On the surface of this question it may seem to be rhetorical or some sort of philosophical jest to ask what dice do, but it is not. It is actually a question I have long wondered about, but never wanted answers to. In part because the answer is a bit more complex than it looks. While dice are random number generators with fixed ranges, there remains the question of what do they do in a game? A joke answer is that they don’t do anything but sit there. The dice tumble and roll by shaking and tossing them with the result being a random number that doesn’t convey anything by itself. Sadly even in the confines of a game’s rules set, the answer isn’t as clear as we’d like.
There are all sorts of random tools one can use in generating the randomness a game needs. For this reason the choice of dice as the workhorse of game mechanics does not immediately stand out as a contributing factor for the prevalence of dice. Random options include coin flips, numbered chits drawn blindly, cards, and spinners, to name a few. Yet despite the extensive list, many are inadequate for most games and the needs for players. Something must truly set the polyhedrals apart.
Let’s look at some features common to all randomizing tools and see if we can better understand what sets dice apart. All randomizers discussed above have fixed, absolute ranges. All of them can be manipulated in specific ways. Where they differ is in the mechanics of randomizing that affects them all. Coin tosses spin through the air, chits are randomly drawn, cards drawn, spinners flicked/spun, and dice thrown. But size becomes a factor in game design. If you have a series of random ranges (as in roleplaying games), a set of chits for every range would be needed to prevent a lag in game play. The same is true for spinners and cards. Dice however take up the least amount of space in a game box.
None of this is to say that these other randomizers aren’t valid systems. This point must be held in mind: the tools used must integrate into the game’s rules and not distract from the experience. If there is little to no modularity in the randomizing tools in the game’s rules, there is no inherent advantages to one randomizer over another beyond probability distributions. After all, the main function of randomizers in games is to serve as impartial judges in areas where conflict between players is likely to occur.
Despite our claims to the contrary, people are vested in their own self-interest. We may claim to be able to avoid personal bias in games and sports, but few can say they have mastered their egos enough to support the desire to win unless they are paid not to. This is surprising behavior for an apex predator species when you think about it. We condescend to allow random chance to determine outcomes and then abide by those choices even if unfavorable. Of all the forms presented here, dice give the greatest semblance of loss of control to chance. It is this sense of fate and its associated sound that helps reinforce what will be explained below as the fungibility possessed by dice that other randomizers do not possess.
So, randomizers are intended to inject chaos into the organized structure of a game’s mechanics. This can include how quickly one can get from start to finish (speed), which way to go (movement), environmental changes (rules) and even pass/fail sequences. The point is to remove arguments and force players to work with adversity by putting arbitrary elements in the hands of an impartial mediator. In this case, none of the randomizers chosen has an advantage over the others.
The question begs to be answered as to why dice are so heavily used. The limitations on the other randomizers physically or mathematically play a role in this reliance on polyhedrals. Let’s start by eliminating a few randomizers and the reasons for why they aren’t used often.
The coin flip has a binary function and is easy to manipulate or feels too limiting. Sure, you can flip several coins and count the number of heads or tails, but this has a serious bell curve flaw that it prevents a lot of the lower and upper ends from occurring. Flip four coins and there’s a 1-in-16 chance of getting a 1 or 4; that’s a 6.25% chancd! Roll a d4 and it is 25%. Not a good spread if your game requires something more varied to keep players interested. Coins can also be palmed, which makes cheating more likely to occur.
The next two are spinners and tops. Their similarity as randomizers are enough on merit conflating them. Both are based on rotation to generate random values. Axial rotation means all of the friction generated to create a random result is on a fixed point. The wear over time means the random values will never be the same despite the distribution of values being equal. This is usually a result of the materials involved. Spinners usually loosen over time while tops have their points wear down. Spinners often take less time to give a result and there is no risk of falling off the table, but are more prone to manipulation. Thus, the time and manipulation issues keep tops and spinners from being used in a wide variety of games.
Chits are only as useful as the material they’re composed of and the manner in which they are marked. If made of chipboard or other paper material, the pieces eventually degrade. Should the numbers be embossed or engraved, a desired number can be picked by feel alone. Eliminating these issues, a game has to rely on using a single set of chits or cause the game to slow down by ensuring only the correct values are available for each draw. Otherwise, you need several sets for different ranges. And these concerns don’t include the possibility of losing a chit.
Cards are perhaps the only randomizer able to give dice a serious challenge. Cards have numerous ways in which they can be used to generate random results from the size of the deck, frequency of occurrence, and values printed on the cards, a deck is a surprisingly effective tool. It’s pretty easy to argue that cards are potentially more versatile than dice. Such argument has merit when you look at all the ways cards can affect games from a random rule change to determining outcome of an event or a hand of play.
Unfortunately, cards suffer from the same issues as chits when it comes to variable ranges. Unless you have multiple decks of varying sizes, cards can only give a static range, which is fine for many games, but is a restriction on complexity. Even with suits, the cards will support only a few types of variables in any given deck. And those rules changes, well unless that is one of the game’s design features like Fluxx, amounts to a very confined mechanic which will need to occur just often enough to not exhaust the deck’s possibilities or make its use irrelevant, which is what some board games do (Starcraft, Twilight Imperium). Either way, these uses impinge on the use of cards as a universal tool.
This, then, is why dice are so popular: dice take on whatever you superimpose on them. A six-sided die can be used in a variety of ways to consult tables of results, determine success or failure, or the distance moved on a playing space as easily as any other sized die or combination thereof. A handful of dice of varying sizes provides a nigh infinite range of possibilities in such a small package that the economics alone make dice ideal. When you consider how easily each table can be customized, it’s little wonder dice are the workhorse of gaming.
When you factor in the variety of ways that the probability mechanics can be designed (flat curve, bell curve, scalable bell curve, exponential, etc.), dice allow for the greatest range of flexibility at a designer’s disposal. Superimpose the rule on the mechanic based on the situation, and the potential of dice outstrips any other randomizer. Thus, what dice do is perform as mimics as they take on whatever characteristics are needed. When you place symbols on the dice rather than numbers, the dice can even model (albeit imperfectly) a deck of playing cards, the letters of an alphabet, and so forth. No other randomizer can accomplish this in such a confined package, which makes dice all the more attractive and the optimal choice since they can also mimic elements of other randomizers.
So, what do dice actually do? They act as scaffolding that take on whatever skin is wrapped around them. Dice allow us to simulate and interpret likelihoods of outcomes with the least amount of retooling; and in the constraints of many budgets, that makes all the difference.
Previous:
Anatomy of Game Design: An Unbridgeable Divide, Part 8.2
Next:
Anatomy of Game Design: Structured Chaos