Gamification BEM: Choice Architecture

One of the main concepts of Game Design (and Behavioral Economics, for that matter), is Choice Architecture. After years of data on how people make choices, we have come to conclude that decision making is something that can be designed, but the design principles behind it are often not that clear. After all, designing a Gamification platform is about making choices over how the system should operate, so everything comes down to that simple fact: gamification is not about behaviors, but about decisions.

Let's begin by understanding the difference between an imposition, a nudge, and a choice.

Imposition and Dominant choices

When the path to success for a company depends on the users acting on a predictable way, impositions tends to be the go-to design principle. Imposition is built upon restrictions and penalties, which is what the brain understands as work and task. Impositions are great to design behavior because of the loss avoidance effect: if there is pain in not accomplishing a task, people will tend to finish that task. But the effects of imposition are often negative in the long term, as it creates pain in the autonomy driver, as people feel disempowered and feel obligation, rather than excitement.

There are actually three ways to design for imposition:

1. Tasks with penalties: if you don't do the task, there will be consequences. The most effective way of creating behavior, but the one that creates the most resentfulness.
2. 0-choice or linear designs: you can take the choice out of the equation: there is only one possibility. How can there be just one possibility? Because the rules of the system and the goals prevent another thing from happening. This is how you create linear games: Mario can't advance if he does not grab the pole at the end of the level.
3. Dominant strategies or illusion of choice: you give a choice to the players, but the rewards are so obviously disproportionate that it is impossible to take any other course of action.

While I recommend avoiding 1 and 3 in gamification designs, number 2 is actually important to create narrative paths and direction. As long as there are choices along the way, you are able to forgive a number of constrains.

Nudges and persuasive Design

Behavioral economics brought this kind of choice architecture to the table, by understanding the inner procedures behind how the brain makes choices. Nudges are about giving information that can be easily avoided and that doesn't require you to act, but play upon several cognitive biases to make the user act as the designer intends. Nudges are not hard constrains, so the designer should be aware of the possibility of a player not doing the intended action, but if well design, they can create directed behaviors without the backlash of the more authoritarian design standpoint of the imposition and the dominant choice.

There are three main strategies for creating nudges:

1. Salience: a piece of information has enough contrast and is placed on a context in way that it directs behavior in a predictable way. The classic example is the spider sticker in the urinals on a bathroom. Something as simple as that creates the urge of aiming towards the sticker in a playful manner, which reduces sanitation problems. Games use this kind of nudges all the time. This is the reason why in-game notifications use movement and light to focus your attention in a particular piece of information on the interface. Once your focus is on the progress bar, your decisions will be nudged by that progress bar.
2. Social proof heuristics: if you see someone else doing a behavior, it will give you license or nudge you towards repeating that same action. This effect can be accomplished in game design in various ways, as for example by creating a character that queues the player to do an action. Let me be more specific: imagine you are playing with an NPC (non player character) when you reach a point where he draws a bow: this will signal your brain that it is time for you to do the same. In Gamification this has several applications, as it can help direct behaviors using social feedback or character design.
3. Defaults: if there is an overwhelming amount of data, you can create defaults so the player can make no choice at all. This has been studied in opt-in or opt-out systems, where people tend to leave the default option to avoid activating their working memory (the slow system of choice making that makes calculations). In a game where you can use multiple weapons but one in particular makes it more fun, you can make that weapon the default choice, so, unless you change your weapon, you will end up using the default one most of the time.

There are actually more systemic nudges as well, like creating goals or arbitrary endpoints, but if you manage to learn to use these three categories in your designs, you can help your players take behaviors that can optimize their play experience. And beware, if the system is just about nudges and taking the decision away from the player, your are working on a Behavior Economics experiment, not in a gamification system, which is okay if that is what you want. But if you want to apply game design principles, you need to understand how to design meaningful choices.

Calculations and Meaningful Choices

Games have the unique ability of activating the working memory more often than not. This doesn't mean that gamers are rational beings that can calculate the outcomes of all their choices. Playing games is, in a big measure, learning how to create skill by simple repetition, which means passing your skill from the working memory to the long term memory (the system that makes quick decisions). But each time a player is presented with a new challenge, she will try to calculate or power through those choices.

Most games present only simple calculations, like three spears allow you to throw 6 dice and you need to get a result of at least 18. Behind this kind of calculations, a particular type of choice emerges, for example, if I have 8 spears, how many should I choose to reach the target roll? This is not a binary "go to the right or left" choice, it is about optimizing a calculation in terms of probability. And probability plays a big role in game design, as fixed calculations might lead to dominant strategies (3 spears do 8 damage while one sword does 5: should I use spears or swords?).

This kind of choices are great to activate the working memory, but be wary that most players will make only part of the calculations and will complete the decision by gut feelings. If in the previous example you didn't took your time to make the math, you are playing from gut feelings, and will try to learn by pattern recognition (if I lose several times because I only used 2 spears, I should correct in the future by adding more spears). If well designed, this kind of choices can lead to learning, but be aware that they can also lead to poor conclusions (I tried 6 spears thrice and lost every time, maybe I should use 7?).

The second way of activating the working memory and create decision making is by creating meaningful choices. This kind of choices cannot be optimized because there is a black box that makes it impossible to calculate the best choice. There are various ways to create this kind of choices, I leave a couple:

1. Partial or incomplete information: why do so many card games have rules around not letting other players see your hand of cards? Because this keeps other players from being able to make optimal choices. Do you want the open item in the market for 3 gold pieces, or would you rather get a blind item from the deck for just 1? There is just not enough information to make an optimal decision, which increases the stakes, the stress, and, thus, the engagement. Players love to find themselves playing the odds or making assumptions, just to find out that they botched the choice (or not!).
2. Information in conflict: Maybe you have all the data, but the choices offer conflicting outcomes, which are both needed or wanted, or awful in their own way. This can often lead to analysis paralysis, so it must be designed carefully. The idea is to create choices that involve outcomes in different categories that both affect the system, but is impossible to calculate how. Think on "The Walking Dead" scenario where you find a strange man in a road taking care of a family. The man was bitten by a zombie and asks you to take the family with you and shoot him. The family begs you not to shoot him and tells you that if you spare his life they can give you their rations of food: the man is their father. If you kill the man, you gain morale for your people, but now have to feed the family members, who will not help you in your missions (you killed their dad!); if you spare his life, you gain the food rations, but you lose morale and other characters will look at you with an untrusting eye. It is impossible to make a perfect choice because there is a trade-off, not only in terms of the calculation (do I need the food or the morale more?), but also in terms of your self-image (can I actually kill the man? could I leave the family with someone that will surely kill them? can I live with the resentfulness of the family or the distrust of my people?).

This kind of choices are part of the principles of game design, and should be part of the principle of game thinking, so a Gamification Designer should incorporate them into their systems. But you need to be careful, as this kind of choices are not directed towards a particular behaviors, as was the case of the nudges and impositions. If you really need your players to do a particular behavior, you can not design a meaningful choice or you might end up losing control over your players. But if you find the choices that can build strategy, that are hard and engaging and that can actually even give you profiling data, you should aim for that.

Design tips

Hopefully you noticed how you should approach choice architecture in your design process, but I will still leave some, maybe obvious, tips.

1. When you need to direct behavior, you can either use linear designs or nudges, depending on how much doing the undesired behavior can damage the experience. If you really don't want people to be rude to others, then don't allow an open chat, but maybe a communication system based on predefined icons (like Clash Royale). The absence of choice is also part of linear design: if you don't have a dislike button, you can't show your dislike, which takes that behavior out of the equation.
2. If you can design effective nudges to direct behavior, this is always better than creating imposition in terms of the experience. Think of playful nudges, as playfulness is part of the toolbox of the game designer. But remember that nudges are not about coercion, so be prepared for players not behaving as you wanted and have rules for those possible behaviors. If there is a 1% chance of a player doing a behaviors, there is almost a 100% probability that someone will do that behavior if you have enough players.
3. Find or design strategic choices around the complexity of your data. Make probabilistic systems where players can actually calculate risk vs reward, but don't punish them so badly that they hate the system. Give them a sense of control by allowing a number of rerolls for example (play King of Tokyo or Yahtzee to understand the concept).
4. Use the unpredictability of complex human or data systems to create unpredictable outcomes. If you have 100 players making a bet and the outcome of your bet depends on the bets of everyone else, you can't optimize by calculation. This is the basis behind investment games, or fantasy football games for example.
5. Create meaningful trade-offs that allows you to create your own gaming profile and gaming style around choices where multiple behaviors are expected from players. Find where multiple playing styles can actually enrich your system, by creating roles or strategies that modify your experience in a custom way. If you choose the warrior you can't do magic but can carry heavy weapons, if you choose a mage you can use magic but only wield daggers. That is a trade-off in game design and this architecture of choice can be applicable to many systems in the real world.
6. And always, always remember to design around choice, not around behavior. The behavior is the desired outcome, but the game is played in the brain and the type of choices you give the players can make them feel like they are working, or like they are playing. Using game interfaces to hide imposition architectures will not yield great results. Using game rules to just create reward cycles is not game design.

If after reading this you feel your designs might be lacking meaningful choices or feel like you could be designing better choice architectures, feel free to contact me or leave a comment and let's build a discussion. I would love to know what you are working on and would be glad to give you some advice if I can. If you feel this could help the design process of other people, feel free to share or tag someone. Giving advice to real projects and discussing designs has been the main way I've used to polish my framework, so hopefully you will be giving me back as much as I might be able you give you in return. I'm really prone to find flaws in designs to try and improve them, so if you are seeking to indulge your confirmation bias, be aware that I might be harsh, but be sure that I will try to give alternatives and useful feedback.