Author: by Alissa McAloon, Bryant Francis

This month, Shadow of the Tomb Raider capped off the new trilogy that began in 2013. As development moved over from Crystal Dynamics to Eidos Montreal, we were curious about the design and production processes that go on at large triple-A studios that bring games like Shadow of the Tomb Raider to life -- and the obstacles these developers encountered.

To that end, we were delighted to chat with narrative director Jason Dozois and producer Fleur Marty last week on the Gamasutra Twitch channel as we played the first hour of Shadow of the Tomb Raider. Here are excerpts from our conversation.

Edited for length and clarity.

Designing highly-flexible difficulty settings was...difficult

Marty: So first of all, you have to understand where [the game's difficulty settings] come from. That's first and foremost, from feedback from the community from the previous games, especially from Rise, and people on one side complaining of puzzles, that the most interesting puzzles being side content, not being on the main path. And also the white paint [which highlights the critical path] being too obvious and things like that.

Shadow of the Tomb Raider's deceptively simple difficulty options

So we were trying to take that into account. But then when you go into playtest and you have more casual players or people who are not familiar with the franchise who try to play the game, then they get frustrated by the difficulty of the puzzles, or they can't find their way around. So we were back at kind of "dumbing down" the difficulty. We weren't satisfied with it at all. Especially since, with this one being the end of a trilogy, we had to throw the most challenging puzzles possible at Lara and the hardest combat; we couldn't really ease up on the difficulty.

We came up with this idea: that maybe for people who love challenging puzzles, but who are not familiar with or do not like the combat or who didn't want to spend too much time finding their way around, they could have hard puzzles but easy exploration -- white paint everywhere, and easy combat. Or it could be completely the opposite.

Of course, that means a lot of changes. That means you're not balancing your game for three difficulties -- normal, easy, and hard. You're actually balancing it for like, nine, and all the possibilities. So that's one thing. And then of course [toggling the] white paint, that means creating systems for which we applied depending on what level of difficulty has been set for the exploration, and making it disappear, and the tutorials, and things like that.

When you think about it, the easiest [difficulty setting to adjust] is the combat, because that's what every game is used to doing. Like easy, normal, hard combat; well, that's easy to figure out: You have more bullets and enemies have less HP. That's why [offering so many difficulty options] was a challenge, and since it was a nearly last minute idea -- not last minute, but it came a bit on the side. That's why we at some point were like, "we're not going to make it, we should just cut it." But then when we saw the reaction from players to it, and we realized that we've got something.

Dozois: Yeah, there was a group called "G.R.O.W.L.", I don't know if you guys are aware: "Get Rid Of White Ledges." That is stuck in my head. So this was really community-focused. You know on the puzzle aspect it was really playtest-focused, like Fleur was saying. You make a puzzle and, we had some hints on a timer, say like three minutes, five minutes. We're trying to find the right number, but there's no right number. Because some people, they want to be stuck and they want to figure it out. And some people never want to be stuck. They think about it for two seconds and they're like "nope, tell me what happens next in the story." So you're trying to find a sweet-spot number, but it didn't pan out.

Shadow of the Tomb Raider lets players choose to highlight the critical path -- or not

So what we did with the puzzles -- people said how do you make puzzles harder? No, we just do a hard puzzle, and we give you more hints. So most of the design was mostly on the writing side. For every step of the puzzle, if you're on normal difficulty and you request it (you have to press survival instincts for her say something) she'll suggest something like she's puzzling it out. But if you're on easy, she almost like gives you the answer: "I think we need to open that door with that thing." It's like almost telling you.

So what it does, it allows you to customize your [puzzle experience], and for us it was like solving a problem. When we went to E3 and everyone was asking about this we're like "Okay yeah, it's good, it's awesome." We liked it because it was solving a problem but people were seeing it as like "Wow I can really customize my experience and that's really cool."

Marty: In retrospect, when you think about it you're like how did we not do this earlier? Like, why?

Dozois: Everything is obvious in retrospect...obviously people are responding and reacting well to it. It allows you to balance it in a way that seems obvious now but was never obvious before, and I think it's gonna inspire a lot of people to do something similar.

To me, [the moment when] I realized that this puzzle difficulty thing was a real success was on some of the optional challenge tombs that I had not played as often. I was playing and I was like, "I have no idea what to do," and I thought "wait a minute." I put it on easy and I found the hint and I figured it out. That's when I was personally really sold. Because on my side, it was [writing] hundreds of lines like, "do we really want to do this?" It's a lot of, "I need to open the door, I need to do..." You know, it's not the most inspirational writing, to give the hint. But when you want that hint, when you want that one sentence, and you get it, it's awesome.

The production "nightmare" of a characters speaking their native language

Dozois: [Immersion Mode, in which NPCs speak their native languages] came from a debate. Because you look at Rise of the Tomb Raider everyone's speaking English with an American accent, and these guys are from basically Turkey, Asia Minor from like hundreds of years ago. So you can either take the Star Trek approach and go, well, everyone speaks English because it's not a story about languages. A lot of games do that, and that's fine.

You can take the other extreme you know like the movie Apocalypto where everything is in Mayan and that's it, it's just a subtitled thing. [But]...where do we put the line?

We talked about saying everything that's a mission, or on the critical path, we would say we would localize it [for its respective market], meaning it would be in the language of the SKU you set it to. And everything that was optional was either in Spanish ... or we'd mix in Mayan later on.

...It's one of those questions of how far do you go, right? Are you going to do all you cinematics in that [language]? Are you going to do two versions of the cinematic? We have two hours of cinematics, so...where do you draw the line? We're tying to make it as immersive as possible so when you're walking around you feel like you're in a place that's alive.

But then one of the problems is that if it's in English, you can understand an overheard, like an enticer saying "Hey can you help me?" -- well, then I don't need a subtitle to know I can talk to that person. But if you have it in a language that you don't understand at all, without the subtitles you're not even going to know.

So then we had the subtitle problem -- that's why we even have a different subtitle setting. We have critical path subtitles and we have immersion subtitles because at some point you can have so many subtitles on screen you don't know what's important. So it was an ambitious thing to try and do. When it works well it's really cool and it is immersive, but I think people will pick their preference.

We didn't really want to force one or another, and it's there. But it wasn't really practical for us to do it everywhere. I mean, the awesome thing would say, "Oh my god the whole game, and even Lara, Jonah, everyone is speaking Spanish, Mayan, and everything and wow it's amazing!" And then you'd watch for a few minutes and go, "Oh well I don't really understand, I don't want to read, I want to be able to see the action."

... At least we said "we're going to immerse you" -- that's why we called it Immersion Mode, but we're not saying the whole game could be overridden and changed to the local language.

Marty: I'll have two things on that. First, one thing to notice, is that some characters, like there's one specific character in Paititi, and I'm sure they're not the only one, that speak to Lara in English because they are considered main characters or cinematic characters. But when they speak to their families they actually speak Mayan when you have Immersion Mode on.

The second thing I wanted to say is that it's been a nightmare on production. There's been so much debate around the team for months and months and months. It was on, it was off, it was back on. We playtested this so much, asking people, "does it bother you to read subtitles," because it's known to be an issue even for people who go to the movies and they don't like to read subtitles.

So are you going to play a 25-30 hour long game with subtitles on, are you going to read all of them, and if you're not reading them how much of the story are you missing? And stuff like that. It was debates and debates. I think everybody on the team agreed it was the right thing to do, to have the languages in, but the headache was how can we do it right and how can we do it so that players actually want to enable that option.

Fixing the pacing of the game -- mid-production

Dozois: That happens, of course. You need to adapt! Either you adapt the content or you cut it or you find a new contextualization for it. Sometimes, we can move sections around and it actually fixes a problem.

If you look at the beginning of the game, that plane crash that you saw, that wasn't at the beginning originally. That was a bit later, right after the first hour of gameplay. And we were like, hm, you start and it's a little depressing. Where's a moment we can get a little excitement and amp it up? So we looked around and we put that there and some people were like "No you can't move it, argh!" But we moved it and we tested it and people really loved it, so we left it there. Stuff like that comes, and you can move things around as it comes...

Marty: I want to add something about the process when Jason is talking about moving things around, and realizing that they actually move better like that. Throughout this entire production we've been a very on-screen focused team. I think it's been the project where I've seen the least amount of paper design my whole career.

Jonah and his intricately-designed almost-cargo pants

One thing that has been done a lot is Jason and Dan [Bisson], our creative director, sitting together and editing the game. Jason will play a full playthrough and record it and then they'll look at it and take pieces and move them around, and see how or if it would solve issues with the pacing and stuff like that. It was awesome for the team because at first we had some work to do around that. But with that edit it was a very clear direction of "okay this is what we need to do."

Dozois: It's much easier for me to move an hour around in a video edit in the video edit than the actual game. They're like, "look how awesome it is!" and you're like "yeah but there's technical things, and good luck, lets do that."

That was a useful tool for us, as Fleur was saying. We did a lot of that the last six months -- just looking at pacing, seeing if this is too long, is this too short. Because you can really focus on a scene or a moment, but when you watch it [in longer increments, you can] see how it feels, the momentum of those scenes going forward. And that really helped us make small edits and little changes. It was a lot of fun.

Marty: The thing is there are so many factors at play here. First of all, we're doing a third [game] and a lot of the people on the team have been working on the franchise since 2015. A lot of them have been working on Rise. So they know how to make a Tomb Raider game. We've been able to have things on screen very, very early.

And that's also because the foundations of the tech were here, because it was the same engine that we used for the previous games. Even if we made a lot of improvements to it and a lot of developments over time, all we needed to make the game, at least a prototype of it, was here from the start.

It really allowed the team to iterate on screen super quickly. And the mindset that the whole team has had all the way through the project is, "ok lets get stuff on screen, lets iterate fast, lets fail early too if we need to."

Dozois: Fail early, fail often, and get it out of the way.

For more developer interviews and editorial roundtables, be sure to follow the Gamasutra Twitch channel.