RPG Errata: Committed Relationships with Dice

Dice are important to RPGs.^{[citation needed]} I’ve talked a lot about the arbitration between narrative and mechanics, but whatever balance a specific RPG system strikes, it is a nearly universal rule that the mechanics require some form of randomization. The traditional method; dice.

This is a very long-winded way of saying almost every RPG uses dice to help decide what happens in the game.

I’ve talked a bit about dice before, but now I’d like to focus on the more practical application of dice in a game. Fundamentally, the purpose of dice is to make the outcome of the story uncertain. Note, not the game. For it to be a game at all, the outcome has to be uncertain, and since RPGs are collaborative, the only uncertainty available for the story is through dice.

Of course, there are hundreds of ways dice can create random results. The majority of RPGs use a “Target Number” mechanic, where there is a number that means “success” of some kind. The players roll dice and calculate their own Number, and hope to roll over or under the Target.

Even this simple framework has hundreds of permutations. D&D uses the 20-sided die and adds a lot of bonuses. Exalted uses varying numbers of 10-sided dice, and you count the number of 7+ results, rather than adding them up. Some systems have varying Target Numbers, others are static. Some have two Targets, one that is a “weaker success” and one that is “stronger.” Some have the players roll multiple dice and keep some number of the highest. Some give the numbers on the dice results on their own, so a 4-5 is a normal success, a 6 is a great success, and anything else is a failure.

Some can get quite complicated. Look at Shattered, where you add your attribute to your skill, check that number on a chart, and then roll anything from 1d6+1d4 to d10. It’s a number-crunchy mess, but it certainly gets you a random number. And can you honestly say it’s better or worse at getting a good bell-curve of results?

You see, we humans are bad at probability.

Lots of ink has been spilled talking about how we, as apes with ideas above our stations, have a poorly crafted understanding of chance. Anyone who has ever played X-Com knows about this: when we see we have a 75% chance of hitting a foe, we feel cheated if we miss three times in a row. Missing at all on a 95% chance feels wrong, more than once in a game? Forget it!

But count up how many shots you make in an X-Com map. Ten? Twenty? If you make twenty shots, and all of them are 90% to hit, you should miss two of them, statistically — but that’s the problem.

Lies, damned lies, and statistics.

One of the primary rules of probability is the inability of one roll to affect the next. If you roll two-hundred 1s in a row, your next roll is just as likely to be a 1 as a 20. But we know better, don’t we? Because if you roll two-hundred 1s in a row, the die is obviously loaded.

Nevertheless, this is the gambler’s fallacy: The chance of rolling two-hundred 1s in a row is exactly the same as rolling one-hundred and ninety-nine 1s and then one 20. In fact, the probability of rolling two-hundred 1s is the exact same as rolling any string of two-hundred rolls.

But we’re not good at thinking like that, and true randomness doesn’t help. The truth about a string of random rolls is that — by virtue of being completely random — there will always be a bit of order somewhere in it: five 10’s in a row, six rolls that are a descending straight, eight alternating 1s and 20s…true randomness doesn’t care.

One of the biggest questions in game-design — at least, game design that displays percentages and deals with random results — is whether to use true-random or pseudo-random results. Pseudo-random results are results that “feel” random, but avoid that pesky order. Rather than reject the gambler’s fallacy, Pseudo-random results embrace it, and make sure that the chances of you rolling the same result multiple times in a row is measurably less likely. Extreme Pseudo-randomness might even make sure that if you roll surprisingly low, you are more likely to roll surprisingly high next time.

But pseudo-randomness is the purview of computer-games. Cold hard dice can’t be pseudo-random. So what happens then?

Confirmation Bias is a crippling short-circuit in our brains. Put simply, Confirmation Bias is a tendency in the human brain to seek out, remember, and interpret information in a manner that proves “we were right all along.” Have you had a string of good luck in your rolling? It’s possible, but it’s also possible you just aren’t counting that one 3 that you rolled, because no lucky run is perfect, right? And sure, that 12 wasn’t really a high roll, but it was all you needed thanks to the bonus you got from your allies. And you don’t even remember those two 10s.

It works in reverse too. If you roll three 2s in a row, you might not count the 19 if it’s followed by a 7, a 4, and then another 2. “Gosh, I’m rolling nothing but 2s!”

This can affect your judgment. If you’re rolling poorly, might you hold back on high-risk high-reward efforts? Or push you to higher gambles if you’re rolling well?

But the Gambler Fallacy rears its head again in a metastasized monstrous form. The dice not only remember what they’ve rolled, but they can be influenced. There are rules for this: we know you don’t touch another person’s dice. We’ve consigned dice to the trash bin because they had “lost their luck.” We’ve bought new dice and “warmed them up” so that when the time comes, they do what we want them to and kill the dragon with a final stroke.

The folk behind HackMaster have helpfully codified these rules into a tongue-in-cheek document that you can peruse at your leisure.

I mean, of course it’s tongue-in-cheek, isn’t it? Proper dice-blowing technique? Two-fisted Monkey Roll style? Cordoning off DRZs (dice-rolling-zones)? 10,000 sided dice? This is satire: a loving look at all the little folk-rituals we’ve built up in our community, made comedic by their codification into a book-of-best-practices.

None of this works, right? Blowing on dice doesn’t do a darn thing. Good Rolls aren’t tangible things that can be brushed off or blow onto dice. You touch someone else’s die? So what? Dice don’t care, they create random numbers. And yet, the HackMasters have a whole section for how to roll a die badly, and Computer generated rolls is right there.

The scientific truth of the matter is that random computer generated dice rolls can be made more un-biased than a plastic die that has been rolled, cracked about, and wasn’t even perfectly balanced in the first place because plastic is never perfectly uniform. Even carving the numbers offsets a die’s center of gravity.

Nevertheless, quote: “Insist on real dice, anything else is OFFICIALLY against the rules.”

Let’s talk about ritual.

Ritual is sometimes looked askance at by our modern sensibilities. The recitation of a bible-verse, the quoting of a speech, the shaking of hands…what does it actually serve? Dissertations have been written on less.

What about Dice? All these rituals of how to “clean” dice, borrow luck from other lucky objects, proper dice rolling technique because rolling improperly will result in worse numbers — because somehow the dice know what numbers are “good” in any given context — all these rituals buy into the gambler’s fallacy and cognitive bias of dice: that luck is somehow acquirable.

But RPGs are, and always have been, games of imagination. They are escapist, yes, but fantasies of empowerment and achievement. Even the darkest games are rife with means and methods of players enacting their will on the story in a way that would never be possible in “the real world.”

In most RPGs, a huge point of the game is allowing the players the chance to affect things that they otherwise cannot. They can save lives, protect ideals, and amass power and wealth in a just and righteous way. They can change the world. Is it any wonder that this desire to control what cannot be controlled spread to the dice as well? The one part of RPGs that can flat out say “no” to your efforts?

It’s longstanding tradition/superstition in theatre that — unless you are in the middle of a performance — you do not say the name Macbeth in the building. If you do, you’re supposed to turn around three times and spit to rid the show of bad luck. No one really does, anymore; now you just get jokingly admonished, or tsked at, or even ignored.

Except I have a friend who won’t ignore it. If he hears someone say Macbeth, he will demand they turn three times and spit. Incessantly.

He knows it doesn’t actually do anything. The art of theatre is flooded with mishaps, misfortune, and clumsy mistakes. No play is made more or less dangerous because of saying the name of a Shakespearean play, but he does it anyway. Why? “It’s our heritage,” he told me once. “It’s something that’s passed down through actors and playwrights for generations. It’s a connection to what it means to be in Theatre.”

I think there’s some truth to that. I think also there’s some truth to the idea that it’s just more fun. Regardless of its practical efficacy, there is something far more engaging about shaking dice and rolling them on the table than sitting at your desk and clicking the mouse on the “roll attack” button. It’s the difference between an .mp3 and a live concert. It’s the difference between watching baseball on the TV versus sitting in the stands with a beer and brat on your lap, hearing the cheers of the crowd and the crack of the bat echo around the stadium. It’s the difference between a vid-call on your phone, and being in the same room with your friends, talking, laughing, and sharing each other’s company.

So maybe the rituals of our dice aren’t something we need to be embarrassed about. I mean, we don’t really believe that we can personally take out a dragon with some chainmail and a spellbook, do we have to really believe that rubbing the dice counterclockwise over a lucky autograph makes a difference?

Or is it all part of the same game?