Half cool
A little over a month ago I wrote a post about the birth and evolution of the Israeli PC indie game scene. Quite a lot has happened since: On the other side of the pond, a leading game studio was acquired for half a billion dollars.
Being acquired by a publicly-traded company enforces an exhaustive due-diligence process and that shows a vote of confidence from an external party towards the great company the management team and everyone else built here. But this also some figures are being made public through shareholder reports. For example the 201$ Million worth of 12-month revenue as of March 2017.
Systematic, Hydromatic, Ultramatic
How does a company reach these numbers based on a F2P model? Like some other newfound and very successful studios, Plarium built its business model on the new emerging markets of Social and Mobile games, delivering constant game updates and enjoying a long product cycle. Traditional publishers from the pre-mobile pre-social era often struggle with the new world of Games-as-a-Service: In their universe, packaged goods were sold for 50$-60$ upfront while game sales were based on buzz, studio pedigree, demos, the strength of an IP, word of mouth and critical reviews.
Online and digital distribution changed things, and so developers and publishers evolved to the offer post-launch content, whereas instead of a sealed “static” product you provide a continuous service for the player and pump out additional content (which is usually cheaper for you to produce than the upfront investment of the initial game).
In the world of mobile and social networks, the studios had to be even more nimble and calculated, as the mass-market trend slowly shifted toward free-to-play. That required them to be sophisticated: Adopting a programmatic approach to content creation, user acquisition, and analytics is what really made a difference. Knowing exactly where each player came from, what kind of content they prefer to consume, and what they actually spend inside your game is crucial information for a developer in such an ecosystem.
This efficiency means that each dollar you spend on advertising your game should generate a larger return on investment, and each piece of content you produce is better tailored to your audience and keeps them more engaged. And so it’s no wonder that there’s a rising demand for skilled analysts, statisticians, and data scientists in the game industry: The ability to crunch plenty of metrics into actionable insights for anything from game design to marketing proves to be a key requirement for a super successful studio in this brave new world.
The death of game design as a craft?
No, no, I wouldn't go that far.
Good game design is still the most crucial thing for making a game actually fun. Can you, in theory, make a financially successful game without making it fun for the player? I’d argue that except for a few outlier examples this would be really hard (and most of these outlier examples actually showcase a fun game which was ruined by altering it and milking the player base dry).
I know there’s some elitism towards the mobile and social markets from PC and Console developers, and perhaps it’s partly justified because of said “milking” tactics. But sometimes it’s easy to overlook the actual hooks and the mass appeal of a certain game or genre, just because it’s not one’s personal taste.
It seems to me that good game design requires a human touch. I’ve been exposed to a multitude of tools in recent years from AI level solvers that try to evaluate difficulty for human players, to machine learning-based predictions of player engagement and spend likelihood - But such an approach can only go so far without having a good game underneath all of it. Algorithms still need guidance.
In a galaxy not-so-far, far away
Case in point: No Man’s Sky. The space exploration simulator rode high on promises to provide the players with ~18,446,744,073,709,551,616 planets to explore, using procedural content generation. That promise sparked many articles before the game launched, including one in Wired on how “games of the future will be developed by algorithms, not humans”.
But when the game was released later in 2016, things didn’t turn out all that great: It’s not that No Man’s Sky made an empty promise, as much as empty space: a space sim of epic proportions hides the fact that space is mostly an empty void to begin with, and that doesn’t necessarily translate into interesting gameplay.
Eventually, it took about a year of additional manual development post-launch to make the game a bit more worthwhile. Perhaps the fault is with the media hyping up things they don’t thoroughly understand, but as the hailed champion of algorithmic and procedural game design, No Man’s Sky instead turned into a cautionary tale.
Where credit is due
On the other hand, you can still make a successful game purely on the merit of fine game design - without even resorting to sophisticated analytics and content automation and whatnot. I was personally surprised to learn that the developers at Playdead, the studio behind the acclaimed Limbo (and its spiritual successor Inside), didn’t even track even the implemented achievements to evaluate player progression through the game. Imagine that.
So yeah, if you’re as skilled as the guys at Playdead then you might stand a chance at nailing your game design just right, but even then it probably wouldn’t hurt to employ some of the more sophisticated tools of the trade: As acclaimed as the studio is, the unfortunate split between its founders puts Playdead’s worth at roughly 15$ Million. That’s nice, but not half-a-billion dollars nice.
And so, I believe the one takeaway from all of this is to rely on your instincts and experience as a game developer to ensure a fun game first and foremost, but also rely on data science as an essential “multiplier” for that billion dollars unicorn dream of yours. Don’t neglect that.
If you wish to learn more about analytics at a large studio like Plarium, and you happen to be around Tel-Aviv this coming Thursday afternoon, there’s an Israeli gaming industry session taking place as part of the DLD festival.
Tomer Miles from Plarium’s analytics team will be there to present some insights on player psychology and common data analysis pitfalls (among other speakers), and I will personally be there to host and moderate a panel with a few Israeli indie developers. See you around!
