Making a Game: What’s in a Roll?

Before we jump into anything, I need to get this out of the way: we are not going to pick a dice rolling system in this article. I know, I said last time that we would be talking about dice, and we will be, but I am not ready to pick the system for rolling that Vikings and Valkyries will use quite yet. Given that dice rolls are the primary way players will interact with the system rules themselves, this is a pretty big decision, and not one that I am prepared to make so early in the development process. Every dice system is going to be best for specific kinds of results and gameplay systems. I will need to spend some time figuring out the systems in VnV before I can determine which dice will fit those systems. I want to design every system in the game so that each piece works even if I change the dice rolling system part way through, so today we will be laying some ground rules to help with that.

What’s in a Roll?

Before we talk too much about the reasons for skipping out on a dice mechanic, I want to take a moment to talk about what a dice mechanic is, and why we might want to use one. Most of you will think this is obvious, but digging into obvious questions can present unique mechanics and solutions. Do we really need dice in this game? What are they for? Dice are random number generators. They are tossed onto a table, and the number that faces up is the number we use. In tabletop RPGs, board games, and gambling dice are used to randomize results. Dice are used in gambling to provide the rush of uncertain victory and to help the casino win big. Dice are used in boardgames that demand risk management from players, so that even the best strategist might sometimes lose to some risky plays and unlucky rolls. This can keep board games fresh because working out a perfect strategy is impossible if a roll of the dice can tip a game one way or another. RPGs employ dice in order to achieve both of those goals, well not the casino one, and also to create uncertainty in actions. Uncertainty is important in an RPG because the players are also the audience. The people playing the characters, and this includes the GM, want to be surprised by the events as they come. Otherwise, where is the drama? All of this is covered in much greater detail in this article. In fact, if you ever want a good dive into specific mechanics in RPGs, the Angry GM is just about the best resource on the net in my opinion.

So we want to have dice rolling in this game in order to allow for uncertainty and drama, and also for fairness. The reason we are going to spend all this time making rules and systems is because at the end of the day we are making a game, and games need to be fair in order to be fun. Now, fair does not mean equal. If I am playing DnD, the Barbarian will be able to lift more weight than the Bard. Some mechanics need to be different, and that means that some mechanics will be unbalanced, but that is okay. Fair means that there are rules that all of the participants in the game must follow. Soccer has rules, video games have rules, and tabletop RPGs have rules. Rules in RPGs are often even more important because one of the players is in prime position to abuse her power: the GM. The GM is the adjudicator of the RPG, which means she has the power to make or break the game, and while she technically has the power to supersede the rules in most tabletop games, we want to design a game robust enough that she does not have to ignore our rules. Breaking rules in a game is tricky. The GM can very easily harm the trust between her and her players if she decides to do something that violates the rules. Those rules are there so that every player understands the working of the world. If that changes suddenly, all bets are off and the players will not be able to make intelligent decisions in the game world. Moreover, the players will not be able to trust that their decisions are being respected by their GM.

The die roll is how the players will understand their odds of success on actions, and also how they will see that their successes and failures are a product of both their decisions, and some randomness. Randomness is easy. The players can see the dice on the table. The numbers that come up when rolled are close enough to random, so that part is taken care of. Being able to see their odds is trickier depending on the system. D100 percentile systems are easy to understand, because they can be expressed as percentages. D20 rolls are a bit more complicated, but the player can tell that they are rolling a skill that they are good or bad at, and plan accordingly from those numbers. Players will also be able to track bonuses that they receive for certain actions, so that will allow them to strategize within the game world. We want dice rolling in the game not only to enforce fair play between the GM and the other players, but also to create a sense of danger and risk that the players can understand.

Variables in Game Design

When programming a video game, one of the most important tools to make use of is the variable. Variables can be used to assign a value to a word. The programmer has a part of the code where they decide what the variable is, and then they can use that variable in the code instead of a number or string of text. For example, if I want to give my player 100 hit points in a game, I can make a variable called “hp” that I set equal to 100. Now, whenever my code needs to look at my player’s hit points, I can type “hp” instead of 100. This is useful if I change my mind later. If, after much testing, I decide the player should have 90 hit points, I can simply go back to the place where I told “hp” to equal 100, and change that to 90. Done. If I typed out 100 manually every time, I would have to change every instance of the 100 to 90, and that is a lot of work.

Now, I am not a computer. I cannot create a set of mechanics with a variable and then simply replace that variable to change all the numbers Moreover, I am talking about the basic logic by which this game will function. I have tossed around the idea of a d20 system, a d100, d10, d6 dice pool, d10 dice pool, a system where dice directly correspond to stats and skills, and on and on. Each of these systems would represent a massive change to the game. So, I cannot simply design the game with the dice system as a variable, so why don’t I pick one now? The reason is that I plan to design the next handful of mechanics in a general state. I want to define what I want each mechanic to feel like, and then I can start assigning actual dice and numbers to them. If I know parrying should be an action that players can take, and that they should succeed roughly 70% of the time, I can set any dice system to get close to that value. This is as close to variables as I can get in the design of such core mechanics. I can create a framework and then plug mechanics in manually later. So that is the plan for the next couple of articles. I am going to design the mechanics of this game with no dice, and then return and apply dice to them.

You may be asking if it would be easier to design the dice mechanic first, and then the other mechanics. After all, that would allow me to design each mechanic in one pass instead of two. You would be right, but that returns to the problem of changing my mind. See, I started this project with a dice mechanic in mind. I wanted the characters to have three action dice each, and for the Valkyrie to have two plus another one for each Einherjar in the party. Players could commit dice to actions and reactions on their turn. The Valkyrie could, at any time, pass any number of her extra dice to a player. So, you might say, I need a dice pool system for this game. Well, not so fast. One of the issues I had with this system is that it mirrored a problem we had when designing Psychout. When we designed Psychout, I wanted actions to be more fluid than the standard turn based structure. So, we gave each player three actions per turn. These actions were represented by tokens. When the player took an action, they had to discard one token. I thought this would cause players to hold some tokens in reserve for defensive moves, but instead the optimal strategy turned out to be spending all of your action tokens on a single overwhelming attack. The best defense is a good offense after all, and being able to fire off three attacks on a turn was usually enough to take down most enemies in the game. We struggled with the system for a while, before abandoning it for a more standard structure, and the game was better for it.

The other issue we noticed with Psychout’s early action economy was that the tokens ended up being kind of boring. See, I expected that players would see the tokens and come up with interesting strategies about balancing their offense with defense. The problem was that the tokens did not translate well into that strategy. It did not feel like a mechanic where a character could sacrifice defensive ability for some extra damage or accuracy. It felt confusing, out of character, and slowed down player turns as each player tried to figure out if they should drop one, two, or three actions right away. Plus, it felt like a mini game stapled on top of the real game. Players were not deciding between diving for cover and shooting at an enemy: they were deciding between one, two, or three tokens, totally divorced from the fiction. This was a system that sounded cool in my head, but ended up being not very fun when implemented.

What is this Article For?

So, what is this article for? The next few articles will be spent detailing the specifics of each of the major mechanics of the game, but in this article, I want to lay some ground rules for what I want my dice rolls to be able to do. This might seem like a strange thing to do: make a bunch of rules about how I will go about making rules, but it is essential to making a cohesive gameplay experience. See, I can hem and haw about my action resolution mechanics all day long, but I will never be able to decide on a system without a coherent set of goals. As much as internet arguments would have you believe otherwise, every dice system has something going for it. Some people may not like a certain type of dice system overall, but if that system is the best for the goals I want to achieve, I will still want to use it. So I am not worried if a system is “good”. Instead I am worried if the system works for my goals. I cannot decide which system to use based on mythical objective quality. I need a set of guidelines to judge the system by.

The Valkyrie and Passing Dice

The Valkyrie must be able to pass dice to players in order to empower their rolls. I have played with the idea of the Valkyrie passing power tokens or something like that, but no. Dice are more elegant, easier to understand, and goddammit this is the mechanic that started this whole project, and I want to get it working. So, the dice mechanic has to be able to accommodate extra dice in a roll. And, I will add another piece to that, the extra dice have to empower the character’s roll, not just allow for a re-roll or extra chance. Getting spirit power from a Valkyrie should be a big deal. It should allow characters to perform superhuman feats, so whatever way we pass dice, it needs to actually make rolls better, and allow characters to perform actions they had no chance of succeeding before.

Binary Success and Failure… On Most Rolls

Look, I have not had the best time with dice systems that make liberal use of degrees of success. Degrees of success are best used to add complications to rolls, and I find this does not pay off enough depth for the complexity it adds. Edge of the Empire’s dice system is a great example of this. Not only can every roll lead to complex results, but as the stakes and skills involved in a roll increase in value, the likelyhood of unexpected results increases. Seriously, check out EotE if you every want to look at an elegantly designed system for creating chaos and Star Wars style movie magic in a game. Despite it being great at creating chaotic results, EotE is sloooooow. Sometimes, I just want my players to be able to roll a check to convince an NPC of something, or to spot a hidden foe, or balance across a rafter. I don’t need checks like that to produce a laundry list of secondary effects that only serve to bog us down in deciding what each and every extraneous symbol means. I respect that on simple rolls like that, I can just hand wave those symbols away with a simple added effect, but then why bother with the system on every roll if we acknowledge that it is cumbersome on most rolls?

The other issue I have, is that the system robs the GM of the power to determine the consequences of the players’ actions. If my PCs threaten and cajole a local ruler into giving them something, I would probably have the ruler try to get back at them later in the game. In EotE, I could do that, but what if my players roll a Triumph, and no Disadvantages? Is it unfair of me to make the ruler have it out for the players? What if they roll a Despair? That would make the ruler having it out for the players make sense, but do I need to have a Despair on the table for that result to be fair? What if I don’t want the ruler to have it out for the players, but they do roll a Despair? Then I have to figure out some other consequence, one that might not even be related to the actions the players just took. The game book certainly doesn’t offer any insight into those questions, and it slowed me down at the table every time. In DnD or Savage Worlds or Dogs in the Vineyard, the GM doesn’t have to worry about any of that. She can decide that the players have pissed off the local ruler and enact those consequences. Done.

Now, Savage Worlds is a game with degrees of success, but they have limited use. If a player rolls 4 more than the target number, he gets a success with a Raise. This acts as a critical hit on attacks, and can give him something extra on other rolls. Continued Raises mean nothing on most rolls. Only damage rolls allow for continued Raising. This limits the effects of Raises to rolls where the drama of a super high roll matters. DnD is a binary system on almost every roll. There is, however, one notable exception: the attack and damage rolls. Attack rolls have critical hits, and critical fumbles, and as much as players like to scream about natural 20s on skill checks, only the attack roll has that mechanic. Damage rolls are a separate roll made after the attack roll that determine how well the player hit his enemy. That mechanic is nowhere else in DnD. Every other roll is determined by a single d20 check. Why do DnD and Savage Worlds use graded checks in combat, you ask? Because combat is the primary focus of these two games.

I am not sure that I will use graded checks for combat in VnV, but I want to be able to distinguish a system or two with degrees of success if need be, so I will primarily focus on binary success and failure, but allow the chance to make more complex rolls if need be. Simple rolls just don’t need a bunch of added effects. We can give the GM guidance on how to create consequences for player actions. The dice just need to tell us if the players succeeded or failed most of the time.

Only One Dice Mechanic

Almost every tabletop game does this, with a couple of exceptions. DnD uses the d20 for every roll, only having a damage roll that is different. I guess DnD also technically has a different roll for determining starting attributes, but as that is a totally different system that is only used in character creation, I can let it slide. I want to see if I can avoid a secondary dice rolling mechanic like damage. I will probably have some variants of my core mechanic, but every check should be made with the same kind of dice.

The Dice Should Encourage Diversity of Strategy

This is not a DnD style game. While I may still use character classes, I expect that these characters will be more well rounded than a standard DnD hero. This means that the dice system cannot overly reward specialization. This does not mean characters will all be the same. I want characters to be able to specialize in weapon types, professions, magical runes etc. I don’t want that to mean that these characters cannot do anything outside of their wheelhouse. For example, have you ever tried punching someone in a high level DnD 3.5 campaign? It doesn’t go great unless you are playing a Monk. Every character is specialized to use specific weapon types, most of which are not punching, and every character is expected to be loaded down with so many magical weapons that their opponents will be able to easily shrug off a non magical, untrained fist. This is not entirely the d20’s fault, but its style of constant upwards progression means that non-specialized strategies fall off at high levels because the bonus numbers can get so big as characters increase in power.

The d20 system is not the only culprit of this. Some dice pool systems can have a weird arc where the deeper a character gets into a skill, the greater the result of putting more points into that skill. I would go so far as to say I would like to see the opposite of that arc: where putting points into a skill early on leads to rapid growth that tapers off as the character continues to specialize. That is a softer goal, but either way, I want to make sure that my vikings can throw a punch and expect to do something with it even if it is not their primary skill. Come on, they are vikings. If every campaign in this system doesn’t have at least one long house drunken brawl in it, I will have done something wrong.

There Must Be Enough Dice for Magic

I know I have not explained the magic system yet, and I promise I will once it crystallizes a bit more in my mind. For now, know that magic is performed by using the dice the Valkyrie passes to the other players. Each player will have a chart of runes that they can place the dice on in order to activate them and create spells. This means the Valkyrie will need to be able to pass out a decent amount of dice, or that magic dice need to persist even when the Valkyrie restores her own dice, however that will work. I would say the Valkyrie probably needs to be able to give three dice to a character for magic. Whether she will have anything left over to give to the other players is something I can work out later. Let’s say for now that, whatever dice system I go for, I want players to be able to place at least three dice on their rune charts to create a spell.

What I haven’t Included

You may have noticed that I have left out some goals that seem important. I haven’t said how fast I want the dice system to be, or how many dice I want the players to roll. Well, both of those questions can be answered the same way: the minimum amount for the maximum depth. Every tabletop RPG worth its salt strives to have the simplest, quickest dice resolution possible. That key word, possible, is what will determine how long it takes to read a roll in this system. I can’t sit here and say a roll should take no longer than 30 seconds without putting some rules to the test at the table. What if the rolls themselves are an interesting exchange of choices, or what if the roll takes longer but provides build depth that would not be possible without that time? So yes, I want my rolls to be simple, not take too long, and not take too many dice, but “too long” and “too many” will be determined by a combination of seeing the game in action, and gut feeling. I will make my rolls take as long as they need to to achieve the feel I am going for, and no longer. If I feel the rolls still take too long, it will be time to reassess the level of complexity I need for the depth I want to achieve. Basically, the dice system being fast is not really a goal in and of itself because I cannot measure it. I can see if the dice system encourages multiple strategies, or if it can be used to pass dice from Valkyrie to Einherjar, but I cannot pick an arbitrary dice roll time and hold the system to it. It wouldn’t be healthy for the game’s development. So, I am going to use these other goals, and make the simplest dice system that meets all of them.

That’s It?

That’s it! This article is just here to lay a foundation for my dice mechanic. I have a set of rules I need to follow, and now I have a document detailing each rule. Next time, I am going to dive right into combat and break down my design to see if I can’t get us the best dang combat system we can have without a dice mechanic.