Red Dead Redemption 2: A deep dive into Rockstar’s game design

Rob Nelson has spent 15 years at Rockstar Games, working on games such as The Warriors, the Grand Theft Auto series, and both Red Dead Redemption and Red Dead Redemption 2. For the past 3.5 years, he has been co-head of the Rockstar studio in Edinburgh, Scotland.

But he’s really been living in the Wild West for the better part of a decade.

Nelson is one of the people who brought us Red Dead Redemption 2, one of the most ambitious story-driven open world games ever made. Red Dead Redemption 2 was in development for eight years. More than 2,000 artists, designers, writers, programmers, and others contributed to it. Analysts believe it could hit 20 million copies sold by the end of December, a level of sales that would make it an enduring hit. Only a handful of games like this will ever get made, in my assessment. It set a new bar that everyone else will have to deal with.

I played all its 105 missions, taking more than 50 hours, and I have spent some time in the recently launched Red Dead Online. The world is incredibly immersive. After all that, I had some questions for Rockstar, and I interviewed Nelson for a good chunk of time.

We didn’t talk about allegations about forced crunch time, as Rockstar has addressed this elsewhere, and I tried to avoid spoilers in our discussions. So we avoided being specific about the ending and how things turn out for memorable characters like Arthur Morgan, who is the enforcer of the Dutch Van der Linde gang and the story’s main character.

Nelson has heard a lot of complaints about the design, such as the long horse rides to get to missions, weird glitches and the length of the story. I also remember how alive the game felt when I went off script, like when I accidentally shot a dog. This was his chance to explain why Rockstar made some very big decisions in the story, the characters, the open world, and the overall design of the game.

Among the gems from our conversation: Rockstar once cast Red Dead Redemption 2’s main design as a procedural world (one that the game’s software generates), with stories driven by the open world and emergent behavior. But the company trashed it because it just didn’t work.

Here’s an edited transcript of our interview.

Remembering the first Red Dead Redemption

GamesBeat: Can you tell me about yourself and your roles on Red Dead Redemption 2?

Rob Nelson: I’ve been at Rockstar for quite a long time, since 2003. I started at the Toronto studio and worked there for about five years as a producer on the games we worked on at that studio. Then I moved to New York and worked as an art director, as well as a producer capacity on all the games we put out up until 2015, like the last Red Dead and Grand Theft Auto V. In 2016 I moved over to Scotland full-time and I’ve been working out of the Edinburgh office. I was a producer on Red Dead 2.

GamesBeat: Were you on Red Dead Redemption 2 at the beginning?

Nelson: Yes, but not directly, not every day. I’ve been here all along, but early on, I was focused on finishing some other things. I’ve been on Red Dead 2 full-time for three-and-a-half-to-four years. That’s the way we work. We roll on-and-off projects as necessary.

GamesBeat: I think we could use some reminder of what the original Red Dead Redemption of 2010 was like, and what it was capable of. It was built around that single hero. Do you recall some things about what you couldn’t do in the first game that you can do now in Red Dead 2?

Nelson: Absolutely. A big challenge for us, and a big part of my job, is to help guide something that’s made by thousands of people over several years and make it seem like it’s made by one hand, or one voice. With Red Dead 2, we had an interesting internal challenge. With our followup to Grand Theft Auto V — this was all the same people who worked on that, with the same goals — we had to follow that properly and have it be an evolution of that game. But it was also the sequel to a game that resonated with people who liked it in the last Red Dead.

For me, I did something that I’d never had to do before. With the GTA games, they’re installments — they’re not direct sequels. They’re new stories every time. But Red Dead 2 was a direct prequel that related back to the last game. In my office we had the last development build of the last Red Dead up, and we had it right beside where we were developing the new one. It’s a weird thing, because a lot of the time, we don’t play the games we make much after they come out. But we were conscious of how the last Red Dead existed in people’s minds. People look back on games that they love with those rose-tinted spectacles.

We made the decision, at least the group that I worked with, to delve into the old game again and try to discern what the essence of it was, the feeling of it. The feeling that people came away with, I think, was the feeling of being able to exist in a wilderness and to survive. To just be. That was the thing that ultimately — along with the story and the tone of the game — that resonated with people. It was the sensation of being in a world that’s disappeared, an America that’s gone.

When we looked at it, it was quite limited. You could kill everything and “skin” it, but all you did with those things was sell them. Then there were some limited things you could do with the money. But the feeling of being able to do that stuff was what was there. We took those ideas and ensured that, even though now we’ve decided to tell the story of being in this game, and living and working in this game, we still wanted you to be able to go out and exist in the wilderness, to be alone, to have space, to survive off the land. Hunting and crafting and camping and exploring, all of those things, were things we needed to take to a new level, because they felt like they were there before. We wanted to really make sure they were there this time.

Leveling up with Red Dead Redemption 2

GamesBeat: It seemed like one key was just fully animating things that happened. I’ve noticed that in Battlefield as well. When you heal somebody, they make you see the medic moving his hands around and wrapping you up. It’s a fully animated scene, where before you might have just taken a pill or something.

Nelson: That sort of stuff was all about the pace of the game, and the world that we wanted to create, a world that you traverse on foot or horseback. We wanted things to mean something. We said this while we were developing the game. Some of the ideas we wanted to explore were somewhat risky, because it’s interesting to change things that nobody’s asked you to change, or that nobody expects you to change. Taking something like your loadout, with the weapons on your character, and saying nope, you’re going to have the guns you have.

That wasn’t a thing that was broken in the last game. It was slightly different from the way we approached GTA, where it’s an invisible inventory. In the last Red Dead, you had a rifle on your back and a pistol in your holster, but that represented all of your weapons. We wanted to increase the bond you had with your horse, and one of the ways — we didn’t want to make it super-annoying — but if you leave your camp, you only have these guns. We could make the horse your “magic” inventory, and that would make you want to keep that horse around. That would make you think about what you want to take with you.

Leaving that stuff — then that creates a problem with your magic horse, that’s going to come from any distance no matter what. The horse couldn’t perish. The horse couldn’t get lost. If you lost it, it was slightly annoying. We had the same ideas for skinning animals. We didn’t want to cut the camera into a three-second scene. We wanted to keep that all in the camera, in the world, in the situation. Any time we would cut it to do a little scene like that, I feel like it kind of knocks you out of the experience a little bit. It reminds you that you’re not this person in the world.

We wanted to explore these ideas, but not take them too far. You don’t want to wait too long to watch this sequence of an animal being skinned. We wanted to reduce that time as much as we could while still making it feel “realistic.” Once you get that vibe or that tone you’re shooting for, when you say, “Oh, it feels cool to do that, to skin that rabbit or skin that deer,” everything else in the world needs to live up to that. Or it’s going to be jarring and feel out of place.

GamesBeat: You get to blend the cinematic scenes and the gameplay a little?

Nelson That’s something we’re always evolving from game to game. The last Red Dead, we were really trying to get away from the fade to black between gameplay and cutscenes. We still had intro title cards for the missions in the last Red Dead, but we’d walk out of the cutscene and cut right to the player walking away. That was the first time we were able to do that, blending cutscenes and missions. We took that much further with Max Payne and then even further with GTA V. Hopefully, Red Dead 2 took it as far as we’ve ever taken it.

We just keep blurring the lines between what’s a cinematic, what’s a mission, what’s a cutscene, what’s gameplay, what is and isn’t interactive. We want to do that as much as possible. Those are the barriers where, when you recognize the convention, you remember that you’re playing a game.

Setting a high bar

GamesBeat: How did you think about the scope of the game? I can imagine conversations like, “Well, let’s make it two or three times bigger than the original Red Dead. We can get 2,000 people to work on it for eight years.”

Nelson: [Laughs] That’s not quite the way it works.

Nobody starts out thinking they’ll work for eight years. We just kind of go. Once you hit this tipping point in development, where you have a world and you have a narrative that you roughly think are going to work, and then you have the elements, the mechanics, the things that are going to support the narrative — the game tells you what it needs and what it doesn’t need, what’s superfluous and what’s not.

Once the team was full steam, really rolling and making the thing–I guess there were things we could have cut to save a bit of time. But really, for the whole thing to feel resolved as a whole piece of work, it needed everything that we put in it. You just have to finish it.

GamesBeat: I imagine there were decisions that made it bigger in scope, like going to multiple protagonists, telling the story of the whole gang.

Nelson: That’s definitely true, but that was a decision that was undertaken from the start. We had a number of reasons. We wanted to make it feel different from the last game. We didn’t want to do another game about just another lone outlaw making their way through the world. We wanted to make a game that felt undeniably Red Dead, but with a totally different set of parameters. Hopefully, people feel we succeeded. It just wouldn’t have been interesting without that.

We loved the idea of Michael and the family needing to get by, and that these characters — not only Michael, but this family that you’d see — they also existed outside of these missions. Again, to serve the purpose of blurring that line between the narrative and the open world. The world is the story, the story and everything else.

By committing to the camp and the big gang — it was hard. But we were able to take all the exposition and extra story stuff about these people and their relationships, if you want to stop and engage with it. And if you don’t you can walk past it and you don’t have to engage. Everyone is going to want to hear something about it and feel more invested — hopefully — in these characters because they’re hearing stuff that they weren’t paying attention to a minute ago. That’s a really interesting and exciting ability that games have and other mediums don’t: that optional, additional narrative that you can choose to engage with or skip over.

GamesBeat: I really liked this game. I played all 105 missions. I had a colleague who felt like he wanted it to be more like Hitman 2, where you can go and do anything, tackle your target through many kinds of means. It seems to me that approach is not what you wanted to do or tried to do. Do you have a view on how you balanced this sort of directed story versus the openness of the world and the missions?

Nelson: I think all of us feel like we want to approach any situation, anything, in any way we want and have it be credible. But that’s a big, big challenge. To do that and have it feel — it’s a big world and a big story, and I think it needed to be a big world and a big story for what we set out to do. But yes, we explored a lot of different avenues early in development, like more procedural approaches to things. Hey, it’d be great if this camp could totally grow and you could hire people from towns and come back and add to your camp. We explored, at one point, if you could take anybody fishing that you wanted at any time.

But what that ended up doing was a very procedural-feeling game. You’d write a bit of dialogue, beta a bit of dialogue, and go and make these AI-type behaviors. It didn’t feel like you’re on a thing with Pearson and Bill, or you’re on a thing with Javier and John. It felt like you were on a thing with AI that just looked like those people, but they didn’t behave like those people.

Unfortunately there’s no procedural system yet that we’re happy enough with to make the worlds we make. Our worlds are handmade. Our artists will use certain procedural tools, but they’re all curated by the artists. It’s the same for the content we make. For it to make you feel anything, it has to be made by humans. It has to be written and designs and shot and acted and processed and put into place very carefully. For things that happen in the world, we have to very selectively know when they’re going to happen.

It would be great if this was all open, but people have to make this stuff happen at some point. It has to be scripted so that it all feels right. I don’t think there are procedural tools that will make it feel real.

GamesBeat: I find that a lot of open world games lose me. I don’t know where the next story thread is, because it’s wide open, and then I don’t care if there’s another mission sometime. In this game, I was actually worried about these characters and what’s going to happen to Arthur.

Nelson: That’s very nice to hear. I’m with you on that. We want as much freedom and openness in our games as possible. That’s the thing that drives us. But it’s a double-edged sword where you’re trying to make something where it feels like people are invested in the characters and situations. You have to pace these things. We always have to balance choice and agency and narrative and investment. It’s a great problem to have and it’s great stuff to explore. We have the same conversations internally when we’re making this stuff. But we still have to make it.

Did it have to be 50 hours long?

GamesBeat: I wonder about the length. How you could sustain the player’s interest through something that’s so long. It took me more than 50 hours. I was very happy with a game like The Last of Us, which took me about 25 hours or so. I felt there was a lot there, a lot of variety. I don’t know that I ever wished it was longer. With yours, and those 105 missions, I wondered if you had to take a sharp knife to the missions.

“We can’t squeeze this one in. We have to cut it.” But there are so many oddball things in it, like the alligator quest or the drunkenness mission. There’s this whole epilogue that you don’t expect. From your point of view around the storytelling, how did you feel about making something that was so much longer than other big narrative games?

Nelson: Well, we didn’t compare ourselves to other narrative games. I don’t even know how much we compared ourselves to ourselves. We certainly went through an editing process in this game and cut a lot of missions, mostly close to the end. Ultimately we made the decisions we made after discussing them closely.

We pulled a lot of the missions out of the main flow and made them optional. We’re always doing that, pulling missions on and off the main flow. In this game we did that more aggressively than we ever have, but it had nothing to do with the production value of those missions. We did it based purely on Arthur as a character and his potential for growth.

We felt like any mission that Arthur would do as a character based on his job and his role in the gang, which is to support Dutch and support his family, that’s what you need to do, because that’s Arthur’s job and what sets him up as a character. Any mission that afforded the player some kind of opportunity for growth or reflection, we pulled off and made optional, so that if we had any sort of revelation that made him feel different or changed, the player didn’t feel like, “Well, I didn’t want my Arthur to do that.” We tried to do that as much as possible.

Things like the Lenny mission, drinking with Lenny — first of all, we didn’t want it to be one-note. We didn’t want it to be all brutal. We didn’t want it to be all massive shootouts and robberies. We wanted to show that this group of people enjoyed being together. That’s why we put in the parties that weren’t missions. They were just in-game, and you could walk around and see that these people love each other and care for each other. The Lenny mission, I think some people might have felt it was random, but we thought it was very important.

We also have a lot of people in the game. You need to spend time with all of them to get to know them. You might not engage with any of the stuff that’s going on in the camp as you walk through, so we needed to make sure that there was at least one or two things with every character so that you got to know them. The Lenny one was important to us. It makes what happens with Lenny mean something to you. If you didn’t spend any time with him — if we didn’t make you spend any time with him — you’d feel like, “Eh, I don’t care about that guy.”

That’s an example of why we made that decision. But we talked about all of these things all the way through. Anything we left in there, we left in for reasons.

GamesBeat: The one where my closet game designer comes out and I get all fiery was the Romeo and Juliet mission. That’s one I thought shouldn’t have been optional, where you resolve the feuding families —

Nelson: Yeah, with Penelope.

GamesBeat: I thought, “Wait, how can I leave this area without finishing that?”

Nelson: That’s a good example. It’s optional, but it wasn’t optional for a while. The question we had was: Why would Arthur help them? Your Arthur might want to help them and see that they get away OK, but my Arthur might not care. It doesn’t serve his purpose. My Arthur’s purpose is to get away with the money and take care of himself.

GamesBeat: You must have had the wolf dreams, then. I had the deer dreams.

Nelson: [Laughs] Yeah. Those are the kinds of conversations we have, and we want people to be having those conversations. We fight over all of these things. “This one has to be mandatory. This one has to be optional.” But it was really fun. Some of my favorite experiences in the game are options within options within options. Some of the best sequences and things you can get are very deep within — if you don’t help this person, and then go and help this person, then a different person will appear much later in the game and you’ll get this heart-wrenching revelation from Arthur. But it only happens if you do this stuff.

I think that stuff would be less meaningful if people didn’t care about helping the people who got them there. I’m not saying that we got it exactly right, but that’s definitely the sort of stuff we’re thinking about, and that we want people to think about when they’re playing the game.

Why was it a prequel?

GamesBeat: Do you ever regretted doing this as a prequel as opposed to a sequel. Would you have more freedom, in a sequel, with how the story could turn out?

Nelson: I don’t think we ever regretted it. I personally remember begging Dan [Houser], let it be a prequel where we play in the gang from the last game. [Laughs] That’s just my personal opinion. But I wanted it to be in that gang doing that stuff. I suppose there are some limitations because you know how certain stories are going to end, things like that. But it never really bothered us while we were making it.

GamesBeat: As a player, it gave me this sense of impending doom.

Nelson: Yeah, I know. That’s a strange case. Certainly there were a lot of ideas like that in this game that we thought were worth exploring, that we thought were interesting. Again, this game wasn’t — it was different from a lot of games, I think, and a lot of games we’ve made. Instead of starting out as this unassuming, weak person who doesn’t know many people, who has to go out in the world and build up their strength, you start as a very strong person who’s very self-assured and confident in his place in the world. And then that falls apart.

Games are aspirational a lot of the time. You’re chasing something. To make a game where the carrot isn’t always obvious — we started out with this perilous situation where we were trying to teach you how to do a lot of things, but then we pop you out into the open world, and we very consciously structured the early chapters to be quite open. You don’t have a gun-to-your-head objective. The objective is to lay low and rebuild your savings, to make money.

We wanted to take all of those things that we felt were what people felt from the last Red Dead — living in the wilderness — and then, how do you do that? Surviving on your wits and making money. The gang needs money. We tried to encourage you, even though some players still rush through it, to go out and engage with the world and make some money. Bring it back to the gang if you want, like you’ve been asked to, or you can enjoy life as it is now, because life doesn’t last forever. You don’t know that necessarily when you play in those early chapters, but we really tried to make it feel like this is open, and it’s about the gang having fun and enjoying life as it happens. Because you never know.

Those were really fun ideas to explore. We didn’t know how they would resonate with people. People might just say, “What’s my goal in this thing?” But we definitely had fun with that.

GamesBeat: How do you feel about some of the reactions from players? What’s caught your attention on how people are reacting to it and forming their own opinions?

Nelson: First of all, we’re relieved that people generally like it. [Laughs]

It’s interesting. We did make decisions that we really had to commit to. We couldn’t do a lot of the things we did halfway, like the pacing. Some of the real living stuff — we knew all of those things had potential issues when we decided to pursue them.

It’s interesting to see how people are talking about the pace of it. I like that we’re in a kind of dialogue with people. We’re seeing the exact same conversations we had internally, when we would fight it out over different things being too long. Was the beginning too long? But if we completely removed all of that stuff, I feel like the game wouldn’t mean as much. Things have to pause sometimes.

Fast travel, please

GamesBeat: My thing: I wished for faster fast travel. I imagine there’s an argument against it.

Nelson: Some of these things are very deliberate. I would certainly say that the way we handled a lot of that was deliberate. It’s a big game, and a lot of it depends on where you decide to put things. Some decisions are more deliberate than others. We react as much as we can in the next games we make.

GamesBeat: In some ways you already have that. You have Red Dead Online, and you can develop that further and change some things. How should people look at Red Dead Online versus some of those decisions you made?

Nelson: I think we look at it as a work in progress. The nice thing about the online games, and certainly with GTA Online — it took us a couple of years for us to find our feet with it. People were patient and they played and talked to us and we listened. We had to develop a more robust set of tools to deliver heists that lived up to what we thought people expected, and once we did that we were able to develop content that was a lot more sophisticated than what we’d previously offered.

The people working on this are trying to make it as good as it can be, but we’re a couple of thousand people, and once you put it out in the wild, if we’re lucky we hear back from a few million people. [Laughs] We adjust where we can. With the single-player game, it’s an experience that we’ve put out and there it is, but with the online game it’s more like a dialogue. We just want the online people to be patient and have fun with it and talk about what they like and what they don’t. We want to make what people want.

GamesBeat: It was nice to see the New York Times talk about it like a work of art.

Nelson: That’s very humbling, and overwhelming. Like I said, it’s a relief. What I think we’re really interested in the kind of conversation that we’re having right now is the nitty-gritty. We want to get better and keep making better and better things. I like to read when people have feedback and think about how it can apply, knowing what we know and considering the decisions we made.

Why did you design …

GamesBeat: When [Rockstar PR] read my review, he told me about a couple of things that I think I should have picked up on. Maybe they just weren’t hammered home to me quite so well. One was how the drunken mission takes place before you rescue Micah, but PR pointed out that Arthur never really liked Micah. He was probably perfectly happy to let him hang. This delay in going out to rescue somebody made some sense.

Nelson: We moved missions around like that. It wasn’t always there. But we thought that was a good time to go and show you a different side of Arthur, so that he wasn’t too one-note. He had a fun side, a more crazy side, or whatever you want to call it. We wanted him to feel like a fully developed person. And we put it there to make sure he wasn’t rushing out to get somebody that he didn’t really care about.

GamesBeat: The other one, right after the bank heist I had all this money. For the first time in the game, the gang had a lot of money, and I had a lot of money. But then I looked at the gang and realized they didn’t have a lot of money. Where did that money all go? And PR pointed out that, well, maybe Dutch is hiding a lot of that.

Nelson: We had three economies in the game. We had Arthur’s money, and then the camp fund to which you can contribute, but that’s just the gang’s operational money. The gang’s savings, those coffers, they build up over time as the gang does every job and they restructure the nitty-gritty stuff. If you look in the ledger to contribute, the gang is rebuilding their coffers, and that money comes into play [later]. The gang fund may be quite low, but that’s just for buying the gang’s supplies.

GamesBeat: As far as some of the surprising things people have noticed that you can do, I think the IGN review showed the horse’s testicles could shrink in cold weather? I wonder if you’ve seen people discover some things that have surprised you too.

Nelson: Of course. To me, the horse balls, what everyone talks about — that was just one person. I think he was in the Toronto studio. He just decided to do that one day. There wasn’t any big discussion about how we were going to do that from a high level. [Laughs] Everybody has a passion for their work that they apply in that way. If it’s going to support the game feeling believable, everyone puts in the time to try to do that. There are tons of things like that.

Working on the game, I was surprised every week by new things. I see people talking about the game and having similar reactions to what I had when we were playing. Riding along and seeing this little critter follow my horse. I thought it was a bug. I aimed at it, because my rifle was still raised, which indicated that it was alive. I went over to it and it said, “Skin possum.” So I went over to skinning, and it jumped up and hissed and ran away. I just laughed. That was another little group of people who said, “Let’s make the possums play possum and have some fun with the players.”

Tons of stuff like that. We often don’t all know everything that’s in the game. That’s the fun thing about making a game this big, but it’s a bit scary. Nobody knows everything that’s in there until very close to the end. We’re playing the game and saying, “Oh, I didn’t know you could do that.”

GamesBeat: I don’t want to get into the ending that much or spoil anything at all, but I do wonder why it ends the way it does. My guess is there’s this desire to have an explicit connection between this game and the first game. But do you have anything that can direct people to why it ends this way?

Nelson: I think that’s right. I think it has to, by the nature of being a prequel — it has to connect. Also, the gang had to be big. If it was just Bill, Javier, Dutch, and John, and maybe two other dudes, why didn’t someone just arrest all these guys? The gang had to be big and it had to be fluid. People had to go in and out. That’s also the type of story we wanted to tell.

I just think it had to end that way. It was the right ending for the game. It had to be a sort of spiritual sequel to the last game as well. There’s a sadness in the last game. We wanted it to be different from the last Red Dead mechanically, but feel absolutely familiar. I think that’s why it had to end the way it ended.

GamesBeat: There was an interesting choice the characters make, to either do something relatively boring or to live the outlaw life. I thought it was interesting for a video game to present the choice in a realistic way. Do you love your family, or do you love being an outlaw? In video games the presumption is that the violent life is the answer. There’s a metatextual thing here where it’s saying the opposite, as the message of the game.

Nelson: Absolutely. That goes into a lot of the decisions we made. If we wanted to get that across in a game, we needed to do it on this scale. If we were going to make a game about these people who you get to know and care about–we’re saying Arthur cares about these people, but how do we make you care about these people? We’re dropping you into the middle of his life where he knows all these people already. These relationships are already formed. You need to learn about those relationships and decide whether you like these people or not.

We maybe try to get you to like certain ones and not like certain others, but you needed to be able to do that yourself as a player. Not in cutscenes, not in watching it passively. Am I going to go talk to Tilly now? Am I going to talk to Susan? Am I going to go help Reverend Swanson? Do I like Pearson? Whatever it is, we needed to give you the option. If you have a connection with the characters, we need to give you the ability to build that interactively within the game.

But then you also have to leave the camp and go out in the world and have experiences with them in the world. You need options for interacting with them beyond the confines of the camp. We had to develop this social interaction system. I really wanted to do that. What does it mean to give you options besides just aiming a weapon at people? All of a sudden, the decisions you make as a player that you maybe made somewhat thoughtlessly before — it’s really interesting stuff.

These are things we thought about all the time, and hopefully they’re all funneled in some way back into the narrative that we’re telling. Thinking about your life, thinking about the way you are, thinking about the way you behave toward people. Those are ideas we were really interested in.