Recently on the YouTube channel that I operate myself and games designer Mike Hutchinson had a conversation about dice pools, which you can view here: https://youtu.be/_jRw3-WTF8A while it’s a great conversation, it doesn’t cover everything I wanted to say about dice pools as game design element (I know it didn’t cover everything Mike wanted to say either, but he has his own blog. As such, I thought I’d post this blog up to cover some of my further thoughts on the subject, and hopefully make a few useful points.
For those who aren’t aware, a dice pool is quite simply a game mechanic where players pool together a set of dice, roll them all, and then look for a number of results over a certain value to achieve a goal or series of goals. For example, a game could ask players to roll 10 six-sided dice (d6 from now on) with a success being any dice that scores 4 or more and it might then indicate that if they get 5 or more successes then they were successful in what they were attempting. Dice pools have a few mechanical advantages, but the most significant is probability control.
If a single dice is rolled the odds of getting any given number is the same as that of getting any other number, which is terrible for game design. Game designers want to give players a certain experience, and to give that experience they need to have some chance of predicting what will happen to their players. That means that totally random results are difficult to use consistently to produce a desired player experience. There are a few ways of controlling the probability of a dice the most common being:
1) Rolling more than one dice and totaling the results, since a 7 is far more likely from rolling 2d6 than a 2.
2) Rolling lots of dice at once, since with more and more trials result will tend towards the mean.
I’ll just explain those a little bit. When rolling 2d6 and adding the scores together the result can be anything from 2-12, but while a single d6 has one way of getting each of its results, each with a probability of 1 in 6, 2d6 have 1 way of generating a 2 but 6 ways of generating a 7, with each single method having a probability of 1 in 36. So, a 7 has a probability of 6 in 36 while a 2 or 12 has a probability of 1 in 36.
Similarly, when rolling 6d6 there is only one way that all the dice can come up 6, but there are many ways for the dice to come up with the numbers 1-6, and even more ways for them to come up with three dice scoring 4 or more. The more dice are rolled, the more likely it is that they will tend towards the mean because to avoid doing so requires a greater level of order.
Tabletop miniatures wargames rather like a big dice pool to control its probabilities, this is for a few reasons. First of all, because they tend to have lots of events happening at once, such as when two big units of probably multiple combatants all fighting together. When this happens, it is simply tiresome to roll each event one at a time, so all the events get rolled at once, this practicality makes it impossible to use adding together dice to control the probability. Clearly, if a player were to roll 4 2d6 it would be impossible without colour coding the dice to know which sets of 2 belonged together. Instead, dice pools are used. In addition, tabletop miniatures games tend to be played on large tabletops with quite a lot of spare room. Rolling 10d6 on a tabletop with the average Eurogame on it will most likely result in so many dislodged cubes as to make play impossible.
In theory one of the advantages of a dice pool is that they should be able to offer a wider or narrower range of options without confusing players more easily than combined dice. Which is to say, 2d6 provide a spread of eleven options, which is a decent range, but its not huge. Most players understand the spread of odds on a 2d6, for example, I know that a 10 or higher has around the same odds as rolling a 7. If a game wanted to offer more than 11 options and so tried to offer 2d8 so on up to d20, while I assume that 9 is most likely on 2d8, off the top of my head I wouldn’t know what number or higher than mapped to rolling a 9, certainly not sufficiently to have a feel for it on the fly at the tabletop. However, with a dice pool, even with d20 I know the odds of rolling a 20, or an 11 or more, or a 16 or more or so on. The problem with using other dice in dice pools is quite how many of them are needed. For tabletop miniatures games, which generally require players to source their own dice, asking players to acquire multiple non-d6 dice is a big request. Also, for a dice pool to even approach effectiveness it requires at least as many dice as would be needed to cover a full distribution across the dice’s faces, and preferably many more. Very few people own 20 d20, and rolling them is far from easy. I find d12s to be very pleasing as shapes generally, but having rolled 12 or more of them in a single go a few times while testing A Billion Suns I can tell you, it’s not a great experience. Dice with more faces are simply not designed to be grabbed up and scattered en masse around a tabletop, they’re not even designed to be transported en masse for that matter.
So, practically, its actually not really possible to use anything much other than d6 for either method of probability control, so how best to offer a wider range of options? The option that dice pools offer is an essentially infinite range of options to players by simply giving the chance to add more dice. For example, with 2d6, aside from a few re-rolls the only option for adjustment is to shift the target number up to 10 places. With a dice pool, if the rules request that a player rolls a pool of d6 needing a certain number of results, while it can only offer an adjustment of up to five spaces on the results, it can offer the chance to add more dice to the pool as long as more dice can be found, and when the physical dice run out, they can always be re-rolled. In addition, everyone understands the odds of each dice, and indeed the improvement in odds as each extra dice is added.
Tabletop miniatures games tend towards using dice rather than cards, simply because they tend to require players to acquire their own components. That their greatest strength comes from constant addition of components, this makes them often less tempting to designers of boxed games, those that do use them tend towards making them a major feature of the design. Sadly, this has led to a negative perception of dice in boxed games, since when used they tend to be entirely random, and so lead to spotty player experiences. In short, if designing a boxed game, consider the option for using and the strengths of employing a dice pool. If designing a tabletop miniatures game, understand the strengths and weaknesses of the inevitable dice pools.