Monday, March 11, 2013

The Dice Post

Balance is a tricky issue.  If you start both players off with the exact same starting position and capabilities, it is likely true that your game rewards skill and experience.  Even in the best of games of this type, it results in a somewhat dry game.  Incredibly deep and fun games, but typically quite dry.  Sadly, many games also fall far short of their depth mark.  Depth in a purely deterministic game is hard to pull off.  It typically requires there to be some sort of emergent quality to the rules.  And emergence is difficult to purposely accomplish.

So most games take a different path, and introduce a variety of mechanics to liven up gameplay, but maintain balance.  Randomness, asymmetry, variable set ups, and so on.  So today I'm going to be taking a long hard look at dice.  How dice effect the gameplay, the depth of the dilemmas in the game, and the nature of the player interaction.

A Roll of the Dice.

Dice are likely to be one of the oldest of game mechanics.  From gambling, to wargames.  From roll and move, to role playing games.  Everywhere you look there are dice.  At it's core, dice rolls are used to differentiate one players experience from another.  If a game gives players the exact same starting position, with the exact same abilities, dice decide the game.  Without dice, if two players always did the exact same action, it would yield the exact same outcome, and the game would never end.  The outcome of the dice rolls forms a core of the narrative of the game.  Some games accomplish this wonderfully well.  And in some games its so transparent that once you see through the theme and the fluff, you realize the game is nothing but luck.  Even worse than that, a game of luck with no real decisions.

We Roll Dice?  That's it?

As a thought experiment, lets imagine a very simple game.  You are rolling a 20 sided dice.  The goal of the game is to roll the highest number.  That is it.  Obviously this is an awful game.  The players make no decisions, and it's entirely up to luck who has the highest number.  So how would one go about fixing this game?

Well, one could add a push your luck mechanic, giving the players 5 attempts to roll high, with the players keeping their most recent roll.  It gives the player a decision to make.  But not an incredibly interesting or diverse one.  It just becomes a game of understanding your odds of rolling higher than you already have, and how reckless you'd like to be.  So we need to make the decision more interesting.

If you made the game go 5 rounds, and allowed players 8 extra rolls over those rounds, it begins to get more interesting.  Suddenly players must now weigh future turns against their actions on the current turn.  Perhaps as an added twist make every unused extra roll worth a point at the end.

Is this sounding like a game you'd at least try yet?  Hopefully it is.  Any game involving dice needs to present the player with an interesting and non obvious dilemma.  It is the quality of that dilemma which will entirely determine the success of the dice component of your game.

The Dice Hate Me!

Backtracking to the beginning of our thought experiment, there is another problem.  The players win or lose through no fault or action of their own.  The game just takes them for a ride.  They are waiting to see if they win or lose, but not really participating in that win or loss.  If a large portion of the players participation in the game is them pretending they can will the dice to a certain outcome, you need a better game.

So how do you give players control over a fundamentally uncontrollable game mechanic?  You introduce re-rolls, die roll modifiers, different probability curves, or options that circumvent the die roll completely.  And you introduce many, many die rolls to the game.  As many as you can to help flatten the result curve out to the expected result curve.

Another Close Game?

So your game with dice has it all.  Re-rolls, modifiers, rolling a pair of six sided dice to even out the probabilities, and a default action that is about 25% as effective as the median die result.  And now every game turns out extremely similar, and the scores end extremely close to one another.  The game keeps players interested longer.  But at the end of the day, the game is still just a ride.  How do we fix this?

You can introduce a resource scarcity.  Players can't just acquire any ol' dice mechanic.  You need to make them scarce.  Make the players compete for them.  Make the divergent paths to influencing the dice mutually exclusive strategies.  And if a player settles for one, it largely excludes another player from following down the same path.

Stone Age does a good job of this.  For an investment of 1 dice off your current turn, you can earn either one free resource on each turn in the future, or a die roll modifier you can use once each turn.  For an investment of 2 dice off your current turn, you get an extra dice for the rest of the game.  Which can be used to flatten out the dice result curve if you throw a lot of dice at a specific task.  And once a player makes one of those investments, it prevents the other players from doing it.  At least for the current turn.

Situational Dice.

Another option is to have the die results mean different things in different context.  A game which takes this route is Farkle Party.  You are aiming for various combinations of numbers to score points.  1's and 5's are worth points right away, along with triples and straights.  You get to keep rolling so long as you keep scoring, with dice you've scored and dice you want to save held in reserve.  So depending on what you've already rolled, you can reason out how many sides on the remaining dice will score you additional points.

Another game which does this is one of my favorite games, Commands & Colors: Ancients.  In Ancients, dice can mean radically different things, depending on the context you've created for yourself.  You will be rolling between 1 to 7 dice, and on each dice what constitutes a hit can be between 1 side, to 4 sides of the dice.  So knowing when to choose your battles counts for a lot.  What truly raises this mechanic up in this particular game is that your opponent will constantly be trying to deny you the opportunity to roll under the best odds, and you will be doing the same to them.

What Do You Mean You Hate Peanut Butter and Chocolate?!

So, we have modifying the dice results in various ways, and we have contextual interpretation of the die result.  Can you ever mix them?  Off the top of my head I've yet to see a game do that.  I suspect because it would water down the effect of the die results too much.  Because in one scenario, you are adjusting your odds of success before the die roll.  In the other you are adjusting your success after the die roll.  If you could do both, it's getting awfully close to just picking the side you want and automatically winning.  The challenge is greatly removed from the game.

One Last Thing.

Are there other sins of dice games?  You know, aside from the list of short comings listed in our hypothetical, dead simple dice game.

Well, when you have been rolling well the entire game, and a single unlucky dice roll at the very end costs you the game.  I've heard of this happening numerous times in games like Battlestar Galactica, where a final dice roll at the end decides the fate of the entire game.  Now I haven't played Battlestar Galactica myself, so I can't speak for how true those stories are.  But if it's true, on the face of it it's a problem.

The game should not just logic OR the results of all the die rolls, and have a single failure mean you failed the whole game.  Seems like a bit much to put on the line for a single dice roll, doesn't it?  Clearly no single roll should reverse all the momentum the game has built up to this point.  A set back is fine, but to do a complete 180 from absolute victory to complete defeat is a bit much.

In Summary.

If you are going to include dice in your game, and do it well, you need to have a few things.  You need to give the player some decisions around influencing the die outcome somehow.  You need to include a lot of die rolls, a significant enough number to level out streaky rolls at least a little.  And please, please don't allow all the forward momentum a game has built up be reversed by a single die roll.


  1. I've playing Battlestar Galactica quite a bit, and while I can confirm that sometimes the game can end on a die roll this is not terribly common and is usually a result of the nature of the game. Allow me to explain: At the end of the game, either the Cylons win, or the Humans do. They can't both win or both lose. So, for example, if the ship is very nearly destroyed and a die roll can cause the last damage to occur (destroying the ship and causing the humans to lose) then this is likely the result of several rounds of well played Cylon turns, poorly played Human turns, or a combination of both. One die roll cannot destroy the entire ship, only several rolls over the course of many turns can do this, with a few opportunities for the humans to reverse the damage without a die roll.

    So, in reality, when the game comes down to a die roll it's not an abrupt stop, slamming a car into a concrete wall, it's the final extra push on the brakes that stops it at the light that we saw approaching and prepared for. The Cylons are pushing on the brakes and the Humans the gas.

  2. I second Greg in the Battlestar Galacitca is in no way a logical OR of the dice.

    Battlestar is a constant struggle for the humans, that last dice roll is synonymous with the final roll to hit in a roleplaying game you have one hit point left your enemy has one left, it's up to the dice to decide.

    If you had acted more tactically then maybe you wouldn't be in that situation at the end of the game.