Every night at 9 p.m., a pay phone would ring on a street corner in Mexico City. If someone picked up, they’d hear a question:
“Who is the mother of all wisdom?”
If they answered correctly — “memory” — a car would pull up. If they got in, a driver would take them on a tour of the city, telling stories about her daughters when they were young.
After an initial stretch, the passenger would realize they’re traveling in circles. The car’s radio would rewind and its clock would jump back in time. The passenger would see a woman outside on one loop, then see a younger version of her the next. The passenger would see someone who resembles the driver on one loop, then notice it was just a mannequin the next.
Eventually, the car would shake, the radio volume would escalate, and an otherworldly character would possess the driver in an all-consuming moment, after which the driver would regain consciousness and take the passenger back to the phone where they met.
Back on the street, the player would pull out their cell phone and return to the game that led them there, hunting for resources, looking for hidden objects, and trying to find other portals into this secret world.
Emphasis on “would.” Codenamed “Hamlet,” this theoretical project was a collaboration between Niantic, the tech-focused Google spinoff known for mobile sensation Pokémon Go, and Punchdrunk, the art-driven immersive theater company behind risqué Macbeth retelling Sleep No More. The two had grand visions of games that could fuse theatrical performance with modern technology and bring Punchdrunk’s storytelling to the masses, populating the world with actors and wild scenarios.
The two groups announced plans in 2020 to develop multiple games together, but they never got the first one out the door. In 2022, they stopped production on Hamlet and went their separate ways.
To find out what happened, Polygon spoke to eight team members who worked on the game at Niantic, each granted anonymity because they did not have permission to discuss it. Polygon also interviewed Niantic CEO John Hanke prior to the company’s recent restructuring, former Niantic executives Greg Borrud and Eric Gewirtz, and Punchdrunk founder and artistic director Felix Barrett.
People described Hamlet as a rare experiment that, at times, felt magical to play, but one that ran into real-world complications regarding scale, safety, and struggles to mesh the two companies’ artistic visions.
Punchdrunk’s Luddite roots
Twenty years before announcing its Niantic partnership, Punchdrunk was a small theatrical startup in London — an “almost deliberately Luddite” one, according to Barrett in a 2014 Wired Business Conference interview. “We were scared of any technology, because we wanted [our work] to be as analog as possible.”
In the early 2000s, the company made its name embracing the up-close, physical, and sometimes uncomfortable experiences people can have at a show. At Sleep No More, which debuted in its best-known form in New York in 2011, attendees wear masks, wander freely through a six-story hotel, witness a murder, get trapped in an asylum, and watch a rave-themed orgy.
Punchdrunk evolved as a company over the years, taking on different shows, divisions, and partnerships — some of them focused on new technology. At one point, Barrett said in the Wired interview, the team spent two days on an island in the “middle of nowhere” playing mobile games, researching how their mechanics could apply to Punchdrunk’s work.
Punchdrunk would go on to work on a mobile app called Silverpoint in 2015, a narrative match-3 puzzle game with a few levels that played out live in London. The app was a small project done in collaboration with Absolut, but in the Wired interview Barrett discussed the potential for something bigger.
“In this digital age, buying a theater ticket and going to something for three hours feels slightly outdated,” Barrett said in 2014. “Whereas the model where you subscribe and you pay weekly and you live inside the show for six months, or your life becomes the show, it feels like — what I’m interested in the next tranche of work is that sort of magic hour, the dusk between the fictional and real, where you almost become your own avatar. You become the star of your own film, and you, rather than going into a building or a theater to go into a show, what happens if the show comes to you?”
A series of experiments
When Niantic announced its Punchdrunk partnership six years later, the marketing line rang a familiar bell:
“Recently, we’ve asked ourselves: What if we were the lead characters in an epic adventure movie? Our new collaboration with Punchdrunk might be the answer.”
Hamlet was to be a realization of Punchdrunk’s long-held plan. A Punchdrunk show, broken into pieces and doled out to players over time — in this case, using Niantic’s technology with a story loosely tied to Punchdrunk’s then-upcoming show in London, The Burnt City. The show, which debuted in 2022, is structurally similar to Sleep No More but based on Greek mythology and the fall of Troy. “Gods and mortals rise for a party at the end of the world,” as Punchdrunk described it.
“The idea was, can we make an experience that has some of that magic of that live theater experience, immersive storytelling, that’s accessible to more people by using devices to expand the reach beyond just a physical installation and actors and a couple hundred people per night?” Hanke told Polygon.
The concept was unproven, but Niantic was in a position to take risks. Thanks in part to Pokémon Go’s success, the studio was on a yearslong shopping spree, acquiring more than 10 companies and signing deals with some of the world’s biggest brands. At the time it announced the Punchdrunk deal, Niantic had more than 10 games in development.
On Hamlet (which went by titles such as “Tracelight” and “The Trust” during development), neither company rushed into production. By the time Niantic and Punchdrunk went public with the partnership, they’d been experimenting with ideas for a couple of years. A Niantic team in Los Angeles and a Punchdrunk team in London ran through dozens of concepts and prototypes, trying to figure out how a crossover between theater and games could work. The studios experimented with audio-only gameplay concepts, where a player would stick their phone in their pocket and control the game by walking, guided by audio in their earbuds based on their GPS coordinates. They also played with releasing a game under the guise of a wellness app called The Meadow, which players would download and start to use before a member of a secret society would break in and ask them for help, transforming the app into a game.
One of the main concepts the teams came up with was setting up interactions with paid actors, letting players experience the sorts of scenes they might come across in a Punchdrunk show. The game would direct a player to a certain location, where they would stand and wait for an actor to approach and act out a scene for them.
Another idea that gained traction was a card collecting system based on an in-universe spin on tarot cards. Team members described it as a geocaching-based game of hide-and-seek: Players would hide physical cards behind bushes, inside mailboxes, or under benches, then offer hints to direct other players toward them. Once those players found the cards, they’d scan and re-hide them.
Over time, these experiments solidified into a game that used a map similar to that of Pokémon Go, with various points of interest and a resource called Tracelight to collect. Actor interactions and the card concept stuck around as central features, though the cards became digital objects instead of physical ones. Niantic planned for the game to be the first use case of its Visual Positioning System technology that would allow players to place cards in precise locations, tying the game to one of the company’s tech initiatives and one of its steps toward building a “real-world metaverse.”
All of these ideas fed into a story Punchdrunk was piecing together, in which players were members of an organization called The Trust. Many of the narrative elements changed over time, but on a basic level, The Trust spent its time investigating a dark alternate world underneath our own called The Lowlands. As players collected cards and different types of Tracelight, they could light up The Lowlands and catch glimpses of the world and characters underneath, with certain actor interactions and big in-person events (such as performances of The Burnt City) representing stronger connections between the two worlds.
In an internal world-building document viewed by Polygon, Punchdrunk laid out a deep backstory detailing thousands of years of history and a series of elaborate concepts for how bigger narrative events could work, such as the pay phone sequence.
In another example in the document, a player would track down a shipping container in a remote desert near Salt Lake City and meet a security guard, who would then take them to his bed in a hidden part of the container and tuck them in, at which point the player could summon a character from The Lowlands to receive a blessing.
In another, a large group of players would gather at an abandoned swimming pool in Moscow for a party, and actors would pull aside certain Trust members to tell them secrets about 49 maidens trapped in The Lowlands.
Most of these were high-concept ideas for how scenes could look rather than how scenes actually played out in playtests, which tended to be more focused. But across tests in London, San Francisco, and Los Angeles over the course of development, team members said that even the smaller narrative moments proved exciting.
“I remember testing a part of the game in Los Angeles,” wrote Barrett in an email interview with Polygon. “I had to hand over a covert package to a stranger — a kind of classic spy story trope. It was happening in the real world with a soundtrack playing in my ears. I was driven to this stranger by the app and the game engine. I had never met the person — and suddenly, I arrived at a location and made the handover while the soundtrack got ever more dramatic. The euphoria I felt was incredible.”
Barrett referred to the musical component as “soundtracking your life.”
“I remember as I approached a real location in the game, the soundtrack built and built, and it felt like I was in a film.”
Development challenges
Despite Barrett’s excitement, a crossover between theater and games was never going to be easy. On a structural level, the teams needed to figure out not only how to design the game but how to build it to work consistently. They also needed to find the right balance of how players would spend their time when not involved in the big showpiece moments. People speaking for this story listed many challenges unique to the concept, with some of the biggest not having clear solutions.
One major concern was player safety, a consideration in many Niantic products. Pokémon Go, for instance, has famously led players into traffic and toward a dead body. Punchdrunk has run into issues with keeping its actors safe as well — a 2018 BuzzFeed News story reported instances of sexual assault from audience members at Sleep No More.
For Hamlet, safety became a particular concern due to the types of interactions the game would set up. With players acting as members of a secret society, meeting people on the street, and searching for hidden objects, there was concern that what made the game unique and interesting also made it high-risk.
“Unfortunately, when you make the real world a game, people start treating other people like NPCs,” said one team member. “It gets ugly.”
The teams had various strategies planned to keep players and actors safe. Players would stand in specific positions, such as holding their hand over their heart, to signify they were participants, and actors would be the ones to approach players rather than the other way around. But there would always be variables the teams couldn’t account for.
“I think that the experience of playing this game would be really different for different kinds of people,” said another team member. “You know, it was different for the men who were playing it. It was different for the people of color who were playing it. It was different for the women. A lot of people, even on the team, were sort of raising these red flags of like, ‘Hey, I’m a woman, and I don’t know if I’d feel comfortable just waiting in a spot for someone to come and whisper in my ear.’”
Multiple Niantic team members said that things generally went well while Niantic and Punchdrunk were testing the game. But team members worried about what would happen when the companies wouldn’t be able to monitor these events as closely. And if the companies swung the design too far toward making it safe, one team member said, they risked filing off much of what made it interesting.
“There were ways to make it less immersive in the Punchdrunk sense, and sort of really over-index on the safety aspect,” the team member said. “But I think that that would have sort of taken away from the spirit of the Punchdrunk-ness of it all.”
Team members also pointed to legal risks if things went wrong, as well as screening and management challenges for overseeing actors all over the world.
“Both Niantic and Punchdrunk’s development teams worked closely with experts in trust and safety, focused on safety by design, and leveraged learnings from multiple products built by both companies,” wrote a Niantic spokesperson in response to questions sent by Polygon. “An internal red team helped test safety scenarios and discover potential risks in order to ensure that the right protections were in place for players. This was done jointly by both Niantic and Punchdrunk, and we indeed had thorough safety playbooks designed to keep our players safe.”
Asked about a moment the companies tested where a player tracked down a parked car, opened the door, and went inside, the Niantic spokesperson wrote that it was only meant for internal testing, utilized a Niantic employee getting into another employee’s car, was monitored by staff, and “was not part of any actual or intended gameplay for the final product.”
“The number one challenge was always safety,” wrote former general manager of Niantic’s Los Angeles studio Greg Borrud in an email interview with Polygon. “We were always looking for ways to push the limits while still keeping safety at the forefront. We knew we had to test the boundaries to see where they were during development. But we also knew we had to reel things back in as we looked to put a live product out there.”
The start of the COVID-19 pandemic in early 2020 brought a new layer of safety concerns. While the pandemic presented challenges for companies around the game industry, the majority of those centered around how teams worked behind the scenes. For Niantic — as well as Punchdrunk and much of the theater industry — the challenges extended to their output, with the companies relying on getting people out of their homes and interacting with others. In response, Niantic set up ways for Pokémon Go players to play from home, and Punchdrunk delayed the theatrical launch of The Burnt City.
For a project like Hamlet, this presented a number of questions. Could Niantic and Punchdrunk still encourage players to go outside? Could they still set up in-person narrative events?
“Where with Pokémon Go, you worried about people walking into traffic and trying to get Pokémon while falling off cliffs, we had to worry about that and also the heightened level of sensitivity just being in public,” said one team member.
Hanke said that on the development side, COVID also made it harder to conduct playtests, which slowed progress on the game. “Sometimes you just need to get a bunch of people together and try stuff and be able to iterate quickly, and we were in a mode where it was a lot of pre-planning,” he said.
Beyond safety concerns, Hamlet also struggled with scale. The game’s base concept required a roster of trained, paid actors in any city where the game was running. That would also require a team to manage those actors and extensive work to adapt the game’s real-world elements to specific locations — costs that wouldn’t decrease as the game spread to different cities.
“A lot of things were very surprising and interesting and on the spot,” said a team member. “But the amount of overhead and coordination that requires per player would [make for] an incredibly boutique experience that wouldn’t be able to scale to a free-to-play-style product.”
Switching from physical cards to digital cards was a step toward making the game easier to manage in different locations (though Niantic’s Visual Positioning System brought its own costs and challenges), and the teams explored ways of making actor interactions simpler and cheaper over time, but compared to other Niantic games where most things were automated, Hamlet added an extra layer of costs.
Hanging over many of these challenges were differences of opinion between Punchdrunk and Niantic. These varied by person and over time, but six people who spoke for this article said the companies often clashed while working on Hamlet, noting that the two never fully got on the same page, which they said slowed progress on the game.
Conceptually, Hamlet was supposed to blend Punchdrunk’s experiential ideas with Niantic’s game design knowledge, but the six people said finding common ground between those two proved difficult, with Punchdrunk getting frustrated that its ideas were watered down or not taken, and Niantic getting frustrated over Punchdrunk’s lack of game development experience and frequent iteration. Four Niantic team members said Punchdrunk was great at dreaming up big ideas for when everything played out the way it was supposed to, but didn’t always have answers for what to do when things didn’t go according to plan.
“Setting a mood is not the same as making a thing where you have to do things, and you’re building up and there’s choices and there’s a story,” said one Niantic team member. “[…] So trying to do something that is Niantic-like, that goes on forever, that has a game-like element, was not in their wheelhouse.”
The team members also pointed to philosophical differences between the companies, noting that Punchdrunk pushed to make things more risky and uncomfortable, while Niantic tended to be more technical and practical.
“I think part of the Punchdrunk formula was very much wanting to push people outside their comfort zones,” said a Niantic team member. “And that was sort of the experience that they wanted, which didn’t entirely line up with the experience that Niantic wanted, which eventually became more like, ‘We want this to be available for everyone.’ And that sort of became a difficult design tradeoff where the more you make it accessible for everyone, the less you can push people outside of their comfort zones.”
Borrud described the differences as “a natural tension between what was a game and what was a performance.” Borrud, Hanke, and Barrett all pointed to the physical distance between the London and Los Angeles offices as a key barrier for a project that required extensive creative collaboration and in-person testing.
“With something this experimental I think we would have been better served by having the entire team all under one roof,” wrote Borrud.
“When we were in the room together the creative sparks flew,” wrote Barrett, “but having to be mainly online with an eight-hour time difference made everything slower and more difficult.”
The final months
By early 2022, despite more than three years of work, former Niantic head of development of games Eric Gewirtz described the game as still being in preproduction.
Niantic and Punchdrunk had taken a number of steps to overcome the various challenges, bringing in a team from Niantic’s London office (which addressed some of the time zone issues), developing a set of moderation tools (which addressed some of the safety concerns), and adding forms of digital storytelling, like prerecorded audio clips, robot calls, and an online ARG (which addressed some of the scale challenges). The teams also continued to iterate on the design and pulled back on some of the game’s more ambitious narrative elements, but they hadn’t gotten to a place where everyone involved was confident in what they had, and they hadn’t produced a lot of content beyond an initial onboarding sequence.
Niantic and Punchdrunk were working toward debuting the game in London, with plans to roll out in other locations like New York and Los Angeles later. During internal testing, some of which took place around shopping district Covent Garden, the staff ran players through different sequences, such as an onboarding sequence where the game would call players and ask them a series of questions.
“Where I think we hadn’t quite found the secret sauce and the magic yet was in the moment-to-moment gameplay,” Gewirtz said. “There were these really epic, magical experiences that would happen that were kind of charted out, but in between [during] the hours of gameplay, it was challenging to keep the players engaged and occupied.”
Some team members were also concerned that, years into development, certain safety issues hadn’t been resolved, referencing discussions like whether actors would have access to player phone numbers (ultimately, they would not) or how closely actors would have been able to track individual player movements.
The teams had set up a number of safeguards to log communications between players and actors and built a feedback system to allow players to report unusual or inappropriate behavior, with Niantic’s trust and safety and legal teams involved in the process. Players would also have had to agree to terms and conditions saying they were of age (“we knew this would always be an adults-only game,” wrote Borrud) and they would have had options to turn off certain aspects of the game, such as being able to receive calls. But some team members remained nervous that they couldn’t anticipate every scenario.
At one point when testing the game in London, Gewirtz was searching for an actor in the middle of the day when the game directed him down an alley. As he started walking in that direction, he discovered a woman defecating in front of him.
“I was like, OK, I think I’m going to go down a different alley,” he said.
The story became a popular anecdote among the Niantic staff, getting passed around and twisted like an urban legend.
“It was just funny that it happened to me because I was really in the moment right there,” Gewirtz said. “I was listening to the audio. I was really kind of getting immersed in the quest, in the adventure, and that really just pulled me out of it.”
“It was a stark reminder that there are things we can control and things we can’t control when you have any type of thing happening in the real world,” wrote Borrud. “Even a place as controlled as Disneyland has things they can’t control when it comes to guest behavior. I don’t have a good solve for those variables yet. But it was something we talked about quite a bit and was a reason we started to move the core gameplay closer to known ‘Points of Interest’ and tailor the initial experience to a specific neighborhood that we knew intimately. Ultimately, though, we also had to acknowledge that there was no way we would be able to control the real world.”
Calling Hamlet off
On June 29, 2022, Bloomberg reported that Niantic had stopped production on Hamlet along with three other projects, cutting approximately 85 to 90 jobs, or about 8% of the company. The news followed Niantic’s struggle to diversify its portfolio beyond Pokémon Go, with games such as Harry Potter: Wizards Unite and Catan: World Explorers underperforming, as well as its March 2022 acquisition of AR tools company 8th Wall, which Niantic called its “largest acquisition to date.”
According to Hanke, Niantic stopped production on Hamlet because it didn’t see a way to make the math work when hiring actors and customizing the game to multiple locations for a wide release.
“We were still struggling to achieve our vision for the product, or bringing that excitement of live immersive theater into a broader setting,” said Hanke. “And when we looked at the numbers of the way that we got there, the way that we got where we did was going to be really expensive, and it was going to be something that we were going to have a hard time doing outside of very specific kinds of cities.”
“We got an experience in London that we were reasonably happy with that did some interesting things,” said Hanke. “[…] I think I would’ve been happy to launch it in London, save for the fact that we just didn’t have a plan to expand it beyond London that made financial sense.”
Prior to shutting down the project, Niantic went through extensive discussions about how to monetize it, looking into releasing it as a free-to-play game, a premium game, a subscription game, an app tied to tickets to The Burnt City, or a game featuring paid street performances. But the game’s unique elements presented heavy upfront costs.
Six developers said they were not surprised that Niantic and Punchdrunk moved on from Hamlet, saying it always felt like a pet project or something the companies were doing for prestige or to reach into other markets, rather than a straight attempt to make a profit.
Multiple team members also mentioned that Punchdrunk was hit particularly hard by COVID and its lingering effects on the theater business.
“Punchdrunk […] were really pushed to the limit to get their play done and launched, coming out of COVID,” said Hanke. “It was very, very challenging for them. So there were kind of bandwidth constraints on their side that kind of made things extra challenging. So when we wound it down, they were able to really just focus on getting the play done, which was — obviously that was the existential thing for them that had to happen.”
“The truth is that two years of working at distance with all the restrictions of COVID took some of the momentum out of it all,” wrote Barrett. “It meant we couldn’t travel literally and figuratively as far as we wanted to.”
Staff also heard mixed messages on the game’s release date. An early plan to release the game alongside The Burnt City hadn’t worked out, and as time went on Hamlet ran into multiple internal delays. One Niantic team member said that, just before the game got cut, they heard it was supposed to be two weeks out from release, which they called a “pipe dream.”
“I mean, it never would have been finished, is kind of my take on it,” said another Niantic team member.
Borrud estimated that, by mid-2022, the teams were four to six months away from being able to bring in an initial small batch of public players, though he added it’s “hard to really say if we would have been ready for that initial cohort.”
“We were ready to try to get there and evaluate along the way,” he wrote.
Moving on
On June 29, Niantic announced another round of layoffs, this time closing the Los Angeles studio where the majority of Hamlet had been built and cutting approximately 230 jobs, or about 25% of the company.
In a statement, Hanke laid out various challenges facing Niantic, such as a crowded mobile market and changes to the mobile advertising industry, noting that employees should “expect a more direct and results-based culture” going forward. As part of a shift in direction, he wrote that Niantic would be scaling back spending on mobile game investments in favor of supporting current games, such as Pokémon Go, and focusing on technology and content for mixed-reality hardware and the future of augmented-reality glasses.
A few months later, on Sept. 24, Punchdrunk held The Burnt City’s final performance in London. In a statement released alongside the announcement of that performance, Barrett wrote that the show “has propelled us forwards and we want to keep that momentum,” adding that Punchdrunk will use the same location as a long-term home for other projects. “We’re fairly sure that The Burnt City will be the last new mask show the company makes, and what comes next will be different and unlike anything we have done before,” he wrote.
On Nov. 8, Punchdrunk announced that Sleep No More’s New York run will end in January 2024.
Speaking prior to Niantic’s directional shift, Hanke said that he hopes Niantic can return to a project like Hamlet in the future, noting that recent advances in artificial intelligence could reduce the operating costs by helping customize the game to different locations or replacing live actors with digital ones.
Barrett agreed AI would have helped Hamlet scale more affordably, though he was less enthusiastic about the prospect. “Sad to say but it is probably true,” he wrote. “I am very resistant to the idea of AI replacing actors and people, however, there is no doubt that some of the time we spent trying to [automate messages], for example, AI would have helped us solve much faster.”
Barrett wrote that Punchdrunk will “absolutely” return to a live video game concept like this in the future. “To my mind, still, no one has created a major crossover video game/theatrical project,” he said. “We truly believe that is the future of shows. We see it in audiences, especially younger ones who have grown up with video games and it feels inevitable that this future is coming.”
“I loved the idea,” said Hanke. “I loved the experience of building it. It was heartbreaking that it ultimately didn’t work out as a project that we could continue to invest in. I hope we’re able to come back to the concept. […]
“I think it’s inevitable that it’s going to happen, but there are definitely some unique challenges there.”
