Implicit vs. Explicit Failure for Educators and Game Designers

Let's talk about failure. This is an important topic for educators and game designers alike. I like to think in terms of two broad categories: Explicit and Implicit failure.

Explicit is 'game over' failure. Running out of lives in a platformer is an explicit way to fail.

Implicit is 'you have not succeeded' failure. Giving up on a crossword is an implicit way to fail.

This distinction matters because failure is a psychologically impactful experience. We should be mindful of its place in our games and classrooms. We should be particularly attentive to the stress induced by explicit failure. Explicit failure is overused, often because people see it as a required ingredient for games and assessments. But it's not!

Think about the Rubik's Cube. There is no explicit failure condition for that wonderful puzzle. You never have to worry that you'll run out of moves, or that the timer will go off, or that you'll click on a mine and the game will be over. You can impose failure conditions like timers and move counts, but they're not intrinsic to the puzzle. You can always keep turning. The only way to 'lose' is to put it down without achieving success. This is a big part of why Rubik's Cubes are so pleasant.

If there was some condition that made the Rubik's Cube flash a 'Game Over' screen at you, I guarantee it would be a lot less popular. Why? Because explicit failure is psychologically distressing.

Maybe the failure condition of the Rubik's Cube is when it explodes under mechanical stress?

The Myth of the Swiss Game Designer

At GDC this year I spoke with a wonderful team of Swiss developers showing off a game about Greek mythology. The game was mostly a visual storybook, wrapped in a thin interactive layer that complemented each myth. I loved it. However, the chapter I played had one real rock-in-my-shoe mechanic.

The player takes on the role of Icarus, flying ever higher in the sky through a field of obstacles. When the player hits an obstacle, their wings break and they fall into the sea. Then the chapter restarts. It took me about five attempts to make it all the way up. "Wow," the developer said with a grin, "usually it takes people more tries than that!"

Forget the fact that I heard the same 30 seconds of narration five times in a row. As soon as I realized I could fail, avoiding failure was in my psychological foreground. I stopped being mesmerized by the gorgeous artwork of my obstacles and instead changed directions whenever they came into view.

Clearly this guy is talented. I will buy that game when it comes out. But this kind of design is a problem, and I told him as much: "I don't think you need a failure condition for this to work."

"Well, we need to have some kind of challenge. Otherwise it's not really a game, is it?"

In his head, game—>challenge—>explicit failure. This guy was a lot older than me. I'm guessing that his formative gaming experiences were in coin-op arcades where games had to rely on explicit failure to make money. I understand his instincts, and explicit failure is sometimes the right design choice... but not for the game he was making!

I suggested that instead of falling all the way to the ocean; the player should falter in the air, fall a few meters, then regain composure and continue flying. That way the player could explore the sky, admire the artwork, and reach the end without ever hearing the same narration twice.

To my surprise, he instantly reacted with grateful enthusiasm to this small suggestion. He told me that testers often complained about this sequence, but he didn't know how to fix it. He thanked me for the advice and asked if I had any other critiques. I did not; the rest of his game was pretty much perfect!

This story shows how easy it is to assume that a way to 'win' must come with a way to 'lose'. It also shows how simple changes can nullify the problem. Game designers should always ask themselves if their failure conditions are appropriate, especially when testers get visibly frustrated.

Why This Matters for Teachers

Educators should take care to challenge students without creating a fear of failure. For English and History classes this is usually only a problem for grammar quizzes and on-demand recall of names or dates. The free-response format of most assessments for these subjects doesn't employ much explicit failure. But math classes... boy do math classes have trouble here.

Most people think about math like an arcade game. Each problem is a level where you make moves until you make a mistake. After that, it doesn't matter what you do because your answer is already wrong. Game over!

Multiple choice is a blunt manifestation of this approach. Every time I see a multiple choice math question I feel a twinge of dread... and I love math! That twinge is the fear of explicit failure that is intrinsic to this format. I know that the moment I click an answer on Khan Academy or Brilliant.org there is no going back if I'm wrong.

If you think you're bad at math, those moments of failure stick with you because they reinforce a negative perception. If you think you're good at math, those moments stick with you because they chip away at your sense of self-worth. This is where math anxiety comes from, even for "math people" like me.

So how can we assess math skills with a less explicit failure condition? One option is the free-response format. An example is "show your work" questions where grades are based on process instead of final results. A related example is "explain your thinking" questions where students describe how premises can be chained together to reach a conclusion. Another format is the "draw a diagram" family of questions. These can be good for topics like geometric reasoning or kinemtatic physics.

These formats are a big improvement over multiple choice or "write the correct number" questions. There is still some notion of 'failure', but it's much less explicit. Think of it like the Greek mythology game I discussed earlier: hitting an obstacle does not mean 'game over'; the player can always keep flying. I distinctly remember these questions as a much less stressful experience in school.

Unfortunately, these questions are difficult or impossible to grade with computers. The AP subject tests do it by shutting hundreds of teachers in massive rooms for hours at a time and paying them by how many tests they process. This sounds pretty hellish to me. Teachers I know who participated in these operations agree.

So are there clever ways to design math problems that are open-ended, encourage exploration, and can be assessed automatically? Honestly it's pretty challenging... but the answer is definitely yes! However, that's a topic for another post. For now, I hope you can see why analyzing failure conditions is such a critical skill for anyone building educational or gaming experiences.