The Compulsion Loop in Game Design Explained
A man can surely do what he wills to do, but cannot determine what he wills.
- Arthur Schopenhauer
John Hopson introduced the concept of the compulsion loop applied to video games as early as 2001. Since then, compulsion loop mechanics continue to get integrated into video games fairly broadly. The compulsion loop concept regained popularity in the 2010-2012 period especially by companies such as Zynga for social gaming. However, I believe we will see increased focus especially in mobile gaming.
To date, there hasn't been (as far as I can tell) a very compact and easy to understand coverage of the compulsion loop concept. Hence, this post attempts to detail the concept, include my own interpretation, and describe the potential future opportunity.
1. Definition
My definition of the compulsion loop differs slightly from the definition put forth by Adam Crowe as follows:
Compulsion Loop: A habitual, designed chain of activities that will be repeated to gain a neurochemical reward: a feeling of pleasure and/or a relief from pain.
You should understand three key concepts comprising this definition:
Habitual: The loop should promote a long lasting and constantly repeated habit
Designed Chain of Activities: The compulsion loops should consist of a set of specifically designed activities within each step in the chain
Neurochemical Reward: Compulsion loop theorists believe that human free will does not exist and that the creation of habitual behaviors can be instituted and programmed
With respect to neurochemical reward: Crowe explains that the compulsion loop derives its power from basic human elements of psychobiology and neurochemistry. Hence, the point of a compulsion loop is to induce a biological response such as the release of Dopamine to help train behavior.
Dopamine is released... as a result of rewarding experiences such as food, sex, and neutral stimuli that become associated with them.
- Wikipedia 2014
2. Concept Origins
Much of the thinking behind compulsion loops stems from BF Skinner's psychological studies on animals such as rats, pigeons, and chimpanzees. Skinner invented the concept of an "operant conditioning chamber" aka "Skinner Box". Skinner used the operant conditioning chamber to study animals for the purpose of teaching animals specific behaviors.
The image above shows Skinner's operant conditioning chamber. Here, the researchers attempts to train specific behaviors to the rat such as pressing the response lever to obtain food. The studies by Skinner concluded with specific observations particularly around contingencies. A contingency describes "a rule or set of rules governing when rewards are given out."
In particular, we should consider two issues regarding rewards:
Ratio: How much of a reward to give based on an activity?
Interval: How long to wait between giving rewards for an activity?
The application of the compulsion loop in games came from the notion that we can map the loop of activities within an operant conditioning chamber to game loops. Therefore, we can think of a user in a game loop just like a rat in a Skinner box.
Hence, the concept of compulsion loops has historically been tightly linked to Skinner but, in my view, are not equal. In my view, a compulsion loop need only support the 3 core principles of the definition I included above. Further, I believe additional mechanics or tactics can be applied outside of Skinner's original research.
3. Compulsion Loop vs. Core Loop
There seems to be some confusion in the industry about the difference between a compulsion loop and a "core loop." Although often used interchangeably, I differentiate them as follows:
The core loop defines the chain of activities associated with the primary user flow. Hence, the actions the user does over and over again.
I show an example below of what I would consider a core loop in a typical RPG game. Further, we can equate this particular behavior to the behavior of a rat in an operant conditioning chamber: press lever -> get food -> satiate hunger.
However, the difference in terminology now comes from the design of the various phases within this particular loop. I like to think of the steps within this particular core loop as comprised of three phases:
Anticipation: The user anticipates some desired user state. For the rats, they seek to satiate their hunger. In an RPG, it could be getting enough gold to buy the Holy Avenger +5 Sword or enough power to kick someone's ass in PVP.
Action: The specific activity we want to incentivize and condition as part of the overall behavior.
Reward: The part where we give the user a reward for doing the specific activity.
4. Compulsion Loop Application to Gaming
As John Hopson describes in his post on Gamasutra in 2001, a number of Skinner research conclusions can be applied to gaming and other application areas where we want to condition behavior. Here is a summary of some of the key conclusions:
Source: John Hopson 2001 via Gamasutra
5. Innovations From Gaming
Skinner's research focused primarily on the Reward phase within a compulsion loop:
However, we have seen examples in social and mobile gaming where additional advances have been made.
Before proceeding further stop and think:
What other features have we seen to optimize compulsion loops in some of the other activity areas? e.g., in Anticipation or Action phases?
The below diagram shows an example within each of those other activity areas:
Aspirational Neighbors:
Designers anchor an objective or desired user state into the mind of the player. See what you could become in the game! A couple examples of this from the folks at Supercell:
Clash of Clans: Providing easy access to visiting the super high level cities of other players in the Leaderboards
Hay Day: Forcing players to visit "Greg" as part of the tutorial and see his more pimped out farm
Clash of Clans:
Hay Day:
Another common tactic, often used in RPGs, puts the player in the role of a high level user at the very beginning of the game. This gives them a taste of what they could later become before then stripping them back to being level 1.
Variable Design in Action Phase:
Giving players a "variable ratio" action can make games with PVE (player vs. environment) more interesting. And, even further, making the variable ratio action a type of reward. In Final Fantasy Air Brigade, for example, PVE can include not just a typical battle but has a chance of encountering and capturing a Chocobo as well:
Many card battle games now routinely mix PVE with PVP or other types of encounters.
5. Future Opportunity
So where can we derive future opportunity with this?
Compulsion loop design can significantly drive retention. Further, some games are purely compulsion loop driven. Examples include Slots games and progression/completion type games such as Mafia Wars or Farmville. Understanding how to design new tactics here can drive a winning game and open up new opportunities. In the mobile game market today, the smallest advantage from a retention or monetization perspective can make or break a game's economics.
Significant innovation can come from taking a closer look at compulsion loop principles and applying them in smarter ways. This doesn't just apply in gaming.
As an example, how can we apply with what we know about variable rewards to restaurant loyalty rewards programs?
Yo Pinkberry that last free froyo should be variable!
We will likely see more compulsion loop innovation especially in mobile gaming where the stakes are so high. Also in mid-core and hard-core games where retention can be especially poor. Hopefully, better design at the compulsion loop level can help avoid situations like Game of War, where they stick appointment mechanic everywhere in the game to do whatever they can to retain players:
How many timers do you see above? :-)
I believe new disruptive compulsion loop tactics can and will be designed.
Currently we are just limited by our imagination. What new idea do you have to create a stronger compulsion loop?
6. Sources and Additional References
Behavioral Game Design via Gamasutra by John Hopson
Gamification Sucks video presentation by Stephanie Morgan
Mid-Core Success Part 1: Core Loops via Gamasutra by Michail Katkoff
The Skinner Box video by Extra Credits
Update (3/23/2014): Thanks to Johan Holmberg for providing some debate and additional links for this blog post.