- ARPDAUPosted 2 years ago
- What’s an impressive conversion rate? And other stats updatesPosted 2 years ago
- Your quick guide to metricsPosted 3 years ago
Hubris, ambition and mismanagement: the first post-mortem of RealTime Worlds
I’ve been told it’s too early, but I think it’s worth doing a post-mortem of APB-developer RealTime Worlds. The company was put into administration yesterday.
The administrator, Begbies Traynor, has said:
“Our intention is to continue trading the company while we attempt to find a going concern buyer which will safeguard the future of the business”
I don’t believe that there will be any buyers. I have added the 200 staff in the UK and 42 in the US to the Job Loss Tracker.
A brief recap
Realtime Worlds is the brainchild of Dave Jones, the creator of Lemmings and one of the few people who can claim legitimately to have been involved in the Grand Theft Auto series since the early days.
Founded in 2002, the company’s first title, five years in the making, was Crackdown. Critically acclaimed, Crackdown sold over 1.5 million copies. Up to the end of 2007, RealTime Worlds had generated revenue of £19.6m in its five year life (it has essentially generated no revenue since then, I believe).
As we all know now, that didn’t happen.
Putting money to work
RealTime Worlds has raised over $100 million from investors including emergency funding of $21 million in January this year, presumably after the planned of launch of APB in 2009 slipped to June 2010. It has also generated revenues of £20 million from its Crackdown advance and royalties and, presumably, from the publishing deal it did for APB with Webzen (which it unwound in April 2008 at a cash cost of $10.4m).
Here’s a table of the rounds of funding that RealTime Worlds raised.
That’s a lot of money. What was it for and how were RealTime Worlds able to raise it?
According to Dave Jones, speaking at the GameHorizon Conference in summer 2009, “I went out and told investors that the world was changing, RTW could be disruptive and I told them honestly that I needed $50 million to make a good online game.”
I was sceptical at the time, especially in light of the emergence of cheap-to-make, cheap-to-market (before Facebook shut down the viral marketing hooks, anyway) social games. Dave had a creative track record with Lemmings, GTA and Crackdown. Could he pull off?
APB is a massive fail
Unfortunately not. APB launched in June 2010 to an almost universally negative reception.
With a Metacritic score of 58, reviewers criticised it for “style over substance”, being “confused and weightless” and perhaps most damning of all “a gamesthat shows that Real Time Worlds understood the lay of the land back in 2005 when APB was conceived.” (That last quote was meant as a compliment by the reviewer, BTW).
Still, games have survived worse results if the fans liked it. Perhaps the hardcore audience would be more forgiving than the reviewers and go out and buy it. Here are the sales figures for the US for July 2010:
Back in the summer of 2009, Dave said that with 100,000 to 200,000 players the game would do alright financially. I estimated that each player would have to spend £500 on the game for the investors to make their money back with those figures, which seemed unlikely to me. (That was when I thought the company had raised $50 million. It would be $1,000 based on $100m of investment.)
But 100,000 players, which I imagine seemed a laughably low target to the management of RealTime Worlds, is 10x the number of players who have even bought the game.
APB is turning out to be the games industry’s own Heaven’s Gate.
So what went wrong?
The story of RealTime Worlds is one of ambition, arrogance, mismanagement and hubris. It will haunt that games industry for a very long time.
If it changes the way that we develop, finance and publish games, that will be fantastic for the industry.
If it leads to idiotic calls for government tax credits to “rescue” the best funded games developer in the UK, it will be a disaster of epic proportions.
There is an excellent deconstruction (albeit with the benefit of hindsight) from an ex-RTW employee in the comment thread of RockPaperShotgun. You should read that (and thanks to Ed Barton for the link), but I will add my analysis here.
Note: I am not analysing the gameplay here, but the company. For the record, I have not even played APB. If you want a discussion of how the *game* could have been better, you would be better looking elsewhere.
Error 1: No business model
Dave Jones made a virtue of having no business model for APB. He said “if a game is built around a business model, that’s a recipe for failure.”
A game and the mechanisms of paying for a game are intrinsically linked. Historically, there were two business models (buy a game in a box, subscribe to it). There are now dozens. The design requirements for a free-to-play microtransaction game are radically different to those for a subscription MMO.
By divorcing the gameplay from what people pay for, the whole reward structure and value-for-money feeling for players come unglued, with disastrous results.
Additionally, by being opaque about the business model for a very long time, RTW allowed gamers to create their own expectations. Charging people with a pay-per-hour concept, after they’ve had to pay full retail price for the original game , was always going to be a hard sell.
(I know there are different ways of spinning this model, but in the end, a free-to-play game generally succeeds by allowing users access to status/emotional/performance items; a paid-for game sells content. APB was a strange and unpopular hybrid of the two.)
A game designer’s job is now at least 25% about building the business model. If you don’t want to do that, you have no place designing online games.
Error 2: Taking too bloody long
That reviewer who credited RealTime Worlds with knowing how the land lay in 2005 got it right. In the time that the company spent tens of millions of dollars building a WoW-beating variant of CounterStrike, the market moved on.
Facebook games. Free to play. GTA IV. Modern Warfare 2. These changed the game. The failures of Tabula Rasa and Auto Assault showed valuable lessons. Heck, the lessons of the ten biggest failures of the last decade could have been learned.
But the development of APB meandered along, and the world changed around it.
Not only that, it was expensive. Between 2005 and the launch of APB, the company spent £25 million on staff costs. Taking too long is very, very expensive.
Error 3: Misunderstanding the try-fail-iterate model
The new mantra of game development is try-fail-iterate. Launch a game, see what users like, tweak it and fix it. That way you’ll get a game that appeals to users by testing it on them.
The anonymous ex-RTW source on RPS says:
That model probably worked fine on Lemmings. It works fine for a game on a social network or the wider web where the initial launch budget is less than $1 million (often much less). It doesn’t work when you have $100 million to throw at wasteful ideas and, above all, when you are not getting transactional analytics of what your users are actually doing in a live environment.
Try-fail–iterate only works if you launch early. Very very early. Not five years late.
(Guy Kawasaki has a great post entitled Why Too Much Money is Worse than Too Little, if you want to find six further reasons why this can be a real problem.)
Error 4. Making a game he wanted to play
Dave Jones said that he was trying to recreate Counterstrike for the GTA generation (or words to that effect).
To my mind there is a big problem with that. Counterstrike is one specific style of gameplay: head-to-head competitive co-op. In contrast, World of Warcraft satisfies people who want to be Achievers, Explorers and Socialisers, not just Killers. (Check out the Wikipedia entry on the Bartle Test if you don’t get what I’m talking about).
In short, a good MMO appeals to many types of people, not just one small sub-segment. As I said, I haven’t played APB, but it seems very focused on single gameplay style. Please correct me if I’m wrong.
Error 5: Hubris
RealTime Worlds went *big*. They went big with their fundraising. They went big with their “game-changing” statements. They went big with their product.
Just at the time when most other people were finding ways to do well, but small. To polish the heart of a product, usually the gameplay, and then add features over time. In short, RealTime Worlds built a boxed-product game with a service-game approach added on top. Neither were done well.
And I have to say that I think that the hubris blinded them to changes in the market. I disagreed strongly with most of Dave’s keynote at GameHorizon. I thought that he was learning the wrong lessons from his success, not the right ones.
It may seem that I take pleasure in the demise of RealTime Worlds. I don’t. There are 250 game developers out of work and investors have lost over $100 million of their money. It will be a long time before a UK game developer gets significant funding for an MMO again. (In fact, I think it may be an eternity – we may never see a developer funded to build a traditional MMO with a boxed product component again,)
But I want the right lessons to be learned.
- Tax credits would have made no difference to RTW. Stop pretending they would and accept that this was a badly-managed company that took a big, well-funded bet. And lost
- Realise that the try-fail-iterate model doesn’t work with massive budget titles. Have an approach that builds on success (and failure), not bets everything on day one launches
- Accept that a massive budget for a single game in a single company is, more likely than not, going to kill that company
- Above all, watch the market. if it changes around you, change with it. Launching a game that knew the lay-of-the-land five years ago is always going to be fraught with risk.
I wish everyone at RealTime Worlds good luck finding new work. If you start-up a new company, go for it, and the best of luck. Drop me a line if you want to chat.
RealTime Worlds was a dinosaur. Its era is over. It will be missed, but won’t be lamented for long.
(If you have anything to add, please comment below. I think that we will be dissecting the lessons learned from RealTime Worlds for a very long time.)