Feedback Loop

For better or worse, platforms such as the iPhone and Facebook are changing the way we develop, publish and even think about videogames. Despite heavy skepticism only a handful of years ago, most industry executives' eyes now light up when you mention these emerging platforms and new business models.

Not all of them, of course, actually know how to leverage this new potential, and some of the businesses they helm will be dragged under in the coming years by a failure to adapt to the new realities of the market. Capitalism, like evolution, isn't always simply a matter of survival of the fittest; the ability to adapt rapidly to changing circumstances is the true key to long-term success.

In all of the talk about how new platforms are changing the business of games, however, we often lose sight of what may be an even more important issue - the way in which these platforms are actually changing games themselves. Any gamer can tell you that Facebook games are very different beasts from their console counterparts, or that iPhone games are slowly evolving a rule-book of their own which is quite distinct from those of previous titles. What's less immediately apparent is how some of that thinking is feeding back into more traditional gaming platforms, and introducing sweeping changes to how games are developed.

Perhaps the most interesting change is the introduction of a feedback loop to game development, something which has been attempted in the past in various genres but which has finally found its feet on services like Facebook. The old development model contained relatively little feedback - you created a game, put it on shelves, and with the exception of a handful of reviews in the press, pretty much the only feedback you got was your sales figures.

Publishers and developers attempt to compensate for working in the dark in this manner by introducing a variety of feedback loops into the development process itself. Closed beta tests, focus groups and even demos, in some cases, provide a modicum of public feedback. Companies like Bungie have created powerful, elaborate test suites which can provide in-depth analysis of the behaviour and reactions of players in their games ahead of launch. Journalists with some insight into the development process have found a lucrative sideline in consulting on games in development, being paid to clearly articulate problems in private in order to avoid public bashing from their peers a few months down the line.

All of these things, however, are mere stopgaps compared to the way in which feedback is built into the working processes of Facebook game specialists such as Playfish and Zynga. For these companies, the feedback loop isn't simply an added extra - it's the very core of how they do business.

The reason for this is that the platform on which they're working, Facebook, is a two-way street. They aren't simply pushing game content to their players - they're also receiving details of all of those players' interactions with that content. In the untold gigabytes of data which move through the servers of games like Farmville, Pet Society or Mafia Wars on a given day, there is a mind-blowing level of detail about player behaviour.

With the right tools, developers can dig into the data and discover how long players spend on each page, which buttons they click and which they ignore, whether different social groups or nationalities play the games differently, or whether there are serious bottlenecks in any of the game processes which seem to stump players.

Even more powerful is the potential to essentially "focus test" new ideas in a transparent way. With a few million players, it's possible to test small changes - perhaps subtle differences to the colour scheme, designed to lead the eye more naturally through a layout, or alterations to how a game mechanic is presented - by rolling them out quietly onto the accounts of a couple of percent of your active players.

The feedback from those players then tells you if your tweak is successful, giving you the data required to make a proper decision on whether to roll it out to everyone - a far cry from the stabs in the dark to which developers have become accustomed. You could even focus test a dozen variations of a game change, and use the data gathered to find out which is the best of the lot - a surprisingly organic and evolutionary approach to game design which has been used to great effect by some of the world's biggest web game developers.

Facebook game designers are on the vanguard of this movement - joined, in some cases, by MMO developers, who can theoretically access the same kind of data, although I'm consistently surprised by how few of them actually invest in building great data-mining tools for the wealth of information their servers hold. However, developers on all platforms are starting to wake up to the potential of this approach.

What has changed, of course, is connectivity. PCs have been connected devices for years, and platforms such as Steam have greatly enhanced that situation. Consoles are now also seen as connected devices, players' experiences even in single-player games now being connected to the rest of the world through Xbox Live or PSN.

This creates the possibility, for developers, of starting to work with the same kind of feedback data - "telemetry", as one leading UK developer described it to me this week - which Facebook developers have been using for the past couple of years. Instead of throwing a finished game into the abyss, developers can now make their products phone home and let them know how they're getting on - sending back vast amounts of data which can be mined to inform the development of patches, DLC or sequels.

This isn't merely a minor change. It's a fairly fundamental upheaval in game development, one which neatly dovetails with the new business opportunities offered by DLC, subscriptions, micro-transactions or in-game advertising - all of which are strong incentives for developers to regard the launch of a game as merely a major step in an ongoing process, rather than the end of the road for a project. Feedback systems mean that not only do developers now have the commercial push to keep working on games after launch, they also have the tools at their disposal to do so intelligently.

That, of course, is going to require some adaptation from developers. It's all too easy to characterise publishers as the industry's dinosaurs, clinging desperately to the boxed game business models of the past, while developers experiment with the potential of the present and sometimes even end up discovering that they don't need a publisher at all. The "death of publishing" is a recurrent meme in the industry of late, and while it's an extreme view, it's not one without some truth at its heart.

It is, however, wrong to assume that all publishers are incapable of embracing change - and even more wrong to assume that all developers are welcoming it with open arms. Developers can be a surprisingly conservative bunch, an inevitable symptom of working in a business which has been so hand to mouth over the past couple of decades. The potential of the feedback loop is immense, but it also requires significant effort - retooling the business processes underlying the development process, developing the data mining software required to make sense of the telemetry itself, and investing in the staff who will become your feedback experts.

For a developer whose eyes are firmly locked on hitting today's milestones, and already gnawing their fingernails over the schedules for tomorrow's, that's quite a tall order. Those developers, sadly, are every bit as stuck in a rut as some of the industry's slower-moving publishers are - and every bit as unlikely to survive the coming years, as a consequence. New business models, new platforms and new technology opportunities, such as feedback loops, aren't just the stuff of idle chatter over canapes at industry receptions any more - they are increasingly becoming the bread and butter of how this industry makes money, how it continues to grow, and how it builds for its future. An understanding of that is becoming essential to surviving in a market that has little mercy for stragglers.