Two Questions to Ask Your Free to Play Game Designer

The free to play game industry has a strong culture of metrics-driven decisions, but metrics do not tell you how to do game design. I propose two avenues of inquiry for improving short and long term retention that treat your game like a game.

Metrics are a crucial part of managing a free to play game, but translating metrics into action items is not as simple as noticing whether a statistic is below the desired value. My friends and I call that step “which line is longer,” because we once saw a game on Facebook called Which Line Is Longer. (Players were shown two lines and asked to determine which one was longer. Sadly, this is not a joke.) So D1 retention is 70%? Break out the Dom P, you just struck gold! D1 retention is 20%? That sucks a lot and you need to fix it or find a new job. You don’t need a PhD to know which line is longer. So what do you do now?

There is no silver bullet. That’s why you have to be smart and tireless to succeed in games, and what worked in one context might not in another. While there are no one-size-fits-all solutions, I think there are questions with enough generality to guide you more quickly to good answers. Metrics are the most general form of that enterprise: they tell you nothing about the solution, but they do tell you what to start solving for. Game design, as a discipline with an object of inquiry and the goal of accumulating knowledge about it, should be able to give us, if not the answers, better questions.

So here are two questions to ask yourself or your game designers. Actually, maybe this is four questions. I’m not a stats guy, OK! Here are two axes along which to analyze your game, with each axis implying two extremes to consider.



This is a simple and perhaps obvious point but I think it is a profound one. The question is, how does your game relate to time, both in a session and over a lifetime? Casual, free to play games can be distinguished from commodity games (the ones you purchase up front and own) by the graphs they create in response to this question. A casual game should be a parabola. It should support very short play cycles, supporting even the activity of merely checking into the app to collect some stuff, and it should support very, very long player lifetimes. There is no need for it to support or encourage long (let’s say an hour) sessions, and most tend to kick you out--or expect you to pay--after a reasonably satisfying amount of play time.

Now let’s compare that to a typical console FPS RPG that you pay $60 for. You will never play that game for only a minute. It takes a minute just to load. You are going to play that game for an hour, minimum, and if you can get the time you will go for twelve hours. And then, it will end. You will reach the end. (Statistically speaking, you are unlikely to reach the end--most people quit before that--but you get the idea). Even games that don’t require you to stop tend to reach an exhaustion point. I took a vacation day to finish the main questline in Skyrim and then quit cold turkey because I needed to be done. And business-wise, it’s no different to the developer. Once I plunk down my dollars they only need to hold up their end of delivering what I expected--which is measured more in quality rather than quantity--and everyone is happy. I got what I wanted from Skyrim and will be strongly disposed to buy whatever comes next from Bethesda because they gave me what I want from a commodity game.

Free to play games don’t work like that. Give your players a great 60 second experience and a reason to keep coming back for a year. These two extremes target the most important zones of being ROI positive. You will lose a bunch of players very quickly. Having a solid short term experience will cut down on attrition at the point where the most people are affected. But only a few of those will become funders. You need to keep those people engaged for a long time. In one game I worked on, something like 30% of our total revenue came from people who had joined in the first month and stayed with us for a year. Keep those players entertained for a long time.


Complexity is on a continuum with simplicity, and as with the extremes of the temporal axis, both have their place in a F2P game. Your game should have a very simple rule set so that players quickly grasp what they are trying to do and the critical path to success, while also having a degree of mathematical complexity that over a long lifetime they will not feel they have solved the game. (Solving a game is another name for being bored with it). The importance of simplicity in free to play is abundantly clear from the top grossing games on iOS, but I think it can even be illustrated by something like League of Legends. At a glance, League of Legends is easy to understand: you drive up a lane until you destroy their base. You need to balance attacking with retreating to heal. That’s the first tutorial and it gives a deceptive, but empowering, sense of comprehension. Then you go play some bots and lose and/or get yelled at for being an idiot. (I’m being autobiographical here). You realize that, while you know the objective and the means, there are myriad variations within that trajectory. Some lead to mind-blowing victory. Some lead to abject defeat. Now you have something--a lot--to learn, but you remain oriented toward your goal by a succinct initial experience.

Conversely, I would argue that the long-running success on iOS also offer a significant degree of complexity in their late game. Here, however, it is more likely to be delivered as modular content than alterations to a holistic ecosystem. Candy Crush and Angry Birds have a bazillion levels, each slightly different. The sum of those variations is its total complexity--the total number of possibilities to consider--in the game. (In this regard Candy Crush can be contrasted with the Bejeweled products, which offer a very small number of discrete game modes).

Clash of Clans, like League of Legends, has a multiplicative model of complexity. Whereas each Candy Crush level is separate from the others, new content in Clash or Leagues is multiplicative because a new unit or item needs to be compared to other possible units and items that could be used instead. Adding one new champion or one new tower triggers a relative recalibration of the value of all other units with regard to the new one. This is good and bad--you get a lot more complexity per content update, but when it accumulates, as it does in LoL, you also tend toward a game that is more daunting for new players. (One more reason the rotating availability of free champions is smart. As Magic: the Gathering matured, it handled this problem by creating different tournament modes that constrained the available set of cards to a subset with a lower barrier to entry). CoC unlocks new units linearly so you are slowly exposed to the complexity of the total system. Over time it becomes a mathematically complex game. My contention is that if it did not increase in complexity--if it offered only simplicity of rule set--Clash of Clans would, over time, lose the players that are keeping it at the top of the grossing charts. It would slide down into the endless heap of infinite runners chasing the imperative to keep it simple.

I have compared a hardcore game (LoL) with casual games to make the case that free to play games have common qualities, mainly the need to quickly engage a player who has made no initial investment and to retain their best players for a long time by giving them a persistent challenge. That said, I think the absolute values for complexity vary widely by how “hardcore” a game is. (These definitions are circular--a functional definition of a hardcore game is that it is more complex). After playing Clash of Clans for months it is still a less complex system than the build of a single champion--and League of Legends is considered the most forgiving MOBA. Despite these differences in scale, I still think both games are exhibiting similar distributions of complexity. And if you are wondering why your game is doing well on D7 but not D30, D60, D90--well, you might not have a complex enough system.

Temporality and complexity, then, go hand in hand. Asking about temporality leads to asking about your simplicity/complexity curve. Whether you are challenging your game design during preproduction or trying to correct deficient stats after a test campaign, these question can hopefully you make the important lines a little bit longer.

Latest Jobs

Double Fine Productions

Hybrid, San Francisco CA, USA
Senior Systems Programmer

Purdue University

West Lafayette, IN, USA
Clinical Assistant Professor in Game Development

Digital Extremes

Lead AI Programmer
More Jobs   


Explore the
Advertise with
Follow us

Game Developer Job Board

Game Developer


Explore the

Game Developer Job Board

Browse open positions across the game industry or recruit new talent for your studio

Advertise with

Game Developer

Engage game professionals and drive sales using an array of Game Developer media solutions to meet your objectives.

Learn More
Follow us


Follow us @gamedevdotcom to stay up-to-date with the latest news & insider information about events & more