Thoughts on Dice – Desirée and Jonathan

Hello dear reader,

I’ve been talking to Jonathan about how we need to post blog entries, and I was going to talk to you about art and he was going to talk to you about some element of game design, and neither of us have actually written those entries. With the launch-date of our Kickstarter looming, I had a thought: our first blog entry should be written together! So I’m going to begin a discussion, between the two of us, about dice.

Why dice? I’m a fan of board games, and very much interested in game design. I’ve come across various attitudes about dice. Some people love them. Some people can’t stand them. Others are in between.

Since The Island of Doctor Necreaux: Second Edition uses dice, I thought it would be fitting for Jonathan and me to weigh in on the debate. Dice: love ‘em or leave ‘em?



I love dice. I love the way they feel in my hands, and the sounds they make as they click together when I shake them. I love the feeling of tossing them onto the table, and the moment of tension as I watch them land, waiting to see whether or not luck is on my side. Before I met Jonathan (a little over 3 years ago), though, I didn’t know what a D-6 was (a 6-sided die; I mean I knew what a 6-sided die was, but the word “D-6” was a foreign language to me), or a D-20, and I had no idea that a three-sided die was even possible.  

What I did know was that in the ancient world, people used bones, stones, and sticks in ways similar to the ways we use dice today. Sticks would be etched with markings that would come up, based on how you tossed them. Stones would be marked with dots, much like the pips of dice, and tossed and counted. So the idea of tossing tangible “things” to create randomness in a game goes way, way back - as far back as the historical record will allow us to look.



Dice, like any other game component (tangible or intangible), are a tool. Like any tool, there are things they’re good at doing and things they’re not good at doing. There’s a lecture I’ve given several times called “Meeples and Metaphysics – Practical philosophy as applied to game design.” In it, I talk about the metaphysical implications inherent in the physical properties of common randomizers (cards, dice, etc.) and their impact on the rules of your in-game universe.

“The metaphysical implications inherent in the physical properties of common randomizers and their impact on the rules of your in-game universe”

Wait… What?

This is my pompous way of saying that the physical nature of dice (or cards, or a spinner) – enforce specific rules for how fate, destiny, and predictability work in your game. By virtue of being shaped the way they are, dice are good are good at:

  • Quickly providing one random result from a pre-set list of possible results.

  • Being unpredictable: Past results do not predict future results. Just because you rolled three sixes in a row, you’re no more or less likely to roll another six.

  • Being consistent and static: The faces on a die don’t change; you’re stuck with what’s on it for better or worse.

  • Being “tokens”: They’re easy to hold, conceal, gather, and exchange.

  • Keeping their state: It’s relatively easy to keep a die on the face you rolled.

  • Adjusting their state (within their pre-set options): It’s relatively easy to change a die from one face to another,

If you need a randomizing element with the traits listed above, dice are a great pick. If one or more of the above traits is undesirable, it’s probably a bad idea to use dice. Do I love dice or hate ‘em? I’ll admit I have an inordinate fondness for dice – especially a translucent red d20 I’ve had since the early 80’s. {{photo here?}} It’s a lot of fun to roll a fistfull of d6es or d10s and tally damage dealt or successes. At the same time, I really, really hate the “X” on the blue battle die in Descent (Fantasy Flight’s co-op adventure quest game).

Huh... Why do I hate Descent’s “X”? This merits some examination. Here’s why I hate the “X.” I despise it because it guarantees a 16% chance of any attack failing. Beyond that, I also despise it because the die doesn’t have a corresponding “automatic success” face, and t feels unfair to have an automatic loss without “balancing” it with an automatic success. Desi and I are in several Descent campaigns. Every time I roll for combat, I do so dreading that “X.” It’s plausible that’s an intentional design decision; they may have wanted the specter of failure to loom over every combat action, regardless of how experienced your characters are. It’s also “fair” in that the monsters have the same odds of failure the players have. Even so, it’s too disruptive. You can do everything right, use great tactics and teamwork, and fail because of a bad roll. And… you don’t even get a booby prize. No extra XP or a re-roll token. All you get is a boot to the head. If you spent extra resources to make your action extra-effective, those are lost too.

Dice are good quick randomizers and the randomness of dice can be fun but if too much hangs in the balance, it’s easy to feel cheated. This is especially true if the rest of the game focuses on player decisions and control. Descent presents itself as a game that rewards:

  • Decisive action,

  • Pushing your team harder than you’re comfortable pushing them,

  • Strong tactical decisions, and

  • Judicious use of limited resources.

Having a random uncontrolled element that (functionally) causes players to waste their action 16% of the time directly contradicts the implicit and explicit themes in Descent. The combat dice in Descent are, in all other respects, used really well – they play to the strengths of dice. It’s that one little bit of uncompensated random auto-fail that exemplifies how dice can feel wrong when they clash with a game’s underpinnings. (In my “Meeples and Metaphysics” lecture, I call this “metaphysical shear.”)


Thoughts on the randomness of dice.

Desirée: Yes, it’s demoralizing to play my cards or use my character abilities to roll a handful of dice and, in that moment of tension, watch as the two D8s land on 1 and the D10 comes up with a measly 2, and I take damage (or whatever negative effect the game doles out). On the other side of it, it feels awesome when I toss the dice and they land on high numbers. And there’s a real rush when it comes just so close. Maybe it’s a game where I can use abilities or cards - or some other built-in game mechanic - to alter the dice. If I use it, and I win, I feel really, really good. I feel like a success. In reality, I’ve done nothing, in either situation, to earn the failure of the success. It just happened. It was pure luck or lack of luck.

Well, except in the situation where I might have used a character power or a card I’ve been holding onto and change my roll. Then, in that case, I may have applied some smart strategizing, some forethought. Sure. But even in those cases, didn’t I have to roll well enough for those abilities to matter? If I’m really pressed to land on one side or the other, I’d have to say that dice are not fair. They are random and can be rewarding as they can be punishing with no rhyme or reason. It’s just how they land.

They’re still fun to play with. I know that statement’s subjective, but I also know I am not alone in this feeling. I like the drama of dice. I’m not always in the mood for it, but some games handle it well. And when a game needs randomizers, I find dice to be a satisfying option.

Returning to the academic side of things, there's’ this great paper by Peter Shotwell that discusses dice and randomness. Board games are ancient, and in ancient times, games were seen as a reflection of the universe. Humans could play with the gods; they could play with death personified, gambling their souls. So in a sense, games have always been a reflection of our world, and maybe the way we respond to randomness in games has a lot to do with how we respond to randomness in the outside world.

There was a striking contrast between the way that dice and randomness were viewed in the Middle Ages by Muslims and Christians. For Muslims, if you played a game without chance, you were presuming to “outwit God’s fate on [your] boards.” Dice, on the other hand, were controlled not by you, but by God, so using dice was akin to “accepting ‘His will.’” For Middle Age Christians, it was the other way around. Dice invite chaos into the game, and wherever you find chaos, the devil’s not far. Medieval Buddhist monks were encouraged to play “mindless” games of chance, because games of skill distracted from their studies.

I find all of that stuff pretty interesting. And maybe I enjoy inviting chaos in. But maybe there are some situations where the chaos would be unwelcome. As a designer/developer, I think it’s appropriate to ask myself: are dice right in this game? Because dice take away some of the power from the player, and sometimes this can be very frustrating. But sometimes it can be a lot of fun.



Let’s start with the obvious: If you don’t want an unpredictable range of random results, you shouldn’t use a die.

Wait. Let’s take a look at that. After all, dice aren’t the only randomizers. There’s spinners, decks of cards, bags of beads, and people. They’re all common game randomizers too. Why do dice get a bad rap when these other elements get off scott-free? Let’s see if we can’t figure it out...

A single die is unpredictable: The only thing I can say with confidence about a standard d6 is that it won’t roll less than a 1 or more than a 6, and it won’t roll a fractional number. Other than that, all possible results are equally likely. But interesting things happen when you try to predict the sum when you roll a cluster of dice. Dice (much like people) become more predictable in clusters than individually. Let’s take a look at the distribution of results of the of 1d6 and the sums of 2d6, 3d6, and 4d6:


Graph of distribution of results of 1, 2, 3, and 4d6 summed.

Diagram from

As you can see, the odds of any result on one die are equal. A 1 is as likely as a 6. The odds of rolling a 7 on 2d6 are higher than any other result, and all other results become less likely as you move away from 7 (8 is more likely than 9, which is more likely than 10, and so on). The odds of rolling a 10 or 11 on 3d6 are higher than any other result, and all other results become less likely as you deviate from 10 and 11. Unlike 2d6, where the odds slope off evenly as you progress toward 2 and 12, the result odds are weighted toward the middle, creating a bell-shaped curve. There’s the same bell curve with 4d6, centered on 14 and more pronounced.

And here’s the thing:

  • A single die is chaotic – All faces are equally likely to resolve on any given roll.

  • A group of dice becomes predictable – to a certain degree – and becomes more predictable as you increase the number of dice in the group.

The sum of a cluster of dice averages out toward the median result.

The unsullied beauty of a die is that all results are equally likely. When you (as a designer) recruit a die as a randomizer, you create an impenetrable veil of chance. You say, “Beyond this point, you can not predict exactly what will happen.” Unlike a deck of cards or a bag of beads – where prior results remove options from the pool of possible future results – a die doesn’t care about its past. You’re not going to “run out” of sixes on a d6. You don’t have to “shuffle” a d20 to put numbers back on it.

Dice (and spinners) do a great job of paralleling the kind of chaotic unpredictability we experience in the real world. That’s great if your expectations (as a player) have been properly set, but it’s not always the right answer. Players don’t always want randomness. Randomness (without a way to influence it) removes agency from the players; it can make them feel like their decisions don’t matter. If I (as a player) do everything correctly, but a bad die roll that the rules forced me to make takes that away from me, I feel cheated.

So how do we address that?

  • Re-rolls: Give players a limited number of re-rolls. One bad roll stinks, but if you re-roll and it’s still a failure, you can tell yourself “Ok. I was destined to blow that roll.”

  • Make rolling optional: You can choose to roll to boost your effect, but in accepting the roll, you’re also accepting the possibility of failure. With great power comes great potential to botch.

  • Let overwhelming force overwhelm: Allow the game to have situations where the player can’t lose, even with a bad roll (ex: My target is 5 or better. I already have a 4, so I’ll succeed even if I roll a 1.). This lets players choose to deny fate its chance to screw things up at the cost of committing excessive resources.

  • Use something other than dice: Give players a hand of cards with the same distribution as the die (or dice) you want to use - for example, a set of six cards numbered 1 to 6. Let them choose which ones they’ll play when; they get to decide when to succeed and when to fail. Players know they’ll have to take their medicine eventually; hold your nose and swallow the bitter pill.

So why don’t spinners, cards, and bead bags have the same bad rep as dice? I think it’s a simple matter of exposure and context. Dice are super-common; it’s reasonable to assert that board games use them in high-stakes resolution situations more often than any other randomizer. As such, we recall moments when we really needed a good roll and the dice screwed us over more easily than the same kind of event with spinners, cards, or beads. (If you’re a Magic player, I’m sure you’ve been mana-hosed more than a few times. Magic players have a counter-example of “cards hate me” specifically because they’ve used cards as a randomizer more often than the typical board gamer.)

Any game randomizer (cards, dice, spinners, beads, etc.) can be used correctly or misused. Many designers don’t think too hard about the inherent differences between each one. Many designers use what’s convenient - and dice are very convenient. They’re easy to acquire; all major manufacturers have literal tons of dice in stock. Players already understand how to use them. They’re quick to use. Because of this, dice get used a lot in games. As a result, they also get misused a lot and get a bad rap: “Dice are too random.”


Games that we love that have dice.


Lately we’ve been playing a Descent campaign, since the app came out to allow a fully cooperative game. Nothing makes us grumble and groan that seeing the blue die come up with the X, meaning a total miss. Some of us have started referring to it as “the cursed die.” But we use the same “cursed die” to roll for the monsters, and when their roll lands on the X, we celebrate.

One of the first games Jonathan shared with me was Roll Through the Ages. I still enjoy that game. There’s a little bit of “press-your-luck” in that one.

I also love Arabian Tales, though you might argue that it’s not really a game. The dice are simply there to create a random experience. It’s so light and easy, and I’m big on storytelling games.

We play a lot of Mice and Mystics. Combat is entirely dice-based.

Castles of Burgundy is one of my all-time favorite games.



Roll Through The Ages (RTtA): In RTtA, each die represents a city you can get certain resources from, and you have to spend one food (one of the games’ resources) to maintain each die (city) you have. I like RTtA for so many reasons… The game clearly understands what the dice are and what they’re supposed to do. Everything else in the game is built around that. The game encourages you to build more dice to normalize your results, but expansion comes with a cost (increased food maintenance and increased chances of catastrophe). The game has a clever structure for transforming the resources you roll into technologies or storing them. There’s so much good design that embraces the chaotic nature of the dice and lets you choose whether to spend time and resources to rein it in or press your luck and spend them elsewhere.

Dark Moon (aka: Battlestar Galactica Express): There’s been a trend of using dice as a popular vehicle for taking larger/longer games and simplifying them into “Express” versions. Fans “expressified” Fantasy Flight’s popular Battlestar Galactica board game (BSGtbg), and this fan-hack was eventually published as Dark Moon. The main reason Dark Moon makes this list is it does what it’s supposed to do very well; it retains the fun parts of BSGTBG and strips out the annoying stuff, resulting in a game experience that’s shorter and a lot more fun. In BSGtbg, you try to figure out who’s a cylon by observing what bad cards get played and reverse-engineering who could have played them. In Dark Moon, everyone has the same dice, and it’s possible for any player to roll a sub-optimal roll. All rolls are made in secret, but (if I recall correctly) what you contribute is public. If the other players see you contribute insufficiently, you better have a good reason. And there’s only so many times, “I had a bad roll,” can get you out of trouble… There’s the clever part of Dark Moon: It embraces the randomness by using it as a mask for the saboteur.

Blank White Dice (BWD): To be fair, this is one of my designs, published by WizKids in 2016. Even though it’s mine, I’m still a super-fan of it. I love BWD because it breaks a core principle of dice. The game’s built entirely around letting players creatE and alter their dice with a set of icons established at the start of the game. You (and each other player) start play with two blank dice you’ve customized By writing on them with a grease pencil. During play you’ll alter your dice and mess with other players’ dice. You get the customization And expansion advantages of a deck-builder with the speed and ease of play of dice. Good stuff!



Of course. How could I leave out Blank White Dice? It was the centerpiece of our first year of dating. It was also my first modern board game, if we’re not counting Boggle and Cranium.


The original printing had dice, as well, and one of the negative responses I saw coming up in reviews was that the dice are too random and don’t offer enough choices. Do you think that adding an extra die for players also adds to the level of decision-making in The Island of Doctor Necreaux, 2nd edition?

Desirée: I loved playing the first edition because it was an adventure game. I remember feeling beat up by the game, and I think we lost or came very near it. Having two dice lets you guess an average, but it’s by no means a guarantee. It’s still possible to roll a 2 when you need a 5, and banking on the average being 7 doesn’t help you at all when those snake eyes are staring up at you. I like the versatility that the extra die adds. There are some skill combos that let you swap dice with other players. Things get more interesting, and I see more room for decisions to be made. But Jonathan’s the designer. I’d like to hear his thoughts on this.



Shifting from one die to two dice is a *huge* change to the nature of destiny and fate in TIODN2e. As I noted previously, the odds of any specific result on 1d6 are identical, but the odds of results on 2d6 are biased toward 7, and become increasingly unlikely the further you go form 7, with 2 and 12 being the least likely. Shifting from 1d6 to 2d6 empowered the players by normalizing the results. Making extreme results less likely literally reduced the amount of randomness in the game. It also had the effect of making one-point-modifiers (like a +1 to combat rolls) a lot more powerful for results further away from 7. Shifting a 2d6 result from 9 to 10 or 10 to 11 is a much larger probability shift than shifting a 1d6 result from any result to any other result.

Shifting to 2d6 was the right move. I’m a little embarrassed I didn’t design it that way from the start.