Linnaeus’s Four Principles of Dice Game Design

In boardgames, game design, mechanics on October 18th, 2010 at 2:26 pm

Like most people in my generation of gamers, I love rolling dice; big handfuls of them when possible. Unfortunately, this clashes with a lot of other elements of my taste in games, and there are very few dice games that I love as much as I love rolling dice. While I don’t think I have all the answers for what makes a brilliant dice game, I do have some thoughts; principles, if you will.

I choose the word principles advisedly. Principles should be followed but, unlike laws or rules, they are provided with the expectation that they will be broken *when there is sufficient justification*. I’m not sure how much the designers of the recent spate of dice games (To Court the King, Kingsburg, Pickomino, Roll Through the Ages, &c.) considered these problems, but all of them, as far as I know, break one or more of these principles, and I don’t think they have sufficient compensation for it.

1. Downtime is the Enemy

The most important thing in keeping a dice game fun is a quick pace. Rolling dice is fun, but watching other people roll dice tends to be boring, especially after you understand the game’s tactics well. If you can manage an average time of fifteen seconds per player turn, you’re in luck. If you go over 45 seconds, consider whether your game is interesting enough to watch to justify the downtime with more than two or three players. If it is interesting to watch what the other players are doing there shouldn’t be a problem. One good solution is to make other players’ actions affect your position. Better yet, give players problems that they can ponder and strategies they can plan during the downtime.

2. No More Than One Roll Per Turn

A corollary of the last principle. The new wave of dice games suffer horribly from boring downtime, and the single greatest cause of that downtime is that many of the games have players roll the dice three or more times each turn. Typically, this takes the form of Roll, Keep, Roll Remainder, Keep More, Roll and Pray. The idea, it seems, is to mitigate the effect of luck by letting players reroll dice they’re unhappy with, and this has a larger effect the worse you roll. Unfortunately, it makes turns go on too long (sometimes almost a minute) and there is nothing worse in a dice game than watching another player consider which dice to keep and which to reroll, since there is nothing going on, and nothing to consider about your own position.

The height of this is To Court the King, which allows players to reroll over and over again, provided they keep at least one die from each roll. Since players have half a dozen dice or more toward the end of the game, and there are special powers to consider as well, turns can take a minute and a half or more and no one but the active player has anything to do.

At the other end of the spectrum is the classic Liar’s Dice, where players generally roll *less* than once per turn and everyone rolls at the same time. Turns normally take less than fifteen seconds, and when a player has to consider his bid carefully, the rest of the table can feel the tension.

3. Give Players a Chance to React to the Dice

While this principle leads to the evil of multiple dice rolls in a turn, it doesn’t have to. It is also the best way available to minimize the role of luck in a dice game without stripping out (most of) the fun. If players have meaningful ways of reacting to what they roll, they can make choices that minimize the impact of their bad rolls.

One course is to allow players to manipulate the dice after they have been rolled. To Court the King, aside from letting players roll over and over again on their turn, also allows its players to apply special powers they have earned to the results of their dice rolls (although some powers focus on adding dice to their pool).

In Macao, players choose two dice out of a pool of six, and they receive resources based on their choices. While it’s possible to end up with no attractive options (especially toward the end of the game), this choice does reduce the impact of luck.

In Liar’s Dice, the players keep the results of their rolls private, and then make bids. While their rolls should have some impact on what bids they make, players still have the option of bluffing (thus the game’s name). There are still good rolls and bad rolls in Liar’s Dice, but bidding and bluffing do a lot to mitigate their effect.

4. Low Rolls Should Not Suck, High Rolls Should Not Rule

Self-explanatory, I hope.

Kingsburg falls for this disturbing trap. If you roll a bunch of ones and twos you won’t get to choose as attractive options as you will if you roll fives and sixes. I believe the theory is that the Central Limit Theorem will even things out in the end, but there are a couple problems with this. The Central Limit Theorem is not a guarantee that things will even out, or even come close enough for skill to make up the idfference. Also, Kinsburg is a sort of engine game, so good rolls early are worth more than good rolls late. Stone Age also suffers from similar problems.

Macao undercuts this problem in two ways: all of the players work with the same dice rolls, choosing two dice from the same pool of results and players get to take the rewards of low rolls sooner than the results of high rolls, playing the investment advantage against high returns from the dice.
Again, bidding and bluffing help Liar’s Dice to avoid this problem, although bidding lower numbers does give the next bidder more flexibility than bidding higher numbers does, since they can raise by increasing the face value of the bid without increasing the number of the bid.

Roll Through the Ages largely sidesteps this issue by using custom dice based on symbols instead of cardinal numbers. The contrast between Urns and Disasters and Food or Workers and the flexible Food/Workers face are still a lesser version.

  1. This is SUCH good material. I am curious what you think of the dice mechanics in Happy Birthday, Robot! (If that’s not too narcissistic.) Here are my own thoughts:

    1. If there are three players, there no down time because one player’s rolls will call for his two neighbors to also contribute to the story. (There is some down time with four and five players, but some groups prefer this, since it gives them a breather between creative bursts.)

    2. Players definitely have the option of pushing their luck and rolling multiple times in a turn. Hopefully this is not boring, because those rolls will affect the player’s neighbors, too.

    3. HBR has less of this, except in that they can share coins with one another to mitigate the effects of bad rolls. There are some other mechanics that encourage coin-sharing, too, but they’re less relevant to this discussion.

    4. Again, hopefully this is mitigated much like Roll Through The Ages. The numbers aren’t relevant, except insofar as the number of dice you have also determine how many words you and your neighbors can write in the story.

    I see what you mean about these being “principles” and not “laws.” It’s easy to break them, a design just hopes he’s doing it for the right reasons.

  2. It’s funny. I’d never thought of HBR as a dice game, but now that you mention it…

    Definitely no downtime with 3, and I think watching what the other players write on their turns is interesting enough and has enough of an effect on them to keep the downtime with four, at the very least, from being onerous. Your point about recovery from the creative bursts is interesting as well.

    I’m not sure if getting fewer dice is less fun than getting more, and of course the coins have a huge effect on this as the game goes on.

    The push your luck mechanic has always made me scratch my head a bit, but I may not be valuing having more words to write as much as other players would. I’ll have to get back to you on it.

    • Using more words isn’t so much the goal (at least for players I’ve seen) as much as the flexibility of having more available.

      RE Your first note: I think if a game uses dice, there are lessons to be learned from “dice games” as a whole. I tend to be very minimal in my designs, so I try to extract as much information and emotion from existing game props as possible before adding any new props.

  3. Kingsburg actually does pretty good for low dice not sucking given that you can split them up. For instance a 2,4,6 is as good as a 15 total and not much worse than an 18 (only missing the soldier). As well the modifiers, such as extra die for weakest player, the 2 chips and the building that allows /- 1 all add great flexibility to the little numbers which can often make them more attractive than the big rolls. Don’t forget a 1 can get you a victory point.

  4. Maybe my analysis of the Kingsburg progression was too superficial, but it definitely struck me when I played a few times that the general trend was roll higher is better, even if looking at side-by-side spaces wasn’t so clear cut. I suspect 2, 4, 6 is better than 2, 2, 2 for instance, and 4, 5, 6 almost certainly is better than 2, 2, 2.

  5. Dice Town fares pretty well against these criteria (though it bends the second): simultaneous rolls, paying to keep more or less than one die, peering at opponents to judge which payoff they are going for, and valuable consolation prizes.

    It’s sounding like Alien Frontiers may have an issue with principle 1…

  6. I have to admit, I’m not that familiar with either of those games, although my Geekbuddies say Dicetown is pretty random, even for a modern dice game. I’ll take a closer look at Alien Forntiers when I get a chance.

  7. I strongly disagree with your first two principles.

    The two dice games I enjoy the most (Exxtra and Can’t Stop) do not adhere to those principles, but I also think they’re invalid in general.

    In my experience, people are far more likely to be engaged in another player’s turn when dice are involved and, as such, downtime is far less a problem than for other games/mechanisms. Furthermore, people seem to enjoy extreme events, even when they’re the result of random chance. Such extreme events happen more frequently when multiple dice rolls are allowed.

  8. Greg,

    I haven’t played Exxtra, but Can’t Stop annoys the heck out of me because it violates the first principle so badly, although I do understand how some people would find it fun. There is some mounting tension as a turn goes along, but, IMHO, it usually takes a while to build up enough to really be interesting, which means that it often doesn’t materialize more than two or three times a game. In many ways a safe stop is anticlimactic, too. I didn’t even enjoy it much as a kid (I got the plastic stop sign version when I was still in single digits and, annoyingly, gave it to cousins before I knew it was a minor collector’s item).

    I rate it a 4 on BGG, IIRC, but if you are more sensitive to the mounting tension, I could see it rating more highly.

  9. Have you tried out that Steve Jackson Zombie game? How does it fair against your principles of dice games? It looks fun, but maybe the subject matter is blinding me to its issues.

    • “Have you tried out that Steve Jackson Zombie game? How does it fair against your principles of dice games? It looks fun, but maybe the subject matter is blinding me to its issues.”

      is that the little game where you are a zombie and youre trying to eat different people (green yellow and red) with different difficulties? If so, it suffers from almost all of these problems. Its basically a worse version of Can’t Stop.

  10. This is excellent and I completely agree. The downtime problem is by far the worst in my opinion, because bad rolls sucking is at least tolerable if the game is very fast.

    I think that simultaneous rolling is a good solution to this, and rerolling abilities are fine if all players are doing all the rolling simultaneously. For example, to court the king would be fine if you all rolled and rerolled together.

    I would say that Kingsburg does a good job of negating the ‘high roll good low roll bad’ thing. In fact, often the player with a below average roll will come about doing better than a player with an above average roll, due to turn order advantages. However, there are two kinds of rolls in Kingsburg for which this isnt true: VERY good rolls (anything capable of getting the Queen), and VERY bad rolls (where you usually end up getting a thing less than others). Even then, various game mechanics tend to cause this player to catch up to others, and thus positions in kingsburg are generally not due to strength of die rolls, except if you roll very good or bad. So I dont think that Kingsburg actually suffers from any of these problems (at least not much). It does have a bad AP problem of analyzing where to place dice to block and not get blocked, when playing with more than 2 players however.

    I applied these rules to the Dice for the Galaxy prototype that I playtested with Tom Lehman at BGG con. (Its a dice game version of Race, created by Tom Lehman and Wei Wah).

    It was quite good and I believe it follows all four principles. Rolling/turns are simultaneous to avoid downtime. Players have the opportunity to modify a couple dice a turn to a side of their choice in order to react to the dice and make very good and very bad rolls less common. There is one roll per turn. Roll and modify simultaneously goes quite fast. I’m looking forward to it.

  11. Simultaneous rolling is a neat idea, especially if there is some interaction afterward. Kinda puts me in mind of a trading dice game…

    I also have ideas about another solution to the tCtK problem that I may bring up in a future post (nothing of earth-shattering brilliance, but enough to deserve its own discussion).

    Kingsburg does make a good faith effort to fix the luck of the draw problem, but honestly, the upper end of the track really does rock, and rolling under about 5 after the first season or so really does bite. It was those extremes I was referring to.

    Thanks for the info about Dice for the Galaxy. I was of mixed emotions about the prospect, since Tom is the designer of To Court the King. It’s been a few years since then, though, so I’m sure he’s developed as a designer since then. I’m definitely cautiously optimistic now.

  12. Hi Greg,

    as a co-designer of Kingsburg, the topic is really interesting. I would say that other people already shown the arguments why your criticism about the game “failing” point 4 has serious arguments (I agree that engine games are not so helped by the “central limit theory) but it may be a little excessive.

    More than punishing low rolls, Kingsburg doesn’t like a lot “same results on multiple dice” rolls, this is why 2,2,2 is usually a tough roll to handle, but even 5,5,5 is not usually a result that will make you happy.

    For a family game, we believed it was just right to leave some room to bad/good rolls to compensate for uneven skills at the playing table…

    I would be interested to know what do you think of the expansion rules, that were mostly designed to further balance the field and make “bad” rolls almost unrelevant to the final score.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: