourself, can imagine to go at it in a way that's not necessarily linear, to give your subconscious freedom to assemble these things as you see fit. Also, the stories changed in their own chronologies pretty significantly. Lewis's story was originally going to be right after, or one or two after, Molly's story, and he was a part of her generation in the family, and then we looked at it in terms of the amount of time that players would spend in stories. It felt like "Oh, we want to have a longer story towards the end to balance out the pacing", and a lot of those stories did shift, also, at various points. Family members moving around generations and stuff when we had a better sensing of what the pacing and mood arc should be.
G_152: You did mention there, other overarching narratives for the game that weren't used. Were there any particularly interesting ones on the cutting room floor?
ID: Everything that generally gets cut is cut for very good reasons. We almost had to cut Lewis and Gregory's story. That was something that maybe a year out from development there was a lot of encouragement to cut because we were dealing with shipping dates that were rapidly approaching, and fortunately, we had some extra time and the budget got a little bigger so that we could do those two stories. But in terms of the ones that shipped, no, I don't think there's anything that was not cut or changed without, in my mind, a really good reason, and mostly they were changed early on enough that I don't think anyone was too attached to them.
G_152: I'm sure. I just wondered if there was any other idea for stringing together the vignettes outside of Edith exploring her family tree.
ID: No, that was baked in from the earliest days. We had the name Edith Finch from day one; we just didn't quite know what her path was going to be through that. Initially, there was the idea that it was going to be a collection of high school students in detention sharing their dreams, but by the time anyone else was actually working on the project, it was grounded in the family and Edith exploring that.
G_152: The game also feels very purposefully paced and very mysterious until right near the end. We can't really know what's going on until we get right near the very end of the story, the last few minutes of it. Was there ever the urge to try and put more of the plot up front to keep the user dialled in?
ID: I don't know. I think things evolved pretty organically based on the pieces we had and how players were responding in playtests. You know, moments where it felt like they were getting a little bit ahead of the story and maybe feeling a little bit bored, we'd lob in a few more cryptic suggestions or what not. A lot of it ended up being dependent on how long it took players to move through spaces in terms of the density that was there. I think we usually found that we needed less than we thought, especially early on when players were just trying to wrap their heads around this enormous family tree and seemingly very complicated house to navigate. So we had the ingredients, and it was just a question of "How do we sprinkle them through in the best way?", but that's something that emerged watching playtests and responding to that feedback.
G_152: If I remember rightly, you had a very impressive selection of playtesters.
ID: Yeah, there's some well-known names that go by in the credits, and one of the things we have found was very helpful is to do playtests at places like the Game Developers Conference where a bunch of developers are already gathered, and we can pull people into a room and have them walk through the experience. We were fortunate enough to get a number of well-known eyeballs on the game at various points.
G_152: Before you put this out, was there the feeling that this was going to be something that was going to be highly affecting for people, going to be taken really seriously in terms of picking up awards, in terms of the critical reception, or could you just not tell at that point?
ID: I don't remember. I don't think that any of us had a good idea of what the reaction was going to be. With Gregory's story, in particular, there was a concern quite early, on the part of the publisher, that that story be handled very sensitively. The main design liaison that we had at Sony is someone who has children of his own, and he expressed a lot of concern about that story being sensitively handled, and we found that was mirrored in future playtesters. That playtesters who had children of their own tended to respond in a much more intense way to the prospect of a child drowning in a bathtub. I think it was a fear that they, themselves, had had, and so seeing it enacted in a game was potentially traumatic in a way that it was not at all for people without children. Or maybe also young siblings too. If you had experienced that fear in real-life then seeing a virtual approximation, you had quite a different experience. So that was a story we were sensitive to and tried as hard as we could to handle delicately, and then, also, brought in playtesters periodically who were parents, specifically, just to make sure that it was shaping up in a way we were happy with.
G_152: I imagine some of the playtests must have been pretty emotional.
ID: Not really. I think emotional experiences are very easily broken, and playing a story where, for example, the shadows are constantly flickering, or objects are winking in and out of existence, or you just get stuck. Where you jump up, say, if you're the frog, and you just hit an invisible ceiling, and the developer has to come in and type some things on the keyboard, and then you go into debug mode for a minute, and you see this neon sea of lines for DebugDraw in order to get over the problem. All those kinds of things are very common in early development builds and have a tendency to reduce the emotional impact. It's like hearing a piece of music with occasional screeching in the middle of it. It ruins a piece of chamber orchestra or whatever you're listening to, and so there's always a little bit of a guess on your part.
Also, people respond so differently just knowing that they're in a playtest, and unfortunately it's pretty difficult to create a situation that feels like the emotional state people would be in when they were in their homes playing it. For example, we would bring in people periodically to do focus tests, like five or ten folks, who'd be paid for their time, random people off the street basically, and they would just play through the game, and we probably saw thirty or forty people play the game in that context, and not one of them cried. We never saw anybody cry at any point during a playtest, but that's something that anecdotally I've now heard from many people that there are moments in the game that made them cry. So, I think it's a very different experience when the game is all there, and when you are playing it in your own home vs. any of the playtests that we were able to experience during development.
G_152: I guess the ideal for a developer would be some sort of nightmare Kinect situation where it's reading your biometrics in your own home so they know what to change where.
ID: Yeah, yeah. We found that it is a useful surrogate to provide another player and a little bit of alcohol. So people playing in pairs with a beer in front of them tend to relax in a way that you don't get when you have one person in front of a TV. So, the nightmare Kinect scenario, halfway there is a friend and alcohol.
G_152: The Unfinished Swan and What Remains of Edith Finch are both incredibly self-contained games. Do you think there's something to this idea of cutting back on the degree to which we sequelise games in order to improve the storytelling?
ID: I don't know. I think there are times where sequels make a lot of sense. I feel, generally, we see sequels primarily for production reasons and not for narrative ones. It's much easier for a team to make something that is similar to what they've done before, and that's not necessarily a bad thing. I think you end up with happier development teams and more elaborate experiences that players enjoy. So, I think there's nothing inherently wrong with sequels, but as a studio, our focus is on creating experiences that people have never had before, so we haven't found that sequels were the best use of our time so far.
G_152: You have a little distance from the game's release now. Was there anything that the audience received better or worse than you thought they would?
ID: Yeah, I mean I would say, overall, I am quite surprised and humbled to hear about how many people were quite emotionally affected by the experience. I think this comes from watching well over a hundred people play through the game, and not seeing anyone cry, or seeing anyone have anything but moderately kind words to say about it. But when it's out there in the wild, and it's exposed to more people, and under different circumstances, it's been nice to hear that there are so many people out there who come into these experiences with an open heart and are willing to be transported sometimes into unpleasant, difficult places. In terms of media portrayal of gamers, and what you hear online even, like if you listen to message boards or whatever, the people there tend to be a more hardened sort, and so it's great to see that there are actually so many players out there who are willing to go into something that may be uncomfortable, and approach it in the way you would any sensitive piece of art, and that there's an audience there that is hungry for those kinds of experiences.
G_152: It's interesting hearing this because I think we have this idea of games as being complete packages which have an inherent emotional experience built into them which maybe some people get, maybe some people don't, but there's some innate set of experiences built into that, whereas what this suggests is that it's a combination of having the right game and having the right environment that really brings it about.
ID: Yeah, and I think looking at the historical landscape of emotional experiences that games have offered, you might assume that that's all they could do, but I think, yeah, you're right that there's a much broader world out there that people are accustomed to, like in movies, and books, and real-life, and whatever. That they're not coming to this as people that only play video games, they have a lot more experiences to draw on, and so, I think there's a lot of room for games to expand into other areas than where they have historically been pigeonholed.
G_152: I'm just really excited to see where it all goes. This is almost a fluff one: Do you have a favourite vignette?
ID: I think for me, Calvin's story is the one that is most distinctive and emblematic of what our game does really well. I love that we were able to spend three years, off and on, tuning that experience into what it is for something that is over within two minutes or so. That that kind of concentrated, well-executed burst is in some ways an extravagant inefficiency in the way that resources are used, but I love that it generally lands pretty well with players and creates moments of genuine surprise.
G_152: Was it a case of tuning the physics with that?
ID: The physics were a big part of it. A lot of it was also players not understanding that they could use both sticks to move the legs. That was something that even pretty late in development, we had problems where, if you're moving one leg, which so far in the game you've only needed to use one stick at a time, so using two sticks is kind of a leap for players. We saw a lot of players who would continue to use one stick for a long, long time. There's a tendency for players to, once they find something that works, to keep doing that to the exclusion of all other possibilities. So tuning the response and the physics and the aesthetics, adding a cast to one of the legs, for example, was part of the solution, so that the legs looked very different, and you had this sense of left and right leg. Tons of little fixes like that that were necessary to address the problem for almost all players.
One of the things that we've found is that, usually, one solution does not cover all cases, so you need to solve the problem multiple times in different ways to address the variety of players that are going to be experiencing that. It's a lot of under the hood stuff like that that becomes surprisingly vexing, or the way that the hands move on the chain was problematic for a long, long time. Just in terms of the hands passing through the chain at points when we didn't want it to. There was a lot of thorny technical challenges that needed to get sorted out and that never seemed to be the highest priority thing, so it would be a year or two before we'd come back to. So yeah, it was a bizarre amount of time and energy that gets spent on something that is over so quickly and most games would never dream of doing that, so I love that we had a chance to do that there.
G_152: Something I've heard Jeff Gerstmann repeatedly say is that it's a miracle that any video game gets made at all.
ID: Yeah, absolutely, and I think most games you're seeing, by the time they get to you, it's maybe a quarter of what the team had intended. That it has been cut down and simplified so many times, and even that is a herculean effort on the part of many people to make it. Yeah, it's a miracle that any game comes out.
Thanks to Ian Dallas and thank you for reading.