Good Games Allow Learning

Some time ago, I read a throwaway line in someone’s review of a board game that gave me a big realization. The line is nothing special, and I guess I subconsciously knew this lesson already, but now it was put into simple words. Words that are easy to remember, and to use to improve all your next game designs.

What was the lesson?

Games are (addictively) fun when they show clear and tangible opportunities for growth/learning after each attempt.

I wish I could find the original context again. Stupid me always takes some time to get around to writing an article, and now I don’t know he source anymore. But I do remember the general lesson and how I’d explain it.

When we talk about what makes a game “fun”, we often talk about “rewards” and “flow” and “problem solving”. And, sure, those things matter. But they are worth nothing if the lesson above is not applied.

It doesn’t matter if a game has a great puzzle to solve, if you’re not sure if you’ve solved it. It doesn’t matter if a game provides many strategies to achieve the goal, if you have no clue how you can improve your strategy after playing. It doesn’t matter if a game gives you all sorts of tokens or money or points for actions, if you have no clue how to earn MORE money and points with a better strategy next time.

It’s about problem solving (learning/growth) and communication. It’s about having agency to try strategies and reflecting/feedback on what did (not) work.

A Simple Example

Imagine this. You’re playing a card game about fighting monsters. You play some cards, another player secretly checks if you beat the monster or not, and then merely tells you the result (“You did not beat it!”). Sometimes you beat the monster. Sometimes you don’t. But … you have no clue why. You have no idea what cards made you win and which did not. You have no clear feedback that allows you to improve.

This is a silly example, of course. Who would design such a game? Why so secretive?

But this is, in essence, the mistake that many (board) games make. They have really interesting rules, lots of strategy, lots of fun problems to solve. Yet they are not fun to play (more than once) because you have no way to learn. Because the game feels to clearly communicate, after each round or match, what you did wrong and ways in which you might improve your strategy.

It’s the same as real life, really. “Practice makes perfect” is a common saying, and also wrong. Learning happens in two stages: practice and self-reflection on said practice. That’s why many adapt the saying to “Perfect practice makes perfect”. Without that self-reflection step, you have no clue what you’re doing right or wrong. You’ll just be practicing the same mistakes over and over, never improving, never trying anything new.

Games need this too. It’s not enough to just provide an interesting puzzle to solve or varied strategies to try. It also needs to clearly show you which strategy you might want to try next to have a better chance at winning. It also needs to give you a chance to learn from your mistakes.

In a general sense, sure, most games will do this by default. If you did X in this game, and you lost, then you’ll know “don’t do X next time”. That is, however, a very long-term feedback cycle ( = you only get feedback at the end of the entire game) and also very vague. Most board games have thousands of cominbations of actions and strategies, so eliminating one at a time is not great.

How To Do This

Instead, I now feel a good game—that people want to play over and over—should,

Build in reflection moments regularly. Don’t just communicate or give feedback at the end of the game (with who won/how many points were scored), but give plenty of that along the way.
Specifically design the rules such that the game hands out exactly the information you need to learn and grow. You don’t solve this problem by telling the players everything. But you also can’t grow if you have no information about why you scored poorly or why your cards did badly. You need a middle ground in which the game highlights exactly the thing a player should focus on to grow and get better.

Because this cycle is what’s fun and addictive. The cycle of finding a problem, trying something, then growing a bit from your mistakes is perhaps the definition of “fun” for humans. The game itself can be extremely barebones or bad, as long as it has this cycle and executes it right, then it will be adored by many and played over and over.

I’ll try to give a very streamlined example of this.

A Simple Example: Part II

Let’s go back to my earlier example of beating monsters by playing cards, and let’s make it look a little more like your usual game.

You can see the monster you’re fighting and its stats.
You check if you beat the monster (no more secret checking by someone else) by following some simple rules.
Say, you calculate attack value, subtract defense value, and that’s how much damage you do.
Over the course of a few rounds, you wear the monster down until you eventually defeat it and get the reward.
(And the reward for beating that monster might be to draw X “loot cards” with some money or points on them.)

This is a more typical kind of game you could find anywhere, both as a board game or video game. I know this, because I have played a few games like this. Or, at least, tried them. But I always bounced off of them rather quickly, finding them boring and altogether “meh”, without being able to state why. They just weren’t that interesting. No matter how many unique cards, monsters, weapons, situations, etcetera could appear.

And I think the lesson in this article is the reason why. Because let’s think about it: does the game regularly give us a clear signal for how to grow? No, it doesn’t.

We only defeat a monster after many rounds or attempts, so the feedback regularity on that is low. Merely reducing a card’s health by 10 each round … for 5 rounds in a row … does not give any opportunity to grow along the way.
The reward we get is pretty random, so there’s no information to get from it. We might beat a monster with great tactics, and then draw terrible loot cards as a result, it doesn’t matter.
If we don’t defeat a monster, what does that tell us? Play cards with higher numbers next time? Randomly draw better cards?

Granted, the example is still simple. A real game like this will have some special actions on cards and special modifiers, but they don’t actually change anything. At its core, it’s still a game where every turn/event does not give clear signals for how to learn from it.

How could we change that?

Instead of health, we might give monsters a number that says: “you must beat it within X turns”. Now players know if they are doing well or not, depending on whether they’re on track to beat them in time. After every turn, they can reflect and say: “I didn’t do enough damage, I must try something better next time”
Instead of getting random monsters, place a map of monsters and their order at the start. Players now choose which ones to approach and in what order. If they don’t make it to the end, they can now reflect and say: “Going for monsters in that order was a bad idea, let’s try a different path through the map next time”
Similarly, instead of drawing some random cards as your reward, clearly print a specific reward on the monster. Now players have another bit of information they can use to decide which monsters they want to fight and when. If they lose the game because they’re out of shields halfway, they can now reflect and say: “I should have fought more monsters that reward me with a shield at the start. Let’s try that next time.”

Hopefully you see the pattern here. We don’t actually change much about the overall game flow. We merely present more information and let players act on it, which plants the seed towards being able to learn from the information and change your actions next time. The seed for learning and growth.

It’s still not a game with perfect information. You might still be drawing random cards in your hand. The later monsters (further down the map) might still be hidden at the start and only revealed over time. Maybe there are even dice involved when fighting for extra randomness.

But it’s a game with enough transparency and clear information to learn quickly from every attempt. And that’s the fun and addictive part of games.

I think lack of this aspect is very common, even in published games, because it’s such a hard thing to grasp and solve. My earliest games are surely missing this, and it’s probably the reason why they sucked even though I thought the rules were very clever and interesting. It’s hard enough as it is to design a functioning game with meaningful actions/puzzles each turn. Now you also have to ask yourself if players can actually LEARN from failing those puzzles.

It’s a lot to consider, yes, but I now believe this second part is just as crucial.

I guess you can also rephrase this as “make failing fun”. Make getting second place fun, because the game clearly shows you “if you try THIS OTHER THING next time, you might win!” Make going bankrupt fun, because the game gives you enough information to think “if I do THAT THING next time, I will be swimming in money instead!”

It’s what makes people glad to have played the game, no matter the end result. It’s what makes them want to play again immediately. That little spark in their head that just had a new realization and found a way to learn and grow.

Real-Life Examples

I would love to give a specific example of a real game, but for the life of me I can’t come up with a good one.

There have certainly been (board) games in the past that looked like they had everything going for them, but felt extremely “meh” when I played them. I only played those once, never again, and I think it’s for this reason. The game gives you an interesting puzzle, but by the end you’ve no clue why you won/lost, and the puzzle will be exactly as interesting next time because you can’t really learn from it. But I can’t remember those game titles right now :/

I guess a video game example is For The King. It’s the one I had in mind when giving my monster example above.

It’s a game that seems to do everything right: dice are rolled, you walk that many spaces around the map, meeting monsters or placing camps (to refill health/repair stuff/etc), and you can fight monsters on the same hexagonal tile for rewards you desperately need later. I’ve played several times with my little brother (it’s a co-op experience). The game had no bugs or lack of clarity, it gave more than enough agency and choice, yet every time we died at pretty much the same point and we were like “eeeeeh so what now?”.

The game gave us no real, clear, simple pointers for how to improve. Get better weapons? Then we need more luck with the dice and more money. Avoid monster encounters to save strength? Then we have no money, which means we also can’t solve bad luck with the dice, and we’re back to square 1. Approach fights differently? We have no clue how, because especially at the start it’s just stand in the same spot and hope you don’t miss. Sure, we get feedback that the attack missed, but that’s not the same as giving feedback that helps us understand how to do better next time.

The game has too much randomness and too many variables, and after many attempts we saw no way to actually grow and gave up. It would have been better if the options were far more constricted. If the random generation was slightly … less random. If it didn’t only give feedback in the vein of “this thing JUST HAPPENED”, but also “this thing WILL HAPPEN” (a bit of why and how behind the information). Then we would have known exactly which actions we took and exactly how that was good or bad, which allows learning from it. We would be encountering more similar situations each time, so we can actually apply lessons from previous attempts.

As for board game examples, I can only come up with Paleo right now. Gorgeous game, good reviews, unique idea with the double-sided cards. Essentially, you are prehistoric humans exploring the wild to find enough food and tools to survive every day/round. Each card has a backside that shows the general kind of card coming up, but you only know the specifics (just how good/bad it is) once you explore it and turn it around.

When I played this game, it was “kind of fun” and “kind of engaging”. Players understood the idea, there were some tense moments and fun reveals, and we inevitably starved and lost on our first attempt. Then, the same thing happened on the second attempt, without much improvement. I remember racking my brain for a way to play better, for some strategy, for some lesson to learn. I asked the others; they had no clue either.

There was just no clear information about how to improve! Sometimes dangers are fine, sometimes they’re not, but that depends on the specifics. Only seeing the BACK of the card—hey, it’s a danger coming up—means nothing! You don’t learn anything from doing the wrong thing. Because next time, that card looks the same from the back as the last time, and you still don’t know what to do.

This problem is compounded by the fact that everyone has such a deck, and tight cooperation is really important in this game. It’s even harder to learn from someone else’s mistakes as you’re so focused on your own deck and decisions. I guess that’s a common issue in cooperative games. They always limit communication and information—otherwise it’s too easy!—but that limited info also destroys your opportunity to see patterns and learn from mistakes.

I am sad to say that, after those two first games, we never played the game again. Because it felt like taking random steps, and then, once you died, having no clue how to take better steps on a next attempt. So why make a next attempt at all? The game did everything right—except giving clear feedback on how to improve, and that kills a game faster than anything else.

Hopefully the article was clear and informative.

Keep playing, keep developing,

Pandaqi