Reposted for discussion:
I do not like Betrayal at House on the Hill. The hook for the game is that halfway through the game a “haunt” will begin and the rules and goals will change to a scenario randomly selected from a booklet. (typically involving a player turning against the others) The issue I have with it is that this event – which is perhaps the most impactful random event of any board game – is impossible to predict. This makes the first half of the game feel like an extension of the game setup, which isn’t helped by the random way resources are distributed to players; there are very few meaningful choices a player can make before the haunt begins.
When a random event has a big impact on the outcome of the game, the game can feel arbitrary. This is especially true in cooperative games where the only opponent is the game itself. Because every mechanic in a board game must be resolved by a player, game AI tends to be simple. To introduce surprise against a predictable opponent, co-op games may have random events of some sort, often using a deck of event cards. But losing because the game drew a particular card at the exact wrong time can feel unfair. The problem is that players are robbed of their agency since they can’t make informed decisions about future events. The solution is to foreshadow these events – let players know what is coming so they can prepare.
Foreshadowing in games needs to be much more explicit than in literature. Where authors may hint at future events, game designers should explicitly tell players what is happening next. Showing players the near future enables them to make meaningful choices about how to react. Hidden events are no more than a guessing game.
The simplest way to foreshadow events is to show players a preview of a random event – for example, keeping the next card in an event deck revealed. There are numerous variations on this – in Pandemic, players play until an epidemic occurs with no idea of where the next infection will strike. After an epidemic, all infection cards seen so far are shuffled and returned to the top of the deck so that players have a general sense of where infection will strike in the near future. Not only does this ensure that infected countries will worsen, it also gives players an idea of what to expect for the next few turns.
Another option is for events to have small consequences when first revealed and then to escalate over the next few turns. This is effective because the importance of foreshadowing scales with the impact the random event will have on the game. Trivial events do not feel unfair when they come out of nowhere because the player doesn’t care about them. Events that can cause players to lose the game feel much worse when there is no way to anticipate them. By starting off small and escalating, the amount of warning the player has for each stage of the event scales with the threat it poses, and the warning is reinforced by the in-game effects of each stage in a visceral way that just showing a card cannot do.
This approach is at the heart of Spirit Island, a co-op game about defending an island paradise from European invaders using magical spirit powers. Every round, the invaders draw a card to determine the type of land in which they will explore. Explorers are weak enemies, posing no threat by themselves. However, the card used to determine lands for exploration in one round determines where the invaders will build in the next, and where they will ravage the round after that. Building entrenches the invaders, making them much harder to drive away, and ravaging blights the land, destroying player resources and bringing the players one step closer to defeat. The brilliance of this system is the natural way it warns the player what will happen next. Where the invaders will build and ravage is revealed not just by telling the players, but by walking them through the mechanical process of the invaders settling in, building up, and blighting the land.
There is a limit to how much foreshadowing players can handle. Revealing the AI’s moves for the next ten turns is overkill. It encourages analysis paralysis by overloading the players with information. Only the near future needs to be known; uncertainty should come from not knowing what will happen in the long term, not from being blindsided by everything the AI does. Short-term foreknowledge lets players make intelligent decisions about what to do now, while knowledge of the game system itself lets them plan for the long term.
Foreshadowing is just as important in competitive games with random events as it is in cooperative games. For example, Galaxy Trucker involves building spaceships out of ship component tiles and then evaluating their performance by going through a deck of random events. During the build phase, players may spend precious time to look at a subset of the upcoming events in order to tailor their ship to those threats. While new players ignore this option, experienced players can skim these cards and use the information to build the right ship for the job. This gives players the satisfaction of seeing their plans succeed (or fail hilariously) on their own skill, rather than the luck of the draw.
Recently, successful indie computer games have taken advantage of foreshadowing to great effect. Slay the Spire and Into the Breach are two examples of games where knowing what the AI will do is a central part of the game. Both games have an “Intent” system in which the next move each enemy will take is clearly displayed to the player. The designers of Slay the Spire found that the less uncertainty there was about what enemies would do, the more their playtesters liked the game. [1]
Randomness for the sake of surprise alone adds very little to a game and can actively detract from the experience. Predictability is not always bad, nor is it all or nothing – an event can be both surprising and predictable if players are given a little time after its reveal to prepare. The best games give players the information they need to make meaningful choices.
[1] https://www.rockpapershotgun.com/2018/02/19/why-revealing-all-is-the-sec...
Hey, that's a pretty cool idea. It is also nice because it avoids overwhelming the player with information, since it summarizes the threat without providing exact information.
You're right that most card games try to make all the backs the same so you can't tell what is coming. There are some exceptions like Galaxy Trucker, but even then you only know the level of the next encounter (which isn't enough to help you make a decision before seeing it).