Games have long been characterised as being, needing or to be, or not being, fun.
Often in game development we’re told that we need to “find the fun” - to stumble across that magical formula that your players will find enjoyable enough to continue playing. That can have varying degrees of meaning across different teams, studios and projects, but the underlying message is that a player needs to find the experience fun enough to continue playing.
If we want the experience to be fun we need to then be able to tell if it’s actually fun or not, and for that to be consistent, it needs to be measurable.
The Problem(s) with Fun#
Let’s take a moment to identify some issues with fun and what makes it a difficult thing to measure and hard to reach as a goal.
Fun is subjective & Can’t be universally defined#
What one person finds fun, another doesn’t. If I describe an activity that I find to be fun there’s a reasonable chance you won’t because that’s just not how you have fun. For instance, I have friends who find a lot of fun in competitive play, however I absolutely do not. These differences between as few as two people makes designing an experience that a wider audience would agree is fun, exceedingly difficult.
Fun can’t be measured or quantified#
It’s impossible to measure how much one person enjoyed something. My definition of enjoying something “a lot” might be completely different than someone else’s. This means that because we can’t properly quantify how much fun I had, we can’t say how much fun was enough across a sample group, whether we need to increase the amount of fun being had, or even how to go about achieving that.
Fun is fleeting#
Odds are, the most fun you’ll have with a mechanic is when it’s new or fresh in some way. Repeating an action in gameplay many times inevitably leads to the mechanic becoming routine and inherently less fun. What’s worse is that any friction in or around the mechanic will become more grating with repetition. This leads to mechanics that were once fun, feeling dull and tiresome - and is often the cause of mechanics being changed or iterated on across development as leadership sooner or later finds that the mechanic isn’t as fun as they used to find it.
Fun isn’t universal#
Not all games are supposed to be fun. What if you’re making a horror game or a narrative experience that is thrilling and emotionally charged? In those cases, and others, fun isn’t the metric you want to chase at all.
What is Engagement#
Engagement is how long and/or deeply a player has interacted with the experience or its mechanics. It doesn’t require an emotional investment, though that can certainly affect it, but instead it grows from mechanics that feed back into the game, encouraging players to continue.
The statement “just one more run” in a roguelike is an excellent example. In this case the player is strongly engaged with the experience and wants to continue. It’s possible they’re having fun, it’s possible they’re not, but we know that they’re engaged. If they’re engaged, does it matter if the experience is fun or not?
Measuring Engagement#
To use engagement as our primary metric we need to be able to satisfy all of the conditions we identified fun as failing at, so let’s rewrite them for engagement.
Engagement isn’t subjective & Can be universally defined#
If a player interacts with a mechanic, they are engaging with it. It’s very easy to measure as it’s a binary condition - either the player is or isn’t engaging with the mechanic.
Engagement can be measured and quantified#
How often did a player interact with a mechanic? How long did they do it for and did they return to it again later? Did they actively seek it out or avoid it? Identifying and measuring these aspects will let you tangibly say if a mechanic is engaging or not, and how much.
Engagement isn’t fleeting#
Contrary to the statement about fun, mechanics that you engage with repeatedly over time should be at least as engaging as the first time. Building a mechanic that grows over time, like a crafting system, should become more engaging as players interact with it. If not, you now have a data point to address a mechanic that players don’t find engaging.
Engagement is universal#
Engagement applies to every kind of game and experience, but in different ways. In the case of a linear narrative experience the player engages simply by playing. The longer they play the more engaged they are. Other types of experiences can be measured differently, but the details will vary from case to case.
Using the factors and metrics laid out here we can reliably identify and measure the ways in which players engage with and experience and adjust it accordingly.
Instead of trying to measure how engaging an entire game is, I suggest starting with individual mechanics or features. Once those are engaging, measuring the overall player engagement across a play session becomes much easier and will help identify where players aren’t engaging with the mechanics you want or expect them to, allowing you to identify and correct the problem areas.
The Dark Side of Engagement#
Engagement can be a very useful metric to use when you’re achieving the goals of the experience in your game, but it can also be used to trap players by compelling them into playing longer or returning more often than they otherwise might.
Lottery mechanics like loot boxes or random rewards, in-game currencies and game mechanics that appeal specifically to addictive or gambling behaviours are all examples of the ways in which engagement gets used to keep players in a game.
The term “forever game” is often synonymous with these types of experiences. Studios like forever games because they represent an ongoing revenue stream with relatively little risk when compared to creating a new experience or IP (Intellectual Property). This is because finding the fun is inherently expensive, time consuming and fraught with failure. So instead, they aim for engagement to keep players in their particular experience for as long as possible.
Engagement Doesn’t Eschew Fun#
It’s important to note that while I advocate for using engagement as a core metric instead of fun, it doesn’t mean that fun doesn’t have a place in how you assess the experience of your game. It may be that fun is an integral part of the player’s engagement and that you will have to measure anecdotal evidence of what players have or haven’t enjoyed, considering repetition and prior responses to help metre your
Now, you might say that “Well, if all of these engagement factors add up to a fun experience, aren’t we still just measuring what’s fun or not?” and that’s a fair point. However I’d counter by saying that unless we can individually measure and assess all of the components of the experience, we can’t possibly have a reliable way to know which is or is not achieving what we want from it. If we jump straight to a conclusion of “this is fun” without understanding the measurable, and thus tunable, parts that got us here, we can never reliably improve the experience. We cannot make it more fun.