5 things devs should know before trying a narrative exploration game

A designer on one of the year's most successful walking simulators really wishes people wouldn't call them that. In a talk titled "Design Constraints in Narrative Exploration Games" at the Montreal International Game Summit this week, Campo Santo's Nels Anderson prefaced his remarks with a rebuke of the common shorthand for the genre.

"I think that term is really bad and it makes me sad when people use it," Anderson said. "Because partially it's like this in-joke. Initially, it was this pejorative riff on games like Farm Simulator or Euro Truck simulator or whatever. And I guess people are trying to reclaim that or something, but if you're not really up on all of that, it just doesn't make sense as a descriptor for a game. And considering this type of game is generally more accessible to people who normally don't play games, especially first-person games, I don't think that weird, jokey faux pejorative is good. It's just kind of dumb, because it's not like the walking part [of these games] actually matters, anyway."

"The player's suspension of disbelief is way more fragile in this type of game"

Regardless of what one wants to call the genre, it carries with it some tricky limitations that designers will have to work around. Anderson identified five major takeaways from his experience on Firewatch, starting with the fact that narrative exploration developers will necessarily be shackled by their premise. Where other developers can wrap their story around the basic gameplay of shooting things or solving puzzle games, the narrative is in many ways the entirety of the gameplay in a narrative exploration title.

"The what and the how of players exploring the actual gameplay is driven by the premise," Anderson said. "In a lot of games, it's pretty easy to back-solve from something that's needed for gameplay reasons to give it a place in the world of the game, even though it doesn't necessarily make perfect sense. Because you know that players will be quite forgiving of things that aren't entirely coherent if the moment-to-moment gameplay is better. But with narrative exploration games, that's not the case. The player's suspension of disbelief is way more fragile in this type of game."

That means developers have access to all kinds of themes and settings they don't have access to with traditional game genres, like Firewatch's Wyoming park in which co-workers Henry and Delilah develop a relationship through discussions over a walkie-talkie. But Anderson warned that it also means developers can't implement ideas that serve the gameplay if they don't fit with the story.

For example, Anderson said Campo Santo originally wanted Firewatch to have a Metroid-like structure, where players would find equipment that unlocks access to new areas of the world, or shortcuts back to earlier portions of the game.

"It turns out that all that optional stuff in a Metroid-type game is actually really important to its world structure. All those optional missile packs and health tanks provide meaning to unlocking those new locomotion abilities and thus new areas. It's not really good design to get a new item, have it unlock these new areas, and then there's just nothing there."

"It turns out that all that optional stuff in a Metroid-type game is actually really important to its world structure"

Campo Santo did add a small number of branching areas to the game with new conversations (and a raccoon attack) serving as the reward, but very few people actually saw them, and logistical issues kept them from embracing the idea in a more significant way. The optional side bits in Metroid work partly because it doesn't really matter when you get a health tank or a missile pack. Whether it's in the first hour or just before the final boss, they operate the same. But in Firewatch, the tone of any sort of discussion between the two main characters would be radically different if it happens in the beginning when they are just getting to know each other versus at the end of the game, when their relationship has developed and more dramatic events have transpired.

"That was very surprising and challenging for us," Anderson said. "When we tried to apply this game structure we were very familiar with, but we ended up having to twist and mutate it into a strange, almost unrecognizable shape because of the game's premise."

The second big challenge for the Campo Santo team was working in the present tense. Anderson noted that in most narrative exploration games, the player is trying to piece together something that already happened, and why. With Firewatch, they wanted to tell a story with events actively unfolding around the player over the course of a summer. And while that's incredibly common in other media, there were fewer examples for them to learn from in games.

The game uses a day-based structure where at certain points, the game cuts and jumps forwards in time. The problem, Anderson said, was that the cuts came at surprising times. While that helped them create a nice cinematic feel to the story in parts, it also discouraged (or even prevented) players from exploring outside of the main plot line and the areas where the plot was guiding them to. It also meant they had to come up with solutions for what would happen when the player found their way to some area ahead of the time when that area would be relevant to the narrative.

"We didn't want to do the old LucasArts adventure game thing where the characters monologue, 'I don't want to do that yet,' when a player clicks on something they're not supposed to be doing," Anderson said, before admitting that there was one spot where they couldn't come up with any preferable solution and actually used that device.

"We didn't have the luxury that Gone Home did, by being set inside a house, where people write things down and have personal belongings and stuff"

As a result, the team had to come up with ways to support the player being where they weren't supposed to be, or not doing what they were supposed to do. For example, there's a part of the game where the player gets a new walkie-talkie out of a lockbox. The game guides players to the box, but it doesn't force them to pick it up. If they opt to skip it, the player then goes on to play "a pretty good sized chunk of the game" that has entirely different dialogue and logic.

"We had to support this, because we just couldn't find an elegant way to force the player to have to go get that new walkie-talkie without just completely shattering the fourth wall," Anderson said.

The third big obstacle in development was that the developers couldn't rely on the usual ways of communicating information to the player, something that Anderson said looped back into the first challenge about being handcuffed by the game's premise.

"We didn't have the luxury that Gone Home did, by being set inside a house, where people write things down and have personal belongings and stuff," Anderson said.

There's not much of anything in the Wyoming wilderness, so they used written documents and manmade artifacts sparingly lest it burst the players' suspension of disbelief. They leaned on Delilah a lot to convey information, but she couldn't be some omniscient figure constantly feeding the player basic information. And because the story was unfolding in the present tense, there were events the player would see that Delilah wouldn't know about, unless the player chose to tell her about it in the game's dialog trees. They found a few ways around this, in some cases having Henry spontaneously deliver lines to nobody in particular. In one part late in the game where they needed to get the player to a location without Delilah's intervention, they created a fictitious science-y tracking gizmo to lead the way.

The developers also couldn't rely on game tropes like radar beacons and constant reminders about what their objectives are. If dialog included crucial information, it needed to get across to the player the first time. They couldn't have players examining objects in the environment or otherwise beginning new dialog to cut off the key information, and they couldn't have characters repeat the information time and again as the canned reminders would break immersion in the story.

"If you're ever thinking about making a narrative exploration game, I would think very carefully about how information is being communicated to the player and in just what ways the system could potentially fall apart, because unlike in most other games, you probably have very, very little to fall back on," Anderson said.

"Playtesting a narrative exploration game like this is really, really goddamn hard"

The fourth big takeaway for narrative exploration games is that the story will be the only thing the player is focused on. As a result, it can be harder to make a game world feel reactive. For Firewatch, Campo Santo built a dialog system inspired by Left 4 Dead, where the game logs when certain events occur and future dialog can be chosen based on what's happened previously. So if a player makes a certain dialog choice or takes a specific action, the characters may reference it later. The more specific the situation the dialogue is appropriate for, the more tangible and alive the game world feels in turn. The reference doesn't have to be done in dialogue, either. There's one exchange between Henry and Delilah where he gives a joking nickname to a hill, and the player chooses that name in a dialog tree from a handful of options. Later on when the player looks at their map, they will find Henry labeled the hill with whatever the chosen name was.

The last and perhaps most daunting lesson of Firewatch was saved for last.

"Playtesting a narrative exploration game like this is really, really goddamn hard," Anderson said.

If something wasn't working, if a scene or moment wasn't connecting with a playtester, it was very difficult to figure out why. Was it because the dialog hadn't been recorded yet and the subtitle text didn't have the same emotional oomph? Did they forget or misunderstand a key story beat from earlier? Was it a technical issue where a piece of dialog didn't actually play?

Anderson said he still doesn't know what the best way to deal with this problem is. Campo Santo had to build "huge" sections of the game to near-shipping quality before they could even properly evaluate them and start learning if they were on the right path. As a result, the game was built more or less in order, bringing each in-game day to a high quality bar before they could test it and move on to the next. It made progress feel slow, but the one good thing that did come out of it is that it actually allowed them to come up with many of the game's reactive moments. When the developers saw playtesters do something unexpected, they were often inspired to add new things in, whether they be offhand comments on the unexpected action from Henry and Delilah, or call-backs for later in the game. The former was fairly simple to implement as the game's voice-overs were recorded by the actors in their home studios and could be added to and changed with minimal hassle, while the latter was doable because those later parts of the game hadn't been built yet.

Playtesting the later sections of the game was also difficult considering testers would have had to play through earlier parts of the game as well, which meant they were either signing people up for marathon playtests when they didn't really need the feedback on the earlier portions of the game, or calling earlier playtesters back in and hoping they remembered all the story they had seen previously.

"Looking at this summarized in this way, there's a lot of 'hard' and 'challenging' in there," Anderson said. "Honestly, probably more than we expected going into making Firewatch. But I think there's also a great deal of unexplored design territory when it comes to making narrative exploration games."

Disclosure: MIGS has a media partnership with GamesIndustry.biz, and paid for our accommodation during the event.