Join gaming leaders online at GamesBeat Summit Next this upcoming November 9-10. Learn more about what comes next. 

Amazon spent five years developing Crucible, the free-to-play team shooter game that launches on the PC today.

The competitive multiplayer game lets players choose from a diverse roster of aliens, humans, and robots to battle on the planet Crucible, where they can collect the resource Essence while avoiding death from fellow hunters and the environment.

Louis Castle, the cofounder of Westwood Studios and co-creator of games such as Command & Conquer, joined Amazon in 2017, and he runs the Relentless Studios game studio at Amazon in Seattle. Relentless made Crucible, which has elements of multiplayer games such as Blizzard’s Overwatch, Riot’s Valorant, and even Call of Duty.

The game debuts with three models: Heart of the Hives (4-vs.-4), where players battle big bosses and capture their hearts to win; Alpha Hunters, eight teams of two players each in a final team standing match; and Harvester Command, where players capture and hold spots on a map. I talked with Castle about how Amazon made the game and the thinking behind its design. We’ll find out soon if those decisions resonate with players who are eager to play new games while in lockdown.


Three top investment pros open up about what it takes to get your video game funded.

Watch On Demand

The game was originally scheduled to come out in March, but Amazon decided to push the launch back to May 20 so the team could deal with the pandemic’s disruption as it finished work, Castle said. He also noted that the game was made from the beginning to be watchable, and all eyes are going to be on this first major PC game release from Amazon.

Here’s an edited transcript of our interview.

GamesBeat: I was chatting with a developer who said this project started about four years ago?

Castle: It’s a little difficult to pin it down. The game started about five years ago, but it’s gone through some pretty big changes. It was announced before I joined the company at a TwitchCon in September 2016. When I came onboard, we had a playable version, and we had a community group already put together, 100 people or so. It was quite a good number. From the very beginnings, when it was barely stood up as a playable prototype, it’s always had community guidance.

When I came to Amazon it was the project that resonated with me as being customer-obsessed, the way Amazon likes to be. It’s one of the things that attracted me to Crucible. All the leadership is tied into community. It’s been built hand in hand with a bunch of advisers.

Above: Louis Castle runs Relentless Studios at Amazon.

Image Credit: Amazon

GamesBeat: How big would you say the team was when you got there? Were you going in just to run the game, or more to manage a larger studio?

Castle: The quick answer is we’re not supposed to give exact sizes of teams. But it’s smaller than teams that would be working on competitive titles in the same space. It’s nowhere near some of the larger teams I’ve worked with in my career. It’s a relatively small, scrappy team.

GamesBeat: Was it a prototyping team or a production team at that point?

Castle: When I came onboard, there were two groups at Amazon. Amazon has a principle of single-threaded leaders, so there were two single-threaded leaders that had development teams under them. When I came on board, one of those groups was being redistributed to start on The Grand Tour and to do some stuff for Lumberyard. The other one was Crucible. It was a reasonably sized, adequate — it was bigger than a prototype team. It was a full-size development team. They had a game that was playable. I had a chance to play it in the lab, and it was quite fun. There were a number of concerns at that time, which have of course been remediated since then, but right from the get-go the game has always been a lot of fun to play.

GamesBeat: How did this get going?

Castle: I was called by my friend Rich Hilleman [an Amazon game executive and former chief creative director at Electronic Arts]. I’d known Rich since I was in high school, so it’s been a long time. He said, “Hey, you should talk to Mike Frazzini [vice president of games at Amazon].” It just so happened that I was contracting for companies like Kixeye and Kabam and a bunch of others. Everything was closing out at the same time, and I was looking for new work. I talked to Mike, and there was a lot of talk about the kinds of roles I could do at Amazon.

He asked me point-blank, “What would you like to do? What would be your favorite thing to do?” I said, “Well, I’ve been in the trenches for a few years now, working on actually touching the code again. I like being closer to the products. It’d be great to be a head of studio again. I like working with the teams directly.” He said, “We can work that out.” I went in and did the interviewing at Amazon. I love the teams I’m working with. I love the work. I’m super happy to be doing it.

Mike Frazzini shares internally with our group the original pitch to the senior team leaders at Amazon, the same way that Jeff Bezos shares that letter to the shareholders. Just a reminder of the north star. We have a real reason to be here. Amazon’s games, for ages — as a company, AWS, is used by about 80 percent of the top game companies in the world. All the features, all the services we offer. We’re also selling games too as a retailer. It’s not that big of a jump to say that we should not just be offering up the stack tools, but also down the stack tools. That’s how Lumberyard got built, and there are lots of cool things still coming from game tech.

We’re all in on games in lots of ways. One of the things you should be doing if you’re making products for customers is knowing you understand the customer’s position. As an internal customer, having game teams makes a lot of sense. It slowed us down in the beginning a bit, because we were trying to do an awful lot of things at once. I’m proud of how they came out and people rallied around the excitement of building something that’s truly obsessed right from the team level.

Crucible is coming on May 20 on the PC.

Above: Crucible is coming on May 20 on the PC.

Image Credit: Amazon

GamesBeat: Was it difficult learning how to use Lumberyard, using a new game engine?

Castle: Lumberyard has a lot of interesting features about it, because it’s been built with a lot of ideas around feedback from customers and things like that. We’re working very closely with that group and with other customers on unannounced products. It’s the same internally. It was nice to have a team that’s dedicated to a technology base right there next to you. It’s also caused some amount of friction, because it’s a brand-new thing.

The nice thing about Amazon is we can build whatever tools we want. We were building on Lumberyard because that made a lot of sense at the time. We may or may not be doing that. I’m sure it just depends on the project. But Lumberyard is a great tool. It’s good at many things that it does. Lots of companies, especially simulation companies that love the highly technical scale issues in Lumberyard — for game development it’s not as mature as some of the other game engines, but we’ve managed to make something that’s as competitive as anything else out there.

GamesBeat: When you came in, did you reboot the project in some way, or re-orient it?

Castle: There was something I did. I first came on board in February. By July we had a big playtest, about 250 players. It was pretty interesting that in July 2017, we had 250 people playing this in the wild. I liked the feedback from all the customers. It was clear that we were on to something exciting, and they were itching to play it.

This was before PUBG had become too big, and way before Fortnite. The game was a character-based shooter. Even back then it had some battle royal aspects to it. It had all these other features you’re seeing that are unique when you play it, the attention to detail in the map and the placement of objects, the MOBA aspect of characters leveling up during the match. All this stuff that you don’t see in other games in the genre. The reason Crucible predates all those games — it has not been developed in parallel to the industry finding all the fun in these competitive games.

I felt like we needed to do a big technology upgrade on our networking infrastructure. Some of the team members convinced me that we needed to change some of the core tech around entity component systems, so that we would be able to have better scalability. For the nontechnical people out there, it’s more about making this a fair place to play. It’s performant over many kinds of connections. We could update the graphics quite a bit with some dynamic global illumination. Stuff that’s in Unreal Engine 5 now, that’s in Crucible. It’s not quite exactly what they’re doing, but it gets the same effect.

There was a visual upgrade. Online play had to be better. That would also require us to rewrite the game. The previous version of the game was live for the full year-plus that it took to rewrite the game engine, just so we could keep playing with our alphas and make sure that we refined the game. We have a game that’s been in testing for more than three years, more or less in the same state it’s in right now. That shows in the quality of the experience and the balance of all the characters.

Above: Ajonah is the sniper of Crucible.

Image Credit: Amazon

GamesBeat: Have you always thought about having small matches? Eight-on-eight, like a Call of Duty multiplayer match, rather than large-scale battle royal.

Castle: From the beginning, it was built to be watchable. It’s exciting to watch anybody playing Crucible. You can jump in, randomly pick a character, watch them, and it’s instantly exciting. The map is relatively small compared to the big open world maps in some of the battle royals out there. The team size, the number of players, is intentionally small, so the match resolves more quickly.

More important, the things that [matter] to you are happening on your screen most of the time. I don’t remember the last measurement, but we do measure how much time, if you’re watching any random character in the match, things are exciting on the screen. If you jump into a battle royal and randomly follow someone, you might see them hiding in a closet for 20 minutes, which is not very interesting. Or they run around like a maniac opening crates trying to find the weapon that will keep them from getting killed instantly. Some games you’re just shot in the head from a mile away and you don’t know what happened.

None of those things happen in Crucible. It’s designed around making sure every game session is exciting and interesting to watch and play.

GamesBeat: As far as what I might be reminded of here, there’s Overwatch in some of the way it looks. It resembles Call of Duty’s Domination mode in the way that the Harvester mode plays, I think. There’s five locations instead of three. But it feels like it’s mixing and matching some things that people are going to like.

Castle: Crucible was built for competitive play, built to be watchable. We have a marketing message that sums it all up, level and adapt. We’re focused on a game experience that we think is exciting and interesting to play competitively with your friends. Each one of the modes is designed around a different set of features that we liked in other games, but it also has its own layers. It’s not intended to be derivative, a direct copy of any other game, which is why people struggle a bit. You can’t really say it’s Game X plus this, because it’s not. It has elements from lots of different genres.

That’s quite intentional. When the game was being designed and built, it was important to us to think about what’s going to make the best experience possible for the kind of thing we were trying to deliver. We liberally borrowed ideas from other games. But we didn’t follow any particular game. It doesn’t feel exactly like any one game. It’s not League of Legends as a third-person shooter. It’s not battle royal. It’s not that easy to pigeonhole. It’s a unique expression of elements from all sorts of games.

GamesBeat: The movement seems slower than some other games. Some of the characters are a little lumbering. Was there some intention to slow down the action in some ways?

Castle: I wouldn’t say slowing down the action. We played with all sorts of different speeds. The game could be as fast or slow as we wanted. We found a nice balance. Another nice thing about Crucible that’s interesting, if you like brawlers, if you like melee, there’s a character for you. If you like [area of effect], there’s a character for you. If you want to be a builder instead of doing headshots, there’s a character for you. Even though there’s only 10 characters at launch, every one of those characters is unique.

We’ve mixed genres up. There’s a sniper kind of character that’s similar to Widowmaker, has some elements, but not the same. That character has a grappling hook to swing all over the match. We have flying characters. It’s not that everyone is slow. It depends on which character you’re playing If you’re playing Earl, he’s a big space trucker. He tromps around on the match, and that’s intentional. But what was also very intentional is that characters don’t have one-hit kills. We’re careful about making sure that you at least have a moment to react to an attack. You’re not going to just wander in and die instantly. That’s intentional.

When you talk about slowing down the game, we weren’t trying to slow down the feeling of reacting. In fact, we’re tuning up the speed of movement for a couple of characters so people have something that moves really peppy. But at the same time, we don’t want people to die instantly. There’s definitely some attention around that. If you play the game poorly and your team doesn’t level up and you’re level three when the other team is seven and you have badly matched characters, sure, you might die very quickly, but that’s the exception, not the norm.

Above: Crucible has some colorful fauna.

Image Credit: Amazon

GamesBeat: There’s no loot. There are powerups, pickups, but you didn’t want to litter the field with things to grab?

Castle: There’s lots of stuff to interact with on the map. That may not be as evident yet, because it takes several times to play through the game before you start to learn the map. It’s quite large for the number of players and the complexity level. It’s a rich and deep experience, with very strategic placement. There are plants that can make you temporarily invisible to sensors, and other ones that heal. All of those are deliberately placed and designed carefully. Their placement is very nuanced.

What it does is it creates knowledge. As you learn the map and learn the way the world is put together, you start to become better at using the environment to your advantage. That’s very intentional, because it gives depth to the game. There are lots of things to pick up and interact with. In fact, you don’t have to be in combat to be useful. You can interact with a device that helps you get more essence and level your team up, or you might be using something else to guard a particular area so that other players won’t be able to use the effect.

We also have the planetary events. There are dozens of those right off the bat, and many more coming. They randomly occur during the match. They’re set up at the beginning to show you what’s coming. You don’t know exactly when they’re going to occur, but you do have a pretty good idea if you’ve learned them a few times. There’s a damage amplifier, for example. There’s two of those. If the other team gets both damage amplifiers, they have an extreme advantage in the late game. It totally changes your strategy. Maybe your team thought they would play the long game and let the other team get some wins early on, but then capture three points at the end. That changes dramatically if all of a sudden there’s a damage amplifier and the other team gets it.

That’s one example of dozens, how they interact with the strategy. That’s the adapt part. No plan survives contact with the enemy. You can go in with a plan and try to execute that plan, but you have to be willing to adapt to the situation.

GamesBeat: I noticed some of the matches can last quite a long time. You have a wide variation there. If you die quickly in the battle royal part, you’re out.

Castle: Yeah, you can watch your teammates or go back into another game or whatever. There’s only 16 players. It’s much faster than a 150-person game, 100-person game. But there’s still-the variation between a fast match and a long match in Crucible is pretty dramatic. It might be as short as 10 minutes, or even less if a team is really on fire. They’re typically 15-ish, right around that range. I don’t remember our average right now. But that’s another thing the team watches very carefully.

We look at both the median and the mean. The median match is the most important to us. What’s the typical experience? There will be extremes. Things like Harvester Command can last a long time, because people can go back and forth and not score enough points. Heart of the Hives typically doesn’t last that long. It’s the first person to three hive hearts. If you get down to two versus two and you’re doing the last one, that can be a much longer match than if someone just sweeps all three and does a hat trick. There are variations, but whether it’s short or long, it’s never boring.

GamesBeat: It feels like a good time to get it out. Everyone’s stuck at home. They need some games to play.

Castle: We actually delayed our launch slightly, which is well-covered out there. We were looking at the launch date, and it was right around the time that things were getting dire in Washington. We were concerned that our team would have to be dealing with, first of all, moving from working at the office to working from home, and then shipping a game, and trying to take care of their kids. Maybe a family member gets sick. It was just too many things stacked on top of each other that could disrupt the health and happiness of our team.

We pushed the date out a little while, just to put it out to where we thought, “OK, we’ll have time to absorb what it means to work from home. We’ll get into our routines.” We planned all our schedules to make sure we have the right amount of content. I’m glad we did it. I was anxious to ship the game, same as everybody on the team. But upon reflection, it was the right idea, to give us just a bit more time. We’re still right in the thick of COVID-19, of course. Nobody could have predicted that. It’s a good time to launch a game. Crucible is certainly ready to be out there.

Above: Summer is a flamethrowing hunter in Crucible.

Image Credit: Amazon

GamesBeat: How does the making-money part work here again?

Castle: Amazon is customer-obsessed, of course. First and foremost, we sell cosmetic items, just like a lot of other games in the character shooter genre. The game is free to play. That’s done so that as many people as possible can play it. Since it is a cooperative team-based game, we want a lot of players out there. Otherwise it can be hard to find the right skill mix. We do have matchmaking. That’s important because that’s one of the other big problems with battle royal. There’s no matchmaking whatsoever. As a newbie, you’re in there with experienced players and you have no chance. Crucible lets you play with people who are relatively at your same skill level, which is always good.

There’s never a pay-to-win element. There’s nothing you can buy that will give you an advantage over others. There are lots of ways to invest in the game to get your characters to have more flair and style. We have a battle pass system on day one. Not many games do that. We wanted to get it introduced to people right away, how you can purchase something that will help you accelerate how fast you get cosmetic items. It equally rewards people who are willing to invest a lot of time, or people who are willing to invest some money and a bit of time. You can’t buy your way through content, so if you’re just looking to pay your way to the top of the leaderboards, or buy your way to the fanciest-looking character, you can’t really do that.

GamesBeat: Can you still hide your way to victory in a game like this?

Castle: It’s hard to do in Crucible. Because the in-match leveling is how your character progresses in skills and unlocks advanced abilities, if you just sit around in the shadows, you’ll fall so far behind in the leveling that you won’t have any chance. Hiding as a strategy, camping as a strategy, there are some ways to do it. The sniper character, with Heart of the Hives as a focus point, you can hide in the shadows and wait for the other team and snipe them. That’s pretty disruptive, to shoot them and grab their essence. I wouldn’t say it’s not possible. But just hanging around and trying to avoid interaction is not a good strategy.

Your team levels as a team. In a weird way, if you do that, you will get carried by your team to some extent. But with only four people on a team in Heart of the Hives, everyone has to be pulling their weight. You don’t all have to be fighting, but you have to be doing something.

GamesBeat: Congratulations on shipping. It’s a big game to get out the door.

Castle: Our earlier game, The Grand Tour, was a big effort. We knew from the beginning we were supporting a show on a network, and there were a lot of things that would limit that as far as exposure. But I’m quite proud of what we offered for the price point and everything else. It was a cool little companion game for the show. But this is our first big triple-A game to go out there. We have New World coming in August as well, another big effort from the guys down in Irvine. It’s absolutely beautiful.

I’m quite proud of where we are with Amazon. I’m happy that we’ve taken the time to make the game as good as possible. At this point we just need customer feedback to make it even better. It’s time for people to go play it, tell us what they think, and get on the long track of adding features and refining things so that people keep having fun with it for hopefully many years to come.

GamesBeat: Is the season approach [for team shooters] the way to go?

Castle: We are actually doing seasons. That’s something we’re committed to. The first one is eight weeks. They’re roughly eight weeks. Sometimes they will be a little longer because of timing on the calendar. But we’re committed to seasons. We’ve already built a lot of content for the future. We’re trying to stay ahead of that wave so that we know exactly what’s coming and we can market it ahead of time.

We have a portion of the team that’s there for backup in case there’s something — maybe a character might be unbalanced, or we might have something wrong we need to fix right away. A portion of the team is dedicated to that, so that instead of just marching on the content treadmill, we’re trying to make the game better with features and improving the things we have.

GamesBeat: Is there anything else you’d like to talk about?

Castle: I’m super-excited to hear what people have to say. We’re a brand-new game studio, a brand-new game publisher. We’ve worked on it a long time, but we’re under no delusion. We know that the minute it goes live and we have millions of people out there playing, we’re going to find out all sorts of things we didn’t think of. We’re anxious to have people do that. We’re dedicated to making sure we respond to customers.


Above: Crucible’s Captain Mendoza is an assault rifle fighter.

Image Credit: Amazon

GamesBeat: We just had our conference with a couple of Amazon folks. Mike Frazzini spoke. We’re discussing our next one, and we’re thinking the metaverse is becoming an interesting topic again. I don’t know if you have any thoughts on where that’s going to go.

Castle: I do. I have a lot of thoughts along those lines. It’s interesting that games are becoming the likely place that the metaverse first occurs. When you look at the concerts in Fortnite, or when you think about the things happening in Minecraft — it’s interesting to me that you have games where hundreds of people in each instance, and then with dozens of those instances you can have thousands or tens of thousands or millions of people all in a shared experience. That’s where the metaverse starts.

Just like people might look at MOBAs and say — that all started with RTS games, when we realized it’s not impossible for people to do lots of things all at once. That led to the creation of MOBAs and these related esports. Similarly, we’ll look back at some point and see that those concerts on Fortnite might have been the first time we had a sense of what the metaverse was going to be. I’m still very enamored of VR and AR, especially AR in the long term. But it’s still frustratingly slow to get to where it needs to be. I’m super excited by that at some point. When you combine the kinds of experiences we’ve seen in these massive games along with the ability to be telepresent in the VR sense, that becomes powerful. I need to be able to see my friends.

GamesBeat: Epic feels like a new platform company, but it feels like the platform companies are the ones that can deliver on this, more than some of the pure game companies. I think of Facebook Horizons. It feels like a bigger effort than just games. That may be what the metaverse is. It’s not just a very game-focused world, but maybe — I don’t know. Fortnite is a contrast to that. But what you expect from the platform companies, with all the different skills they have in all these different areas, it’s an interesting competition there.

Castle: The companies that create spaces for people to get together — Facebook is certainly headed in that direction. I think you’re right. Those companies will create the opportunities. But if you think about it, Facebook became even more popular when Zynga made games. There was something for people to do while they’re there. Building a social platform in and of itself doesn’t guarantee an audience, but once you have an audience, a bunch of people doing a thing together, it’s much easier to get them to try other things.

I do think there’s going to be the marrying of people who know how to create content that’s compelling for hundreds of people — it may not be games exactly — but on top of a technology base that’s built by somebody who understands scale and platforms. That’s one of the nice advantages of being at Amazon, because we have expertise within the company across all of these verticals.


GamesBeat's creed when covering the game industry is "where passion meets business." What does this mean? We want to tell you how the news matters to you -- not just as a decision-maker at a game studio, but also as a fan of games. Whether you read our articles, listen to our podcasts, or watch our videos, GamesBeat will help you learn about the industry and enjoy engaging with it. How will you do that? Membership includes access to:
  • Newsletters, such as DeanBeat
  • The wonderful, educational, and fun speakers at our events
  • Networking opportunities
  • Special members-only interviews, chats, and "open office" events with GamesBeat staff
  • Chatting with community members, GamesBeat staff, and other guests in our Discord
  • And maybe even a fun prize or two
  • Introductions to like-minded parties
Become a member