BEM Gamification: The uncertainty paradox

A big thing in BEM gamification is working around the concept of uncertainty. This began the moment I read the "uncertainty system" definition of games put forward by Zimmerman and Salen, and contrasting it with Cowley's paper on video games and flow. From there, I am always working on creating by design a proper level of uncertainty in my systems to keep the game interesting, but, when you work with positive feedback loops and progression systems you find that uncertainty, the important kind, has an inner paradox that might be what's killing motivation in the workplace, what makes people stop playing games after a while, what creates a stopping point in TV shows, and is probably something that should not be "solved" but accounted for.

Two types of uncertainty

In the BEM model I work with to types of uncertainty: random-based and knowledge-based. Random-based uncertainty has its own problems, specially when contemplating the different motivation approaches towards Discovery and Empowerment, because too many random elements takes control out from the player, but too little randomness and the game can become a puzzle with no replayability value. Understanding the differences between input and output randomness, how the player's mind calculates probability and how "beating the odds" creates thrill is really important to design random elements that won't feel unfair.

But that kind of randomness is an artificial artifact that is easier to work with, because you, as a designer, can control it and maintain it. But knowledge-based uncertainty is harder to work with, because it stems from the player's previous knowledge of a strategy and skill sets towards completing a goal, as well as from the narrative values and game lore. If you think about it, in education this is the kind of uncertainty you want to diminish overtime, as you want your players to reduce their gaps in knowledge when achieving learning goals, but this can have an unexpected effect in the long run.

Speculation vs prediction

The prediction game can be really motivating because there is a natural veil posed by future events. In games, beating the odds is part of what makes a game fun and predicting the future correctly is usually a rare event. But our brain deals with this kind of uncertainty by creating models and testing them, by speculating about the future, which builds strategic value to the world. It is impossible to watch a TV show and not be making predictions on what will happen further along in the plot. Most of the time you get it wrong, but when you manage to guess, that is exciting!

Speculation is a big part of the learning cycle in games, as every new pattern in the level design begins with the player making hypothesis and reacting on partial and new information. You know how everyone says that losing on a safe environment is a big deal in games, but it is not about creating losing conditions and say "game over". Losing events must be accompanied by information elements that help the player understand "why" they lose. And here we start finding that the motivational power of speculating is not flat: if there are no clear feedback elements that help me understand why my model failed, it can become frustrating. Games are uncertainty machines, but maintaining the uncertainty at the same level over time can kill your game.

Fluctuation of uncertainty

Imagine a platforming game where there is a row of enemies you need to cross by, but everything you try kills you. There are no clear indicators on how to solve the problem, you only know what kills you, but you can't find new information on how to clear the obstacle. After a few repeats the game over screen is no longer motivating you, on contraire!, it is killing your motivation. This is because the uncertainty level is not fluctuating, so you feel you are not learning. Imagine an even worst scenario: you are being killed by some invisible force but you don't see it or hear it, and you can't predict when it's going to kill you. You start thinking it might be a bug, because a game would never kill you without giving you enough information on "what is" killing you. You are not only not learning (sorry for the double negative), you are actually speculating on a direction that places the blame on the unfairness of the game. You are relieving your uncertainty by making conspiracy theories that helps you feel you are learning something.

This is dangerous, of course, because the built-in positive feedback loop of learning means that uncertainty should always be in fluctuation towards certainty, but controlling the "certainties" your player is incorporating into their models requires a clear feedback and information design, or they might learn the wrong thing. And this happens a lot when we learn stuff empirically, because most of the times the real world doesn't care if you get the correct feedback and many systemic relationships are subtle and not immediate, which hinders learning the right stuff. The game designer can control the information space, and should never cut corners around this issue: the game most give me enough information to be able to solve the problem in a few tries.

The outcome

A game is a system that has ruled outcomes. It can be a reward, but it also can be the certainty that a certain action will allow you to change 4 sheeps for 1 wood. The rules of the game, for the most part, have a certainty that is required for the player to feel justice, but some black box elements can bring some randomness to the table. But again, there are layers here. You might know what you will get but might not know how to use it. In games like the Witcher 3 or Zelda: Breath of the wild, collecting items for crafting is a big part of the mechanic, but most of the time you never get to know how to use those elements. In this case, you might know how to get a herb or ingredient, but you might never know how to use it (unless you search on the Internet). Useful elements like tools and functional rewards might be interesting if you don't know at first how to use them, but you need to build that certainty over time, or it will lose its motivational power.

Functional rewards are one of the few things where more certainty creates more enjoyment, and uncertainty can bring overwhelming frustration. Winning extrinsic rewards like monetary prizes can have another effect, because the knowledge of the possibility of losing something you care about might make you anxious. When the element of uncertainty can impact your life in a serious way, like the prospect of not winning money or losing your job or failing a grade, or a pandemic, you cannot enter flow states so easily (this is called the perceived importance).

Muscle memory and learning patterns by repetition

Learning is slow, because our learning mechanisms require repetition and the use of a skill to improve it. Muscle memory is a great metaphor for this, because you can clearly see that we admire the people that can move their body with above the average precision. Being fan of a sports player or going to Le Cirque du Soleil captures this idea perfectly: our muscles are erratic which creates inner uncertainty in game play. If a soccer player could always kick the ball perfectly it would become less interesting. In Formula 1, a constant criticism is that the vehicle technology overcomes the drivers skill uncertainty to a point that it feel like is the car winning the competition, not the human. Detroit, Become Human, has a whole controversy around android entering sporting events, because they have an unfair advantage over humans. In muscle memory, years of training may not lead to perfection, and the game must remain imperfect to be interesting overtime.

You can extrapolate this problem to chess, where the skills are not about muscle precision, but about pattern recognition. To learn chess means to be exposed to many, many patterns and be able, by repetition, to create predictions on best moves by combining all those patterns to find a novel solution. But you learn slow, because, again, only repetition can give you the required skill. So while video games train muscle memory a lot, board games train pattern recognition the same way, by repetition.

Both of these skill are key for motivation, because the faith of growth built into a kind of "Growth Mindset" is what keeps a player playing. In our platformer example, let's say we corrected the game and showed the player how to solve the obstacle. Now, knowledge and execution are not the same thing, and the player still needs some precise button taping to pass the obstacle: so the player "knows" how to pass the challenge, but can't do it still. Skill building usually lags over knowledge building, so, while you reduced the uncertainty on what you need to do, you now need to work on the how you need to do it. These actually keeps the player engaged, because her expectation is that she can now solve the problem if she tries enough times to get it right (even by chance). This keeps the uncertainty on "when will I make it", but reduces the uncertainty of "will I make it", so again, there is a fluctuation on several levels that works towards keeping you engaged in the task at hand.

Is repetition fun?

Repetition is a kind of certainty and can become boring really quick. Our brains don't get excited on repetition, but only if certain conditions are met:

1. You can predict that you will achieve the task.
2. You can predict the outcome of the task (reward or not).
3. You know there is no or little probability for error.
4. There is no problem solving involved anymore.
5. You are not invested on the skill you might learn from repeating the activity.

If I ask you to repeat a word 100 times so you can learn its correct orthography, you might only do it if you are invested in learning the orthography. Otherwise, the task is repetitive and boring, because there is no problem solving and little chance of error while doing the exercise. Repetition can kill fun when certainty becomes the main rule!

Now, for a sports player, their life can become the repetition of one game into another, but it might not get boring quickly because each match is filled with uncertainty: players can't predict they will win the match, there is a high probability for mistakes, it requires problem solving and they are invested in the task. This creates a recipe for long-run engagement, but this is hard to recreate in the real world. Most games become dull by repetition, like playing the whole Assassin's Creed series, which has been criticized because the sequels tend to maintain the same formula. Games try to keep it novel by several techniques, like changing the plot, changing the conditions for a mission, limiting some game moves or using novel mechanics in certain parts of the game, but exhaustion can kick in and that is why sequels are hard to design and are released some years apart.

Positive Feedback Loops

Positive feedback loops are uncertainty reductors by nature. If you get knowledge, that will help you solve future problems with less risk. If you get a power up, you will have lower chances of failing an activity. If you get more points that everyone else it will become harder for the other players to catch up, so the result of the game becomes predictable. Positive feedback loops increases predictability, while negative feedback loops can accomplish the opposite. If the front runner of a racing competition consumes more gas, it becomes less apparent that she will be declared the victor at the end.

In any storyline, the nearer you are to the end, the more information you as a viewer have, which means you already have most of the knowledge about the mysteries you were trying to conceal. That's why TV series and narrative games always have a big reveal under their sleeve for their last chapter, which can also serve as a cliffhanger for future sequels or seasons. The further you are along in the story, the more satisfying the reveals must be, and character growth and having parts of its past concealed is a way to manage this. This is so important many series actually fail to do this well and start plot lines that break the suspension of belief of the viewers, with many characters that were all good starting behaving erratically or evil, or a friendship you care about breaking down for the sake of moving the plot. And this is hard to keep up because your viewers have seen those techniques in many other shows.

And so appears the paradox

The problem seems to be that learning is engaging in itself, but is a road for boredom! Look at the mastery paradox: mastery should feel like an asymptote, you should never reach it, or you will lose interest; but also, mastery can't feel stagnant, if you feel blocked you will become frustrated. How to deal with both of those scenarios! They seem to be the only possible outcomes for game progression: or you master the game or hit a roadblock! Either way, no game can truly be eternal for a player and, if you play enough games, gaming in itself can become dull with time. Yes, this happens more than you think. For me, for example, finding an engaging game now is harder than ever, because most of them feel like repetitions of games I already played. This can happen even with some masterpieces, like Sky, from Thatgamecompany, which I feel is too similar to Journey.

The fluctuation of uncertainty creates a kind of design certainty: if your game is well designed, your player will learn, which reduces uncertainty, which kills part of what makes a game enjoyable. And if every game in learning becomes a trivia with points, badges and leaderboards, that would be the quickest path into losing the interest of our learners! Again, artificial uncertainty by randomness can help, but that requires more thought on its own problems.

Two years ago I spent some time talking with parents about Fortnite, because they thought their kids were addicted to that game. I remember I told them that a couple of years later whey would stop playing that game that much, and now Fall Guys and others have been taking the scene. Unless there is a real addiction process happening, the brain can't stand the repetition of the same activity for too long, and will start asking for novelty again.

You can't even keep the motivation by storytelling, which is one of the most powerful tools at our disposal! It is rare for a TV series to survive their sixth season. A series like Dr. House manage to maintain the same formula every episode for 8 seasons, which is remarkable from my POV. The detective genre tries to keep the mystery through creating weird cases, but you still get used to the characters, so you start feeling like you need something new. And even if the viewers could manage an 11th season of Friends, the actors and staff will not! If we can't keep viewers engage in one of the most engaging media there are, how can we expect to keep people engaged in the workforce? Gamification might introduce novelty for a while, but will not work forever, trust me, I know.

Designing around the paradox

I believe the first thing we need in our field is some humility. Gamification expert might play a weird and marvelous game, and we have a lot of knowledge about engagement and experience design, but we must know our limits. We never should stop struggling to push those limits back by research and experimentation (or we might kill the uncertainty and get bored!), but we must understand that humans are hardwired for change and novelty. A challenge that keeps repeating itself will stop working and might have the effects of extrinsic motivation. We must be aware of this effects so we can design around them, and so, I believe we can try to figure some stuff out again from the gaming industry. I leave you some tips.

1. Design with an expiration date in mind. Loyalty program sellers like to tie their clients in long-term relationships, which requires selling things that you can't stop using afterwards (that's the basis of a communications company). Gamification should not try to replicate this model, unless a real game-as-a-service scheme can actually bring long-term benefits to the company and their users. Most games should understand the lifespan of their players and be aware of fatigue.
2. If you need to teach a particular skill, you don't need a game that last years. Model the game around the learning curves of your players. If the gamification project relies on learning cycles, and not in engagement cycles (like habit building), you should be able to remove the game from the player's life and its effect should be long lasting. If you feel that removing your game will make your players stop behaving as you intend, you will have more problems in the long run.
3. If you need your players to practice a skill for a long time, you need to make sure to create a game based on dosification. Mobile games tend to be long lived because they give the game to the player in small doses. This will reduce the perception of repetition and will reduce behavior extinction.
4. Avoid relying on dark patterns. Again, creating meaningful learning for your players should be your goal. Dark patterns might work on the short-run, but if your players start feeling manipulated, you will lose them hard. Again, your player might learn things you don't want if you are not transparent with your feedback systems, which kills uncertainty but in a bad way.
5. Focus on "how" things will happen and not in "what" will happen. If you watch the Hunger names you know that Katniss will be a Victor in the first movie from the very beginning. There is little uncertainty there, but getting to follow her path is exciting. Our brain will even forget that certainty in moments of grave danger which creates temporal excitement. If the show is interesting, getting the reward will not be the brain's main focus and it will become less extrinsic by nature.
6. Avoid never-ending progression systems if you can. They are fake and become grinding machines for task repetition. If your progress mechanisms are well designed, it doesn't matter you can only get to level 10. Intrinsic progression systems are more important and this is accomplished by feedback and challenge design.
7. Learning and experience goals should prime over engagement goals. Engagement is a tool and should be aligned with the purpose of the activity. People are suffering from all the engineering around growth hacking and as gamification designer we could be a cog in the problem if we start feeling that attention is our currency. This might be hard to read, but if your goals are centered on experience and not in attention, usage, frequency and so on, you might have a more positive effect in humanity and our field as a whole. That means, if learning reduces uncertainty and uncertainty ends the game, you should let it end.
8. If your building a game that will be integrated into a core necessity service, which means that it might not end, make it systemically meaningful for the user. If you are building a loyalty program, be sure your player's will get out something out of it. A loyalty program that mines your data but never gives you enough in return is, again, a bad system. You don't have to return only money, you can give knowledge, experience, purpose, challenge and much more, but make sure you are giving something back.
9. Potentiate creativity and problem solving. Every project I work on requires my full knowledge on gamification to solve novel problems, which means that my knowledge is the base to solve uncertainties, and is not the goal itself. I suppose that is why gamification designers are motivated on the long-run. Trivia-like games will never achieve this, because they are only knowledge-based, and are not skill or creativity based. If you only need to give the knowledge, that's fine, but plan for a short uncertainty curve or a high frustration curve and minimize those effects in your design.

Now, this paradigms are part of the BEM model and you might disagree with some of this points, or might be incentivized to ignore them, but try to make sure you are not increasing short-term gain in spite of increasing the odds for long term failure. Your players will learn from your design, and the moment they stop learning or getting meaningful things from it, they will stop using it. If you need to keep them engaged, think on how you could create a more powerful learning experience, and avoid resorting to behavioral tricks if possible. And again, be creative! This is a hard problem to tackle, but the challenge makes it worth your while!

Javier Velasquez

Founder of Free to Play and designer of the BEM Gamifcation framework.