Earlier this year, the UK lost one of its longest-standing game developers. Lionhead Studios, founded in 1996 and bought by Microsoft in 2006, was closed on the 29th of April, with the loss of around a hundred jobs. The game Lionhead was working on, Fable Legends, a four-versus-one fantasy-themed multiplayer game that was already in closed beta, was cancelled. Despite the reported emergence of several interested buyers before Lionhead’s closure, neither the studio nor the game was saved.
The Lionhead studios story is bigger than its last game. (For a great read on the studio’s full history, I recommend Wesley Yin-Poole’s recent Eurogamer article). But Fable Legends’ development is emblematic of Lionhead’s troubles in its final years, as well as some of what made the studio great. It was an ambitious game, and beautiful, and it had a great sense of humour – but it was pulled in so many different directions over the course of its development, largely at the whims of the studio’s Microsoft bosses, that it struggled to come together.
Over the last four years of the studio’s life, Lionhead was facing the same intimidating problems as many other similarly sized studios: vastly increasing development costs, a changing marketplace where the money was shifting to free-to-play and mobile, frequent changes in leadership. Fable Legends was a bet that didn’t pan out.
Before Fable Legends, Lionhead was in a strange place. After Fable 2, the studio’s most successful game, Lionhead was divided. Its owners, Microsoft, were keen for it to keep working on yet more Fable. But its famously esoteric founder, Peter Molyneux, had grown bored of the franchise, and between 2008 and his departure from Lionhead in early 2012 he was spending much of his time on esoteric passion projects. One such project was Milo and Kate, intended as a showcase for Microsoft’s then-new Kinect technology, an odd piece of software that intended to create the illusion that you were communicating with a young boy who lived inside your Xbox. (It was cancelled in 2010, and much of its tech was put to use on Fable: The Journey, also intended to be an illustration of Kinect’s wondrous powers).
After Milo and Kate, Molyneux and a large team of others at Lionhead were working on something called Project Opal – a sprawling project so unfocused that even the other designers working on it had no idea what it was actually supposed to be in the end, according to some people who were with the studio at the time. (Due to non-disclosure and redundancy agreements, the people who spoke to me for this article did not wish to be identified.) Opal was to be a vast, interconnected game across PC, mobile and console, some kind of cross between a village-building, resource-gathering game, action game and management sim. Molyneux retroactively describes it as “a game that never existed before”, intended to “connect millions of people together”. Thematically, it bears some resemblance to what Molyneux went on to attempt with Curiosity and Godus.
For some people at Lionhead, this troubled period was when the studio’s problems started. The studio was burning through time, money and goodwill with its owners at Microsoft, whilst failing to come up with anything that could actually be released. A group of Lionhead’s key staff left in 2012, unsure of the studio’s direction or its future. Molyneux himself left in March of that year.
“For some people at Lionhead, this troubled period was when the studio’s problems started.”
Something else happened in 2012: former Sony executive Phil Harrison joined Microsoft as corporate vice president in charge of its European game development efforts. Harrison has a long history in the games industry – he is a part of the story of a great many high-profile games of the last two decades – and he had a very clear vision for where games were going. Phil passionately believed in games-as-service: in other words, long-tail online games that evolved with their player bases, and were probably free-to-play. This belief would be what determined Lionhead’s direction (and, judging by Sea of Thieves, Rare’s too).
The next Fable project was originally intended to be an expansive single-player game that built on the series’ heritage, set in a sprawling London-like city. But that pitch was rejected, and at some point between July and October 2012, the next Fable game had become Fable Legends.
“I thought I was going to be working on a single-player game, a more advanced version of Fable 3,” says one source. “But when they went to get that game approved, the three senior designers who were pitching it were told that ‘you will not be given permission to make Fable 4, or something that is a shadow Fable 4’. Phil Harrison’s vision for all of his studios in Europe was now for service-based games. That’s what he thought was the future of games. He didn’t want to make anything that was a £50 box, fire and forget. He wanted long tails of revenue, even if there was a smaller up-front burst of revenue.”
“Harrison had a vision for the future where everything was games as a service,” says another. “I think many parts of Microsoft really struggled with that approach, despite the fact that Windows 10 was going in that direction. Harrison absolutely did not want a boxed product.”
Given Harrison’s background, this approach was understandable. Between his stints at Microsoft and Sony, he had been special advisor for a company called London Venture Partners, which invested in new business models in the gaming world and counted Supercell as one of its early seed investments. He had seen, close-up, how successful games-as-service could be. Lionhead had to come up with something that would fit the model.
“The pitch [for Legends] was along the lines of Dungeons and Dragons with four friends and a dungeon master,” says a source. “How do we make that digital? That was the initial idea around it. We weren’t aware of any other games in this space at all. The elevator pitch was very much Fable meets Left 4 Dead meets Dungeon Keeper meets League of Legends. Actually, at that point it was also meets LittleBigPlanet, because we had the idea of having a comprehensive user-generated content approach. But that took more of a back seat as we focused on the core challenge of actually getting the game right.”
“Lionhead was the studio that Microsoft was willing to risk.”
If that sounds a little unfocused, consider that Lionhead had absolutely no experience in making games of this nature. Its staff had no experience with servers, community management, ongoing development support, free-to-play monetisation, player retention, or any of the other new specialties that the games-as-service world had created. Indeed, they’d never done multiplayer-specific game design.
So with that in mind: why on earth did Microsoft choose Lionhead to spearhead this radically new approach?
“The positive thing you could say about it is that Lionhead had such a great history of innovation,” said one source. “We’d already done experimental things with stuff like Kinect, with Journey, that proved we could pull off different ideas.
“The other thing you could say is that Lionhead was the studio that Microsoft was willing to risk.”
Let’s take a look at Lionhead’s position in around 2012. The studio had about a hundred employees: not huge, but not small enough to be cheap either. It was Microsoft’s fourth-biggest studio. It was headquartered in Guildford, and though the UK is not the least expensive place in the world, it’s certainly less expensive than a studio in, say, California. Since its last successful release in 2010, it had spent a lot of its years burning money on incubation projects that never came to fruition. Lionhead was far enough away from Redmond to be insulated from it, but also far enough away to feel remote. Its position was by no means enormously secure.
If Lionhead did not follow Harrison’s new direction, what was it going to do? Fable Legends was the game that they had to make. And though not everyone at Lionhead was on board, they were going to make the best of it.
“We go into it with our eyes open, and we try to make the best of it, with the awareness that it was a different kind of title,” says a former Lionhead employee. “Culturally it was a challenge. We had to change the company, we had to invest in new types of skills that we hadn’t got before: product managers, community people, business managers and service engineers… all the stuff that goes with a free-to-play game.”
“The studio was mostly behind it – by process of elimination,” says another. “We shed 10 per cent of our staff in the first year. Those were the people that rejected it outright.”
Much of the reluctance, says a source, was down to the fact that what Legends was attempting was not only new for the studio, but altogether new. It felt risky to some, even then. “It was a tough time, but an interesting time,” they say. “Even on the leadership level, there were people who were quite old school: they were used to what they were used to, which was boxed retail products. Within the team and within Lionhead there was a number of people that just loved making those games. The idea of moving away from that scared them. It was pretty much always planned to be free-to-play, and the idea of free-to-play scared a lot of people really early on. It was synonymous with Facebook and rip-off Skinner boxes. I guess things like League were around at that point, but certainly a lot of the members of the team weren’t that familiar with it.”
“The idea of free-to-play scared a lot of people really early on.”
But those who remained ended up getting behind Legends, and the unique challenges that it threw up. “The core design of the game was fun to work on – it was full of interesting design challenges and things to solve,” says one person. “What was more worrying was trying to bring the fans along on that journey, to convert them to a really new idea that nobody had done. When we announced it, even Evolve hadn’t come out yet. That asymmetrical design was new at the time.”
“As time wore on and we got to understand more about games as a service, [we felt that] what we were trying with Legends was a really cool idea,” says another former Lionhead developer. “Previously if we shipped something and the consumer hated it, we were fucked at that point. But the whole thing with Legends was that we were going to work with the community on iterating on it.”
The former Lionhead staff who spoke to me maintain there was nothing fundamentally wrong with the idea behind Legends. (As someone who played it at several different stages in its development, I’m inclined to agree; it was unexpected, certainly, but it was fun.) They felt that Lionhead was capable of rising to the task and making something unlike anything they had ever made before. And let’s not forget: they almost did. Fable Legends was just weeks away from open beta when it was cancelled.
The core problem wasn’t with the game’s design. The core problem was that Fable Legends was much, much too expensive for what it was. Fable Legends could have worked as a small-to-medium-scale free-to-play game, if it had been released quickly and iterated upon well. But with every adjustment to the Xbox division’s corporate team, Microsoft’s changing priorities created a very dangerous situation for Lionhead.
“The game was pitched to Harrison as a medium-scale game, a AA game. It was supposed to be out significantly earlier; it would have been last summer,” says a source. “But the size of the game just kept growing, and the fidelity value of the game kept growing. And that was because we were the servant of two masters. We reported to Phil Harrison, the master of Europe, but we also had another person that he does not report to: Phil Spencer [head of the Xbox division]. And he wants a beautiful AAA quality experience that he can use to sell Xbox Ones. So now we’re making a free-to-play game that’s as expensive as an AAA game. Very dangerous.”
“The original pitch was for a really cheap game – it certainly wasn’t the $75m we ended up spending,” corroborates another source. “There were going to be three phases of release. But as time wore on, there were various voices that made it more complicated. For example, Spencer was very keen on having the Fable features: it was crucial that it could be played single player, for instance that was suddenly a big important thing. It was also supposed to be “the prettiest ever online game” – that was Harrison, he wanted it to be prettier than anything else out there.
“The original pitch was for a really cheap game – it certainly wasn’t the $75m we ended up spending.”
“The cost was considerable. It was a big, big game. Art was a lot of it, audio was a lot of it, but there was quite a large engineering team – in part because pretty much everything we created had to work within an multiplayer environment and client-server environment. We had to invest a lot of time and effort on anti-cheat things to prevent hackers from screwing over our balance. There was a lot of infrastructure work – we needed to create new systems, some of which didn’t exist within Microsoft, around analytics (your life-blood for a F2P game, the key to understanding players and their behaviour).”
So here was Lionhead, making an experimental free-to-play game that was quickly becoming as expensive as any major first-party game you could work on. It would have to be very successful to succeed.
Fable Legends was announced at E3 2013, which seemed pretty early to many members of the team. But then, showing your game at E3 can be something of an insurance policy. If your publisher makes a big, public fuss about your game, it’s a show of confidence. Even though the game wasn’t ready.
“We were announced early, in 2013, so that we could be there as an advertisement that, yes, a fantasy RPG-like game is coming,” reckons a source. “So that’s announced at least six months to a year before we’re ready to announce anything. It was just a trailer. But even at Gamescom in 2013, we were showing early gameplay.
“Then we show off in 2014 at E3, for the same reasons, and we’re showing off a cool fun level to play – but that’s the only cool fun level we’ve built. We’re showing four heroes, but we’ve only built six. We’re showing the world everything we’ve got. All of that stuff should have been a year later. But our hand was forced.”
It was around this time that Microsoft came up with a new priority for Fable Legends: Xbox One and Windows 10 cross-play. There is reportedly an initiative within Microsoft, codenamed Helix, that centres around Windows convergence; the eventual aim is for all of Microsoft’s products to run the same software. In the shorter term, all Xbox One games were to be adapted to run on Windows 10 as well. Fable Legends was to be the first game that would do this.
This would take more time and money – and it wasn’t the only showcase feature that Microsoft asked for over the course of Legends’ development. Fable Legends was also to support DX12; back in 2013, it had to make use of the cloud and Smartglass, Microsoft’s ill-fated second-screen initiative. But it was the Windows 10 cross-play that proved the most damaging demand, because it would greatly restrict Fable Legends’ potential audience if the game were only to be available on PC through the Windows Store. Originally, the plan was for Legends to release through Steam – but the Windows/Xbox convergence strategy put an end to that.
“When we started the project we weren’t aware of Windows 10,” says a source. “We were going to ship on Xbox One initially and then we wanted to come out on PC at a later point, most likely though Steam. But we got burnt quite badly.”
“Without Steam, without other platforms, it was just painful.”
“Without Steam, without other platforms, it was just painful,” says another. “The Windows Store is a giant disaster. It’s on fire. 98% of PC copies of Rise of the Tomb Raider, a flagship Windows 10 game, were bought on Steam. The same is true for Minecraft. That hurt us, too. The store’s a mess; the number of people who couldn’t even install the game from the Microsoft store was… significant.”
Besides the technical problems with Windows 10, though, Lionhead’s bigger problem was that for the entirety of Fable Legends’ development cycle, the potential audience for the game was shrinking. Xbox One sales were falling far short of projections. Windows 10 installs, too, were nowhere near what Microsoft had planned. The scale just wasn’t there. And for a free-to-play game, scale is everything.
“Let’s be honest – we make our projections based on a series of assumptions,” reflects a former employee who worked closely with Microsoft. “There are supposed to be 2x as many Xboxes out there as there are right now. There are supposed to be 2x as many Windows 10 installs as there currently are. So now, when we look at how much money Legends could make in the free-to-play universe, you have to halve it. Because we can only reach half the audience that was projected.”
Fable Legends went into closed beta on the 16th October, 2014, and it would stay in closed beta until the servers were shut down on April 13, 2016. One Lionhead employee I talked to maintained that it could have been released pretty much at any time. The studio, on the other hand, wanted it to launch in its best possible state.
“I was of the opinion that we could have released Legends at any point. It was good to go. But we intentionally didn’t release it, because the last thing we wanted to do was release it when there were problems with the game that meant many people wouldn’t come back,” they say. “You wanted to release it when the game was right. But I guess we ran out of time.”
Testimony varies on the state of Fable Legends during its protracted beta. Some say the game was pretty much in place; others maintain that there were still big problems with balance, and that it was still lacking significant, crucial features.
“It was really clear during the latter half of the closed beta that player retention was a challenge,” says a source. “We knew it was going to be a challenge, because the early closed beta versions you just got to play the one quest, which was a good fun experience – but it needed a metagame on top of it to get people to come back for another go. A lot of those systems came in quite late; monetisation wasn’t turned on until October last year. Our monetisation stats were pretty good, and we had some incredibly passionate players who racked up over 1,000 hours playing, but it would appear that during the closed beta, there was a much larger number of people who would load it up once and wouldn’t come back.”
“We knew that we were supposed to be part of the greatest ever holiday line-up on Xbox One. We hadn’t made that, and we were on borrowed time.”
“Much like any free-to-play game, you don’t project that a very high percentage of people are going to pay for it. We beat some of our projections; people who did play the game did spend money and spent more than we expected, but the size of the instal base was a large part of our pain,” says another.
“That’s what we were trying to fix in the first few months of this year. We were making progress, but we knew that we were supposed to be part of the greatest ever holiday line-up on Xbox One. We hadn’t made that, and we were on borrowed time. If we didn’t get to a point soon to release, we’d be in trouble.”
Some people high up at Lionhead had begun to realise that they were losing Microsoft by the end of 2015. John Needham (the studio head who had been brought in to oversee Legends) was suddenly whisked away back to Redmond in 2015, partially to replace Phil Harrison, who left Microsoft that year. Hanno Lemke, general manager of Microsoft Studios Europe, was now overseeing Lionhead. Over the course of Fable Legends’ development, incidentally, Lionhead had four separate studio heads.
“We know that we were losing Microsoft, at the leadership level,” says a source. “They were unhappy with delays. They were worried about projections. We’d turned on monetisation in the beta, and the retention and monetisation numbers weren’t as good as they’d hoped. That wasn’t a surprise; that’s why we’d soft-launched, to fix these things, to deal with them. But our ability to iterate in that initial period was nonexistent. We wanted another six months; we thought we had longer cycles to fix problems.”
But by the beginning of this year, Lionhead had run out of time. A new financial year was approaching. Somewhere along the line, a decision had been made. On March 7th, staff were called into an all-hands meeting and informed by Hanno Lemke that the studio was being shut down and Fable Legends was to be cancelled. People at the studio found out about their fate just half an hour before the wider world got the news. People at other Microsoft studios found out when the press release went out.
“Nobody knew about the closure,” says a former high-level Lionhead staffer. “The day the world found out, the studio found out – about 30 minutes in advance, at most. The press release went out at the same time as the studio all-hands meeting.”
“The day the world found out, the studio found out – about 30 minutes in advance, at most.”
One source reckons that the final decision was probably made about a week before, as Microsoft started doing its fiscal planning, deciding what products and studios to fund and not. Lionhead was not the only first-party studio that was having problems. Something had to give.
“I reckon [Phil Spencer] feels bad about Lionhead, but I think for him, he has to run all of Microsoft Studios as a business,” they say. “First-party studios isn’t doing so well. Halo 5 is a big miss, versus projections. Minecraft is a big miss, versus projections. Compared to either one of those, Lionhead is practically a rounding error. But I think if your division is under-performing, you have to go to your boss with something on the altar.”
There was an outpouring of sadness over the announcement of Lionhead’s closure. Fans and former employees penned their remembrances; fans expressed their dismay that the studio would be no more. Meanwhile, the UK games industry – as it often does – came together in support of Lionhead’s employees. People who were there in the final weeks say that barely a day went by when some company wasn’t in Guildford, either at the studio or at a nearby pub, pitching their open positions.
On the last day that Fable Legends’ servers were live – April 13th – one employee spotted a Twitch streamer broadcasting the final hours of the game. “It was totally against the NDA, but at that point, who cares?” they say. “She was playing it for the last hour, before it closed down. We had it on the big screen. It was just so emotional. It was depressing. It was really sad, a real shame.”
There were those at Lionhead who knew that Legends was turning into an increasingly dangerous project. But nobody that I spoke to thought that Microsoft would actually shut the studio down. The worst-case scenario was that Fable Legends could be disassembled, and all the beautiful art, audio, characters, creatures and environments could be used to build a more traditional, single-player Fable 4. But Microsoft decided to just cut its losses.
“There was always a scenario where Fable Legends wasn’t a success, because we were trying something brave and innovative, and that was what Lionhead prided itself on doing,” says a source. “I think we felt that there were options; one thing we could have done pretty damn quickly is taken all the assets and made a Fable 4 out of it pretty easily. We felt like we had a life-raft. But that didn’t happen.”
“There was always a scenario where Fable Legends wasn’t a success.”
Throughout the mandatory six-week consultation period, there were offers coming in from companies that wanted to buy Lionhead. A bunch of employees banded together to shop Legends around, but they were unsuccessful. The problem, according to one high-level source, was that Microsoft refused to sell the Fable IP. They were only willing to license it out.
“As soon as the closure was announced, anyone who had an interest in saving this studio was holding water dripping through their hands,” says a source. “Everyone in the studio had half a dozen people calling them up and saying they’d be a great fit somewhere else. The universe came calling for so much of the talent – so many people came through the doors, and Microsoft wouldn’t sell the IP.”
It’s easy to speculate that Lionhead could have been saved if Microsoft had been willing to give up the rights to Fable. But from a business perspective, it’s also easy to see why it wasn’t. The possibility of Fable on PlayStation would be a blow to its pride.
“The stranger decision is, why would they get rid of a studio that had that expertise just as they were trying to pivot other teams towards games as service, like Sea of Thieves?” says a former employee. “The knowledge and learnings that we’d had at that point would have been great for those other projects.”
When a studio closes, it’s not just the products that you lose. It’s the people, and the culture – and Lionhead’s culture had been developing for 20 years. There aren’t many other studios like it, especially not in the UK, where mid-sized developers are struggling to survive. It’s sad that we will never get to play Fable Legends, the last, almost-finished game from a legendary UK developer; it was so close to finished that you can still buy the official art book. But it’s sadder that we’ve lost the inventive studio culture that produced it – that almost made it work, despite the constantly changing demands of Microsoft and the ever-more-challenging market conditions that it was launching into.
But then, for many fans, Fable Legends was not the Fable game they wanted. What has hit them hardest about Lionhead’s closure is that we’ll never see a Fable 4 – at least, not one from Lionhead. And Fable 4 is something that could have happened, if Microsoft had been willing to let the studio work on it, back in 2011 – or if it had agreed to sell the studio on.
“Legends was an incredibly innovative game that took the franchise to new heights, but a lot of the fanbase wasn’t ready to embrace that,” reflects a source. “You can make a reasonable argument that we took the franchise to a place where Microsoft wanted to take it. But perhaps that was beyond where the fans were willing to accept.”