Board game design

Though some great games can only be played a single time (e.g. T.I.M.E. Stories, though there technically it’s a scenario that can only be played once), replayability is generally seen as something to strive for when designing games.

So the rules are simple…
In a recent post I argued that why we like “deep” games is that they are replayable.

Then I recently read this article, which does an amazing job in tying depth, strategy, tactics and many more subjects together (warning, long read!). In it the author defines depth as:

“Depth is a function of the layering of heuristic understanding necessary for effective play”

This seems to coincide well with my own ideas. I never gave a formal definition, but informally it’s something like: ”A game is deep if it requires a lot of play-throughs to master”. I guess my loose definition is a bit more easy to understand, while the definition from the article is more rigorous. Take your pick…

Both ideas center around how much you need to invest to really learn a game. And thus learning is one of the central points here. Which will be the subject of this post.

How humans learn

Kids learn through rewards and punishments: “Hold your hand against a burning stove and it burns (don’t do that again!).” And: “Eat a piece of candy and you get a delicious taste in your mouth (more please!)”. There is a very quick feedback between the action (touching a stove) and the result (pain!). Because of this you really only need a single encounter with a hot stove to be very careful around one for the rest of your life. Likewise, if we get a candy and it doesn’t taste good, we’ll be slightly less inclined to have more candy in the future.

As we get older, we can learn when there is a bigger remove between our actions and the results: “Be nice now and get desert later.” Or: “Don’t do your homework today and get punished at school tomorrow.” Still, there is a clear connection between our action and what comes our way because of it; the teacher is very clear that you have to go see the principle because you didn’t do your homework.

When we are grown up we can handle even bigger temporal differences: “Study for a year and you’ll be in line for a promotion after that.” And as long as we do actually get the promotion, we’ll have learned that studying leads to advancement.

The examples above all have a clear link between our actions and what we get out of it. But those connections aren’t always so clear cut, especially when timeframes are longer: “Did I get my promotion because I worked hard, because I have skills that nobody else in the team has or because I sucked up to the boss?”

The joys of learning

Humans really like learning stuff – being able to form a connection between an action and its results. This makes sense, because else life would involve random acts and random results, meaning you’d never be able to predict what would happen next. People without the ability to predict don’t survive very long (“Let’s see what happens when I go pet that lion?!”)

You might disagree that people enjoy learning. That is because most people associate learning with school and school is anything but fun. This is because when studying the result is very indirect. What you learn in school doesn’t help you to predict the world better, it only helps you to do well on an exam.

Learning the correct conjugation of a French verb in school doesn’t do anything in real life. But compare this to someone who moved to France who has a bit of insight and through interacting with the locals is improving his language ability. This does immediately impact his life and as such is much more satisfying.

Learning in board games

So what does all of this have to do with board games (yes, the subject of this blog really still is board games!)?

The most important thing to learn is not to play this game…
When you first play a game you suck. And you’ll readily lose to someone with more experience. But while playing your first game you’re gaining insight in how the game is played. You should play this card and not that, going for points early makes you lose steam for the end game.

This is not some dry learning you’re doing in school, no, you’re learning something that is immediately applicable, in the next round or in the next game.

My belief is that this is one of the most important reasons people enjoy board games so much: They give the best kind of learning experience. The kind that can be used right away. You are now able to predict the future that much better, well done, have a shot of dopamine!

How to learn to play a board game

So you want people to be learning while they play your game, as that gives a pleasurable rush.

How do players learn a game? Through a bit of insight, but mostly through trial-and-error. When you play Catan for the first time (with others who have no experience) you haven’t a clue on what the best choices are, so you place your villages almost at random. Then a few rounds in someone remarks that rolling sixes and eights seems more likely than twos and twelves, so you’ve learned to focus on the big numbers (big as in that they have a larger font size on the tiles). Good, have your jolt of pleasure!

Then you learn that you need to spread what kind of resources you get (buzz!, but that a bunch of meadows combined with a sheep harbor is also a good idea (buzz!). Then your neighbor blocks one of your roads and takes a juicy spot, taking the game to another level (more buzz!).

When you’ve played a lot of different board games you start to recognize meta-patters: More resources is generally better. Getting more actions (e.g. more workers) is almost always is a good investment. Etc. But you still need to dig into the game to really learn its specific ins and outs. Which means playing, trying and failing. Until you stop failing.

The measure of success

What does it mean to stop failing? What does it mean to succeed?

When learning French your aim is to be able to have a conversation with that hot Parisian. When learning to play a game, your aim is to win.

The learning feedback loop

During a board game you do a lot of things. And at the same time your opponents are also doing lots of things. You’re playing cards, gathering resources, bluffing and moving tokens about. Depending on what game you’re playing you might take between a single and hundreds of distinct actions.

Sing it back, bring it back. Sing it back to me!
Only when all these actions are taken does the game reach its end. And only then can you determine whether you did well or not, because only then will the winner be known.

The result is that the feedback loop on whether any single action was “correct” is relatively long: Only after the game is over can you determine if that action was part of a winning way of playing.

More interestingly, giving feedback (win / loss) only at the end of the game obscures the information about any single action immensely. Because was it this action or that one that made the difference? Was it their combination? Or were both of those actually sub-optimal but did you win because you did a few other things right?

Worse, you can have won because of luck. Or because everybody else was playing like wet rags.

The result is that it can be very difficult to figure out what an optimal choice is at any given moment (which is of course exactly what we want; it’s well known that interesting (read: difficult) choices make for good games).

Too long a feedback loop

However…

If the feedback is too obscure, if you really can’t figure out how or why you won (or more likely, lost) then a game will lose its appeal. As written above, learning is fun if you can use what you’ve learned. And that means that something actionable has to come out of the learning experience: ”Next time I won’t place my first village between the dessert, the sea and a two…”

Luckily, games generally provide shorter feedback loops as well. In Catan you can see when someone else is getting more resources than you are. Being human we instinctively understand that more stuff is better! So it might only take few turns to regret placing our village at a two instead of at a six, meaning we will have learned something.

A layering of feedback loops

The ideal game then has feedback loops at many different “levels”; there should be extremely quick feedback (having more villages means I get more stuff!), intermediate feedback (placing a village at a six is better than at a two), long term feedback (taking a number of development cards is a good idea as that obscures how many points I have, meaning I won’t be the target of the robber that often) and every level in between.

“Deep” games have many layers of such feedback loops, resulting in interesting learning experience for absolute beginners, but also for people who have already played a game for a hundred times.

What this means for game design

So how does all of this help us design better board games?

I think it’s a light bulb…
Telling players how they are doing can help create short feedback loops. If you gain a few victory points with most actions then you can very quickly see your progress and measure it against your opponents’. This helps to quickly progress through understanding, which can be a good or bad thing, depending on your target audience. If you’re trying to create something quick and light then this is definitely the way to go. But if you’re catering to die-hard gamers then it makes more sense to obscure any form of progression, as these people can more likely glean the “basics” quite easily and would in fact be more enamored by having to learn through long-term feedback loops. This then means that it would be better to give as little information as possible about “who’s ahead”; no victory point tracks (or perhaps no victory points at all). Imagine for example a series of hidden objectives which stay hidden until a player has achieved all of theirs and declares herself the winner.

It also means that self-testing of your game-under-development is difficult if not impossible. It’s your game so you’ve probably played it many times and know the ins and outs, meaning you aren’t learning anymore. Or even if you are, it is most certainly not at the same level as a novice player. You might argue that different versions of your own game will require new learning and you’d be right about that, but that learning is helped immensely by all the learning you’ve done on previous versions (I’m going to assume here that versions are actually quite alike; if not you’re basically starting on a new game).

This is not to say that self-testing is completely useless; when balancing game elements you can probably get reasonably far just on your own. Just remember that you will be playing as an expert and thus that the “balance” you’d create would be for an expert. The result can be that the game would actually be quite unbalanced when playing for the first time if an opponent happens to stumble upon a strong combination that an expert would easily deal with but that will simply kill you when you’re new. As an example consider the Fool’s Mate in Chess which is not fun to get served up when you’re learning the game.

Closing thoughts

As mentioned above, I believe that “learning” is one of the main drivers of enjoyment for games. This is in general not something you need to think about actively when designing; it’ll happen automatically. But when you’re going a layer deeper, this might be exactly the thing to think about: What is the learning path, what would players pick up first, what later? Is there an entry level that’s interesting enough to get to the deeper stuff? Is there deep stuff that keeps the game interesting for a long period?

Indirectly this also answers why I love writing this blog so much: I’m learning – not about any given board game, but about board game design. I hope you’re enjoying your learning as well! 🙂

Further readin

A while back I wrote about the different ways in which learning a board game can be enjoyed: The joys of learning board games

And perhaps you’d be interested in creating something that requires different skills to be learned? 12 Skills you can design board games around

About the author

Bastiaan_smallHi, I’m Bastiaan. The goal of this blog is to learn about game design. That’s hopefully for you as the reader, but just as much for me as the writer.

Help me to learn (because hey, it’s fun!)? Leave a comment (below) or connect with me on Twitter? You can also subscribe to this blog (see the sidebar) or like it on Facebook, to get updates when I write them.

And perhaps you know of others interested in learning? Share this post using the buttons below:

Facebooktwitterreddit
Board game design

Some time back I wrote an article about scarcity in board games. The premise was that board games revolve around resources (in many different variants) and that scarcity of (some of) those resources creates interesting choices.

What is scarce becomes valuable
In this post I want to go into an example of this, based on Voluntarios (volunteers in Spanish), the game I’m working on. Specifically I want to go into a somewhat novel form of scarcity: The scarcity of space.

A first scarcity of space

In Voluntarios you’re at the head of a group of volunteers (workers) and you have those volunteers do work (in a slight twist on worker-placement, the details of which I won’t bother you with) on reconstructing a village after an earthquake.

There are two types of workers: Experts and Normal workers. Most of the work that needs to be done is “normal” work, but there is also some work that can only be done by Experts. Experts can do the work of Normal workers, but it costs the player a Karma (= victory) point when doing so (because when you’re an expert, you don’t want to do simple menial work, now do you?!). And of course Normal workers can’t do Expert work.

The number of spaces where Expert work needs to be done will ebb and flow through the game, but on average there are slightly fewer spaces than the number of Expert workers available. This means that there is an incentive to “get rid” of Expert workers as soon as an Expert space becomes available. This then creates an interesting decision, where a player needs to decide between placing an Expert or doing something else that might be more useful to do, but with the risk (or even certainty) of losing a point when the Expert is forced to do Normal work.

Expert spaces are created through buildings that can be worked on by all players, but where the player controlling the building will get the benefits when it’s finished. However, making such spaces available generally does not mean that a player can take advantage of them (by placing an Expert) immediately. Thus, there is a choice between making Expert spaces available (which will go toward finishing a building a player controls) and waiting for someone else to do it so that an Expert worker can be placed.

A second scarcity of space

A prototype construction project – with space for 1 Expert (green) and 2 Normal workers (blue)
One of the advantages that finished buildings can have is that they give the controlling player an additional worker. This means that the number of workers increases throughout the game. The total number of spaces for workers however goes down as much as it goes up, meaning that there can be moments where there are more workers than spaces to place them.

Of course, letting things go to waste is well, a waste. Therefore there is a mechanic that players lose points when they have unused workers left at the end of a round.

The result is that players need to think very carefully on whether they want to control buildings that give additional workers; they may be beneficial, but when space runs out they are very much a burden. This then creates interesting strategical choices on whether to invest in more workers or to go for other types of buildings.

And it’s not a stand-alone choice, it very much depends on what the other players are doing as well; if everybody else is investing in other buildings, then having a few more workers of your own means you’ll still be able to place them without too much trouble.

The good, the bad and the unexpected

One of the two types of space scarcity works like a charm. The other has some… Side effects.

Can you guess which one is which?

Well no worry, I’m going to tell you!

(No) space: The final frontier!

When there is not enough space for all workers the result is that all possible worker spaces get filled and thus that all associated actions are taken. A number of those actions benefit not just the player taking them, but other players as well (think of role selection, but through placing a worker instead of taking a card). For any given action there are only two options: I take the action or someone else takes the action.

This is actually quite a bit less interesting than what happens for most games: I take the action, someone else takes the action or the action doesn’t happen.

When all actions are taken there is still jostling for getting the actions that you really want, but there is no tension about which actions exactly will have happened when the round has ended.

Worse, when there are limited spaces at some point players are “forced” to place workers in spaces they aren’t really interested in or even worse, would actively prefer not to take. Technically they have the choice of not placing a worker, but if the downside is high (which it was in my game) then it’s not really a choice at all. And thus this mechanic took away player agency and resulted in a lot of frustration

Bad choice of space scarcity! And thus “too little space for all workers” will be taken care of in the next iteration.

Without space everybody can hear you scream!

What do you mean that space can be scarce?!
The other option for making space scarce, jostling for positions for Expert workers, however works well. It indeed creates a sense of urgency about “having to get rid of” your Expert workers.

So why does this work but scarcity of space for all workers doesn’t?

The fundamental difference I feel is that Expert-space-scarcity doesn’t take away options. An Expert worker can be used for any space, though at a penalty. You’d like to avoid that penalty, and so you’ll work towards that, but you don’t have to. This gives the player control.

For the full-space-scarcity on the other hand at some point it becomes clear how many spaces there and thus for how many workers you’re going to have to take a penalty. There are some actions you can take to increase spaces during a turn, but they help everybody (almost) equally and thus do not really give a sense of control.

Salvage space?

So full-space-scarcity is a bad idea.

But…

I like the idea, as it is novel and actually goes against the idea of so many games that “more workers is better”. So might there be a way to salvage the mechanic?

What if this scarcity generally doesn’t happen, but only shows up every once and awhile? Like in 1/5th of all the rounds?

Would that be enough to force players to take it into account when choosing how many workers to go for? Or will it happen “at random” and frustrate players just as much as when it happens regularly? Perhaps if players are veterans they would learn to plan for this, but for rookies it could still be a big downer? Would it be bad enough that the rookies never turn into veterans?

I haven’t fully given up on the idea, so who knows how and where this might show up?

Closing thoughts

Scarcity is one of the fundamental building blocks of board games – something has to be scarce for there to be any competition.

In this post I gave two very concrete examples of scarcity, one that worked and another that didn’t. The fact that I used the somewhat unusual scarcity-of-space hopefully doesn’t detract from the lessons that can be taken from this.

The most important of these I feel is that when working with scarcity, allow your players to work with it, instead of simply having it forced upon them. In this sense there is a similarity to randomness: Forcing players to live with (the outcomes of) randomness is tedious, but once you give them some control after the randomness has happened, the game becomes a whole lot more interesting.

I hope that after reading this you’ll take a look at where the scarcities are in your own game and how your players get to handle these.

Further reading

For more on Voluntarios, read this post which uses Voluntarios as an example for strategy in board games, or this post in which I realized Voluntarios had too few interesting decisions.

And here is the original post on scarcity.

About the author

Bastiaan_smallHi, I’m Bastiaan. The goal of this blog is to learn about game design. That’s hopefully for you as the reader, but just as much for me as the writer.

Insights are scarce so help me to learn? Leave a comment (below) or connect with me on Twitter? You can also subscribe to this blog (see the sidebar) or like it on Facebook, to get updates when I write them.

And perhaps you know of others interested in learning? Share this post using the buttons below:

Facebooktwitterreddit