I’m an ardent supporter of clarity in games, to allow players to focus less on what they can do, and more on what they want to do. A big part of this is to only add rules when they are necessary, to keep the structure as light as possible.
This week, playing Arkham Horror LCG, I found one rule I think you should always add: the “up to”. Compare these two cards:
These two cards are very different, but both share the “Gain X, where X is _____” format. I love those so much, mainly because of how satisfying they are: the timing aspect gives it a push-your-luck vibe, and you feel like you’re getting away with something when you get it to the high values! You feel clever, there are opportunities for awesome combos, and your game is memorable.
Such cards are hard to balance (but, remember, balance schmalance), because the timing aspect is very limiting, but the value is also very variable. However, it feels good to pull off a high number on these, so I’d assume they are balanced assuming you’ll hit pretty close to that maximum value.
Crack the Case is based off of your location’s shroud, which usually hovers between 2 and 5. Search for the Truth is based off of the number of clues you currently have, which is almost limitless, and therefore, they put an upper limit: “up to 5”.
You could say that both are more or less the same. Crack the Case has a similar cap of 5, but it’s hidden: you need to have played a few games to understand what those shroud values usually are. Putting the upper limit on the card feels less elegant, because it’s a rule they have to spell out rather than let you learn, but it also raises the barrier to entry. If an experienced player uses it on a 3 and reveal a 5 on the following turn, it’s a “fun” frustration: you played it safe and shouldn’t have. If a new player does it, it’s just frustrating, because they weren’t shown what these values could be.
Still, you can just say that this is part of your learning process: your second game will be more satisfying, and that’s just part of the depth of the game. Sure.
However, one place where calling out the limit makes it clear is by pushing you towards action. Without a limit, you can spend turns and turns boosting up a single action, which leads to a static, stale game. Sure, that one action will be cool, but will it be cool enough to warrant that big of a warmup?
By putting this hard cap, you tell players exactly when they should stop planning to boost it. Without a limit, players start planning dozens of turns ahead, which means they plan dozens of possible outcomes for each, which quickly pushes them towards AP. With an “up to” clause, you tell your players that they can plan… up until this exact point.
When they get to that point, they use their cards, and then get planning cool combo #2, looking 2, 3, 4 turns into the future instead of 20.
When designing a game, the balance between depth and weight is a tricky one. Every system, element, or even just rule, you add to a game brings strategy, tactics, balance, but also complexity. Fun : weight ratio, which many prefer to think of as the complexity budget, is an important aspect of game design.
Here’s the thing: there’s a lifehack to this. I have a tool to share with you which can add a lot of depth for minimal complexity, in an almost “too-good-to-be-true” infomercial kind of way.
That thing is Space. More specifically, spatial relations. Our brains think spatially, and a lot of these things are so ingrained in us that it’s harder to put in words than to actually interact with.
Think of the tiles in Isle of Skye: it takes a second of looking at it to parse all that it contains, not only the scrolls, buildings and barrels, but also the land types, roads, and their position on the tile. If these were cards in a tableau building game, no mechanism could make up for that depth.
Now, of course, tile laying games are built around that spatial aspect, and not every game can be about tile laying, or route building, or map skirmish. Yet, even game types which are not intrinsically spatial can be enriched, at no extra complexity, by leaning into our brains’ innate spatial understanding.
Not only good for tiles
Early in With a Smile & a Gun‘s development, you would draft a die, and place influence in the district of that value: 6 districts, each associated with a number. It was easy to grasp, but a bit boring, a bit stale.
I added a lot of strategic depth to the game by making each die value affect two districts, and each district be affected by two values (using my favorite tool, the power of combinations):
By using adjacency, players grasped the relationships just as easily as they did the first version. The decision were more interesting, because of the depth this added to the game, but players didn’t have to spend more cognitive bandwidth on this system.
Version 3, the one the final game still uses, adds a lot more depth without much weight:
Now your die moves your meeple around the city, and you place cubes in the row/column in front of it. In addition to the added thematic aspect of moving your meeple, here’s all of the strategic difference this adds:
You affect 3 districts instead of 2: the most obvious one, but a pretty big impact on depth;
Relative numbers matter, not absolutes: going to a spot might require a 2 for me, but a 4 for you, which means I can keep you from going without having to go myself;
Distance matters: You place 3, 2, and 1 cube, starting from your meeple, making the choices regarding the area control more granular, more dynamic;
The possible combinations are clear: not every combination of 3 districts is possible, and that is clear to every player, even on their first game.
Not only did this system add the thematic resonance of the movement, and all of these interesting levels to the game, but by and large, players found this version simpler to grok than the first two. Of course, other effects come into play, but still: this added a LOT of depth, for virtually no complexity.
Even more intrinsic
I’ve brought this up in one of the designer diaries for the game, but it bears repeating here: there are fundamental differences between the corner districts, the side districts, and the central one, simply based on spatial aspects.
First, the 3/2/1 placement mechanism means that while a side district can receive 3, 2, or 1 cube, depending on where you’re hitting it from, corners can only get 3 or 1, and the center only gets cubes by 2s. It affects how swingy the majority for those districts are.
Second, the distance between two spots that affect a given district vary. The 4 spots to influence the central district are exactly 3 spaces apart. However, after you pass a corner, it can take a while before you get back to placing on it, especially if you need to place 3 cubes.
Now, these aren’t large differences, but they’re still there. What’s more, I didn’t even design them in: it’s just inherent to the spatial design of it. It took me a bit of time to even realize it was the case.
Remember when we were talking about depth : weight ratio? I added depth to my game without meaning to. Isn’t that some magical, snake oil kinda tool!
Tzolk’in famously uses these large, interconnected gears to represent the passage of time. In addition to the gimmick and eye-catching aspect of the gears, this entire system would have been so much harder to represent without a spatial aspect.
Pandemic is one of many examples of games which feature movement on a map, and you could say that’s an inherently spatial game and I’ve broken my own premise. However, I think it’s worth pointing out how the links between cities naturally create different experiences based on which cities are targetted at setup: 3 cubes on Santiago, with its single path out, is not the same as on Baghdad, with its 5 neighbours, or Madrid, which also has 5 neighbors, but 2 of which are of different colours. If the map didn’t include those spatial oddities, the setup variety would not be as meaningful.
Battle Line (also known as Schotten Totten) is, to put it simply, a mix of War and Poker, where the two players play cards in front of a line of 9 flags, trying to make a better poker hand than their opponent. To win, you must either gain 5 of the 9 flags, or 3 adjacent ones. That extra winning condition adds a lot of tension, makes some sites more important, makes the ends of the line feel different from the middle, and all of that at no extra cognitive load.
Think about the design(s) you’re working on right now: is there a part of your game that you could improve (either by simplifying without making less engaging, or by enriching it at low complexity costs), simply by presenting it spatially?
“Gain X points if you have the most Y” is a phrase that sounds familiar to every board gamer. Majorities are a great way to add indirect interaction and tension to a game, and is often used as either a central mechanism, or a secondary way to score points where it serves to obfuscate an item’s scoring value. I have used many majority aspects in my games, and have learned a few lessons along the way. This series of posts is meant to highlight those axes on which you can play with your system and offer different experiences for the player. Today I’m talking about choices which affect the cognitive load of the mechanism:
Can it be secured? In Acquire, there are 25 shares of each color: at most, I need 13 to secure a majority. Sometimes it’s not a maximum amount to be spread out, but a maximum value a player can have: the Cult tracks in Terra Mystica, and the UN track in Energy Empire, both have a final spot only one player can reach.
Securing it allows you to focus on something else, but it’s also resources and/or actions you could have used to do other stuff: mitigating that risk becomes part of the strategy. It also allows you to file it away in your brain and keep on focusing on something else. It allows you to better evaluate an action’s value: if I need to invest 3 actions in this to gain 10 points, it’s worthwhile, but it isn’t if it takes me 8. And two players who keep one-upping each other is funny to watch for a bit, but it’s slowly taking them out of the running.
If it can’t be secured, that means the tension will never relent. If it’s a secondary scoring aspect, or if you want a more tense, more chaotic, more tactical game, it’s probably not worth adding that math element and making players count those items, because if they can, they will. If their focus is better served elsewhere, don’t give them that information.
How many actions are we supposed to invest in this? It’s a related question. In With a Smile and a Gun, both players compete over Business tokens, which score points for whoever has the most of each type, and Reputation tokens, which are just worth straight points. I tried multiple things for Business tokens, specifically in how many of each type I put out there. In a version, I had 8, 10, and 12 tokens of a type. This diluted their value so much that no one ever cared about them, no matter how many points the majorities were worth. Even when they were brokenly overpowered, players still preferred the sure thing. “Who cares about this one, I can get the next one instead” is a line I heard times and times again. It also often led to majorities being clinched very early in the process, making them worthless for the second half of the game. In the end, I brought them down to 4, 5, and 6, and
I think it’s partly risk aversion, partly optimism, but also related to the fact that, without the majority, those tokens were worthless. In Clans of Caledonia, your main way of scoring points is completing Contracts. In addition to their standard bonuses, the player who completes the most gains an extra 8 points. In games like this, it doesn’t matter if the majority will be won with 4, 8, or 14: you’ve already gotten stuff from your investment, that majority is just icing on the cake.
These are the aspects that go through my mind when I add a majority aspect to my games. Like most great mechanisms, majorities can be customized to do exactly what you want them to do. Are there other twists on the mechanism I have forgotten?
Tuesday’s post was presenting two types of non-decisions, which look like decisions from the designer’s POV, but not for players when they experience the game. Those were Arbitrary and Automatic decisions. Today we talk about avoiding them.
So what causes Automatic or Arbitrary decisions? They are two extreme of the same spectrum: Automatic is what happens when you have perfect knowledge of the impact of your decision -making it just a matter of comparing the numbers and choosing the one with the most benefits; Arbitrary is what happens when you have no knowledge of the impact of your decision -meaning all options look the same.
Therefore, stand in the middle. Done, blog post over. Nailed it!
You want a bit more? Oh. Okay.
So I think there are four important things to keep in mind in avoiding non-decisions, and in each, an example of a game I love which does it well:
Opacity is how I describe a game that, through layers of complexity, makes it hard to evaluate impact because of the limits of the human brain. Opacity is often achieved through multiple layers of math, mostly in economic games, but can come from important mechanisms which are not apparent from the get-go. Some will argue it’s a feature because of how clever you feel when you understand it, but I think it is quite problematic: either it leads to arbitrary decisions, because you can’t figure out what each move is worth, or it becomes an automatic decision once you can. I am a fervent advocate of games requiring as little brain juice to understand the rules, to have more left to understand the strategy: to me, opacity is a necessary evil that should be limited as much as possible.
A game that achieves depth without being too opaque is Russian Railroads: the impact of an action are as straightforward as can be, whether it’s gaining a new piece, or moving a piece forward. The impacts of a move are also straightforward: this will give you an extra point per turn, but gets you closer to you !-bonus. Instead of having very complex, interconnected systems, it presents them to the players in a really simple way.
Comparability is the ability to compare results. Back when I played Dungeons & Dragons, I had this buddy who just loved big numbers: he had an attack that dealt 20 damage, an attack that dealt 15, and one that dealt 10. Every combat, he’d do his 20, then his 15, then his 10. On the other hand, I had an ability to push enemies, one to stun them, one to teleport to another enemy I could see. Which one was best depended on the situation. Comparing them was more about gut feeling, making the decision interesting.
Wingspan is a great example of a game that makes choices hard to compare: even leaving out round objectives and egg maximums, are you going for the 2-food, 4pt, forest bird which gives you a bonus card; the expensive, 7pt plains bird which can give you an extra point every time you activate it, or for the quick to play, 3pt forest bird which has the same special ability, but a lower chance of success? How are you even supposed to compare those? Because they vary on multiple axes (without becoming opaque), and because of the special abilities which feel very different, it makes those decisions based on the situation rather than a clear cut better option.
Uncertainty is not having an exact value for an action, because its value depends on other things: a random factor (draw the top card from the Building deck and build it right away -but how good is that card?); other players’ actions (the player with the most Honor gains 10 points -how safe is my lead?); or your own future actions (gain an extra wood every time you go to the forest -how often will I get it?).
One thing to keep in mind is that uncertainty can lead to arbitrary decisions if taken too far, and if it isn’t combined with another mitigating factor. “Gain a point every time a blue card is played” is an interesting effect of uncertain result, but if I had to choose between that and “gain a point every time a red card is played”, unless I know what people are likely to play, then it’s arbitrary. If you mitigate it with comparability (gain a point per blue card vs gain a wood per red card), then it’s less arbitrary.
One game which uses uncertainty to make decisions interesting is Libertalia. Because of the simultaneous selection, and despite having perfect information of what others have in hand (barring some specific special effects), you can never be certain what the exact impact of your move will be, because the effects of the cards are interrelated.
Interchangeability is a really long word. I really like Euro games, but too many of them end up increasing the number of resources without making them feel different, and in so many games, the choice of a wood vs a brick is the color of the cube. Many games even allow you to pay resources of any type for a lot of things, making what little difference there was even thinner. Most of the time you care because you’re building towards something specific, but by giving each resource a specific niche, you’re making sure those choices stand on their own, not only when an objective requires a specific one.
Le Havre is an example of a game with a load of different resources, which each feel different, mainly through its two-level, multi-use system. Resources are upgadeable, each type leading to a specific “second level”; some are used for building, some for food, some for energy, and most change use on the second level; some reproduce on their own, while most don’t. All in all, it means that the decision between a Fish and a Coal is never arbitrary, and never automatic.
I think by keeping these four things in mind, the ends of the spectrum become easier to avoid.
One of the most discussed difference between hobby board games and mass market is decisions. When in Monopoly you roll a die to see where you go, in Tokaido you just… choose where you go. That choice is, really, what modern games are about, where strategy and tactics and story and interaction come from. Recently, I’ve been testing a lot of less-than-polished games from other new designers, and I’ve seen two types of non-decisions: they look like decision, but really aren’t.
I’ve played a lot of prototypes that were spinoffs of the card game War. Imagine a game where you had a handful of monster cards, and played them against one another in duels. One player would play one, and the other would respond, highest number gains a point, discard the cards, go again. This would be a crappy game, and a perfect example of both types of non-decisions: the first player had an arbitrary decision, and the second, an automatic decision.
An arbitrary decision is one where you have insufficient information to make a decision. Not incomplete information -that is actually interesting-, but insufficient to push you one way or the other. If I knew what my opponent had in hand, then I could try and strategize which monsters I’d force them to play, opening things up for the following duel. If the monsters had specific abilities outside of the duel, I could base my choice on those abilities. If the prize differed from one duel to the next, I could choose when to go all in and when to play it safe. But as is, no information means my decision is based on… nothing at all. And therefore, a non-decision.
An automatic decision is the opposite. It’s not only perfect information on the situation, but also clear, directly comparable result. If you play a 6, do I have a card of higher value? Then, I play the lowest one that is still higher. If not, I play my lowest card. That’s it: sure, I can play my 10 on a 3, but that’s just sub optimal.
And you might think “of course, that’s a crappy example of a game just to make your point”, but really, (1) I’ve played games that are pretty close to that, and (2) you’ve played a few prototypes, and probably even a few published games, that rely on non-decisions like that. Will you turn right or left at this intersection? Will you take a bonus red or blue cube? A few decisions like that in a game are annoying, but too many and it’s a problem.
That begs the question: how do you avoid those? If you want to be on the spectrum between arbitrary and automatic, in the good, crunchy Goldilocks zone, how do you make that happen?
Well, as a true blogger, we’ll look at that on Thursday.
The goal of this blog is to talk about game design, and the lesson I’ve learned during the process. So why not start with the latest lesson I learned. I had to learn it a few times, too: a mechanism that players don’t interact with is just artistic pretentiousness.
While working on Cybertopia, we ran into this situation twice in the last month. The first time was with the Corporation boards, where players place their Virus tokens, covering certain spots for instant bonuses, and eventually trying to have the most Viruses on it when the Corporations goes in lockout.
While working on Cybertopia, we ran into this situation twice in the last month. The first time was with the Corporation boards, where players place their Virus tokens, covering certain spots for instant bonuses, and eventually trying to have the most Viruses on it when the Corporations goes in lockout. We had 4 Corps, so it made sense to us to have each look different, work differently, and trigger scoring from a different mechanism.
We felt pretty good about it, until we showed that version to a friend: it took me the regular 8 minutes to teach the game, but then it also took 6 minutes to explain how each Corp worked. They liked the idea, and actually found it interesting for every one of them to feel separate. But here’s the thing: it never guided any strategic decision. You didn’t play on the race because it was a race, or on the square because it was a square: you chose which one it was because of the bonuses you could get, and the opponents’ positions on them. It’s not just that it added rules but not strategy: it’s that it was a mechanism which only existed to make us, the designers, feel smart.
We’ve since made the change to make every Corporation work in the same way: a 5×5 square, where players start in the middle, and which scores when 2 lines (rows or columns) are completed. We lost maybe 1% of strategy, but cut the length of the teach by close to half, and the rules mistake by a similar margin. Sure, the system offered differences, but never influenced how players played: that’s where the word interact from the intro comes in.
Worst thing is, we then did the exact same thing: we had a set collection aspect (which is since then gone, but the point still stands), where in order to make every set feel different, we gave them all a different mechanism:
Blackmails score 1pt per Secret you own;
Blueprints score 15 points if you have 4, 0 if you have less;
Dossiers score triangular points (1, 3, 6, 10…);
Financial Info comes in 3 flavor, and scores a lot, but you only score the flavor you have the least of.
Sure, it looks cool, but in the end, it’s all set-collection. It’s just “get a lot of these”. So we instead went with streamlining them all to the same mechanism: A scores if you have 2, B if you have 3, C if you have 4. Again, the difference drove no decision: you still wanted to have a bunch of them, and they were not all worth the same to everyone -which is super cool!-, but again: no interaction with the players. It just was there.
It reminds me of when I was DMing D&D and used packaged adventures. They had huge backstories for the bad guys, pages upon pages of who they were, and how Bandit King was Random Bandit #3’s son, and Daddy Bandit just found out his wife died, and then PCs come in and fireball and it never mattered.
In each of those cases, it was technically different: the Corps had a different feel; the Set collections had different risk-levels; the PCs could have a discussion with a faceless goons and use this knowledge to their advantage. But in reality, it’s like Schrödinger’s cat: until they interact with the differences, those differences don’t actually matter.
Except to us, game designers, who can pat our own shoulders and yell “LOOK AT HOW CLEVER I AM!”