As players, we sometimes feel like we’re at the mercy of the dice–but as designers, dice work for us. They make what’s going on under a game’s hood explicit, and in doing so enable us to see and work with the often-obscure probabilities underlying the game experience. When used correctly, dice can let us tune a player’s experience to the Goldilocks standard: not too little of anything, not too much, just right.

The great merit of dice is that they give the designer direct access to the player’s chance of succeeding in doing something. A bullet has pretty good odds of harming a soldier, but only a very small chance of disabling a tank. Bullets fired by a modern rifle are more effective at both of those things than a weapon from the Napoleonic era. Using dice, a designer can reach into a game and set those percentages: if history demonstrates that the odds of a bullet stopping a tank are a little less than 3%, requiring players to roll a 12 on two dice will model historical events accurately. If playtesting then demonstrates that bullets need to stop a tank about 15% of the time or else tanks are too strong, requiring a roll of 7 solves the problem.

Managing the odds of success in this fashion does more than just let designers model armor penetration; it provides a way to establish the feel of a game. When something is more likely to happen, players will naturally trend toward strategies that favor doing it. Conversely, actions that are unlikely to work will be a minor part of the overall experience.

Take Warmachine as an example. In Warmachine attacks are made by rolling two dice, adding an attack stat, and trying to equal or exceed the target’s defense. An average Warmachine soldier has a ranged attack skill of 5, and a defense of 12. Thus, on average dice the soldier will hit her target (ranged attack skill 5 + roll of 7 = defense of 12).

Since the average roll is a hit, Warmachine skews toward offensive play. Players are aggressive because they know attacking is likely to be rewarded. The game as a whole ends up feeling very active; attacks are frequent, models are steadly removed from the table, and the game constantly progresses.

Yet, making aggression good on average wasn’t enough for Warmachine’s designers. They added a mechanism by which players could roll three dice to hit instead of two. With three dice even bad rolls are enough to make contact, which lends the game even more energy; attacking isn’t just favored, it’s much better than hanging back.

Of course, too much offense would be a problem–Goldilocks’ lesson is just as applicable here as it was to heat in soup–and the precision with which dice odds can be manipulated enables Warmachine’s designers to add just the right amount of defense back in. The (arguably) best defensive spell in the game adds 2 to a soldier’s defense. That’s enough to warrant going to the trouble of getting that third die, but not so much as to make hitting impossible.

Imagine what Warmachine would be like if the average defense was 13. Now players generally need an 8 to hit–or, looking at it conversely, the average roll misses. With the best defensive spell defenses push up to 15, which can only be hit with any reliability when using the third die. It’s hard to envision that game being a high-energy affair. More likely it would mimic trench warfare, with players waiting to attack until they had a dominant position.

Going beyond the overall feel of the game, dice can also be used for subtler applications. The average defense is 12, but Warmachine’s important leader figures often have defenses in the 15-16 range. As a result, they’re hard to hit. Players are thus incentivized to push their leaders forward and get them involved in the action, which focuses attention on these thematically important pieces. The designers have used dice math to support the narrative of the game.

Subtler still. Not every soldier can add that third die. The ones that can have a much better chance of hitting a defense of 15 or 16. Hence, the soldiers who can “boost” their to-hit rolls are well-suited to knocking out enemy leaders, while those who can’t are usually sent against line troops or relatively immobile heavy targets. By setting where soldiers’ attacks fall on the probability curve, Warmachine’s designers establish their tactical functions.

Subtler still. Rolling a handful of dice is fun. People tend to like picking up and throwing more of them; it’s exciting to see if a big pile of dice will spike to a huge total, or collapse stupendously. By giving the ability to “boost” to thematically important soldiers, the designers link those soldiers with the excitement of the big pile of dice. That encourages people to play them, further reinforcing the game’s narrative and intended theme.

With all of that said, Warmachine’s approach to dice isn’t appropriate to all games. A game about World War I trench warfare probably should favor defense over offense! The take-away point is that dice, correctly implemented with an understanding of the probabilities involved, enable designers to build and modify games with substantial precision.

It’s often hard to judge exactly what effect an element of a game’s design will have. The beauty of dice is that the effect is right there to see: the probability of success is now X. Rather than fearing the randomness of dice, use the macro-level predictability they offer to shape the game they’re in.