BEM Gamification: Measuring Motivation (part 1)

Any gamification project will have a significant number of KPIs, specially when you are not applying gamification to an already existing project. Are you players learning? Are your players enjoying the interface? Are your players using the right behaviors? Are your players taking advantage of the system functionalities? But a question you will have to tackle every time is: are your players being motivated by the gamified system and how much?

In the early days of Gamification, Zichermann talked about measuring the big E, for engagement. I find talking about measuring engagement difficult, as digital analytics are already using this term for some metrics, like frequency of visits, bouncing rate and so on. I would argue this is not engagement from the point of view of a game designer, but why enter the discussion when we can just talk about measuring the big M: motivation. Is a player who returns frequently to a site being motivated? Yes, but a player that won't return to the site regularly might be too. Take for instance a player who can't afford a game but seek every YouTube video about it. That player is being driven by the game, but she is showing compensatory behaviors and channeling her motivation towards other content. Or maybe the player has the game but is more engaged with the "meta game", and is browsing every wiki she can get her hands on, so her engagement metrics might be "poorer".

Engagement is actually a triad composed of motivation, emotion and attention. In the century of the attention wars, many are focused on measuring attention, as this is a big monetization driver. Meanwhile, emotion has been treated as a means: what to be viral? Enrage people! If emotion was a KPI maybe we could live on a better world, one with less biased algorithms and AIs. But motivation is seldom measured or taken into account, as tracking motivation seems difficult, at least for the guys in marketing. I want to give some understanding on how motivation can be measured, specially for you who are working on gamification projects or general applications for change.

Let's begin by understanding motivation and why it is so difficult to measure. For this, a like to use the quote from Jefrey Nevid:

Motivation is the process that initiates, guides, and maintains goal-oriented behaviors. [...] Motivation doesn't just refer to the factors that activate behaviors; it also involves the factors that direct and maintain these goal-directed actions (though such motives are rarely directly observable). As a result, we often have to infer the reasons why people do the things that they do based on observable behaviors.

When we talk about motivation, we can't directly give the reasons for the behaviors, so different motivational theories and frameworks have been dealing with understanding taxonomies for this reasons. Maslow was the first to give such a taxonomy, but we currently work heavily with Edward Deci's Self Determination Theory, which gives as Mastery, Autonomy and Relatedness (yes! Purpose is not original to the theory as it is understood as a part of Autonomy). From there you can find things like Octalysis' eight core drivers, or Appelo's Champfrogs, or even my own BEM seven drivers. This taxonomies are useful as they can help infer why are people doing stuff: we can't still pinpoint the reason, but we can give several informed hypothesis.

In this sense, one of the earliest ways of measuring motivation in game design was actually Bartle's 4 player types. Bartle's study was an effort to classify players according to their predominant behaviors according to four play styles: seldom did he knew that he was working in a similar framework as SDT: achievers are mastery driven, socializers are relatedness driven, explorers are autonomy driven, and killers, well they are still mastery driven, with an specific empowerment drive. Bartle's study was the first real attempt to measure motivation, by making "behavior clusters".

I'm not sure if behavior clusters is a thing, but it is a thing I use. The idea is to categorize behaviors into groups that can relate to different drivers. Why? If we take a student that loves solving math problems, we might think it can be related to mastery, but we could also explain this as a way to express autonomy (she thrills in the discovery of new ways of expressing herself through numbers) or even hedonism (she finds equations beautiful). Most individual behaviors can be explained by almost any driver, which becomes a "list of reasons" why someone might engage on an activity. And believe me, doing such lists is a great way to train yourself when designing gamification systems (that's part of the core of BEM). However, that is not great for measuring motivation. But if we know that the same student loves to solve puzzles, and is driven by hard chemistry problems, we can start discarding possible causes, or at least give a more probable explanation.

Let's leave that for now, as this only solves part of the problem: we can't answer why a player is doing something, but we can try to infer it better. But now we need to actually create a measure or scale. There are many indicators of motivation, but I usually work with three: energy, perseverance and direction. I will try to make this as pragmatic as possible:

Energy and Direction

If a learner is not driven by doing an exercise, but is driven to avoid being punished with a grade, you can actually measure this by the level of optimization of the task. When our brain is pushed towards something painful or boring, it will try to find the quickest way to get a positive result that will avoid the bad outcome. This example can help us illustrate two measures through reading the perceived quality of the task performed and then using the pain-desire scale.

When our brain is engaged with the activity, and not an extrinsic pressure, that energy will probably go towards doing the task better, as the quality of the job done will be a measure of fulfillment. But when our brain is being pressured by an extrinsic goal, it tends to save energy and becomes creative in finding ways of getting to the goal without having to do much effort. Is a player who tries to get all three stars in Angry Birds more driven? Maybe. We need to understand our previous taxonomy to actually give a better answer: if we measure this behavior through the lens of mastery, then yes, the player is more driven, but if we measure it through the lens of discovery and exploration, then maybe not, it is just driven by a different driver. Energy can't be measured on a vacuum, as we need to know where that energy is channeled. A student that cheats in an exam is not driven towards answering the exam, but she is driven to spend energy to find better ways to cheat.

This takes us to the other indicator: direction. In academic terms, we have three directions: apathy, avoidance and approximation. Apathy happens when the brains just doesn't care or doesn't care enough to try to get to a goal. If you place some reward behind an enemy but the player just decides to skip it by avoiding the enemy, we can say that that player was probably apathetic towards that reward. You can actually measure this and it can be convenient to find better ways to create drive towards a goal, but be careful, this apathetic attitude is not the same as failing in your design, but more on that later.

Now if a player is trying to get a big coin guarded by several obstacles, even when she could try a simpler approach and avoid the activity, we can infer she is driven towards approach. Approach is better explained as a goal that is arbitrary and not required, but that generates desire: approach is psycogenetically intrinsic as the decision is presented externally but its not pressured, it stems from wanting psychological satisfaction. This is the hardest direction to grasp in gamification projects, as the designers usually want to pressure behavior, but it actually is created by suggesting optional goals. To measure approach direction, then, we need to create optional goals that players might not interact with in the game.

Finally, avoidance is about removing a pain or avoiding a future pain. Most games have some sort of enemies and health bars to create this desire for avoidance. Curiously enough, avoiding probable pain is rewarding in itself, but not being able to avoid it might be demotivating. What is important to measure in this sense is not if there is avoidance drivers present in your design (they will surely be), but if those pain points are creating the desire to retry or to leave the game. Every pain you create in you system (losing XP, losing a challenge, being hit by an enemy, being surpassed in the leaderboard) has a probability of making you lose a player. How you manage your frustration threshold is important, as lowering this odds can actually hinder the experience for players driven towards approach through mastery. You have to select a range of punishment and difficulty and you will never be able to make it universal for every player.

Probable? End of part one.

As you can see, motivation is a game of probabilities. We can gather queues, but we can never get the driver itself, even if we ask for it. Self-knowledge and self-awareness are actually skills, and most people can't understand why they are driven by something, as we have not been taught to understand this in our schooling system. So, our measures of motivation cannot be given in absolute numbers. Knowing that we have 1000 players will not help us to understand motivation, but knowing that out of 200, 30 players are still in the game can give us some light. In my next article I will talk about perseverance and understanding motivation as game of probability, and where to put your expectations. Hope to see you next week!