Results and relations - The purpose of a gaming system

In my last post I put up a small thought experiment about character generation. The aim was to point at two different roles that a gaming system has.

My own answer to the question is that even if I make a purely random character, making the rolls myself gives me a stronger connection to and better understanding of the character. This implies one thing: The rules has two purposes

First of all, it helps me to decide what happens in the game. If my character fails a Jump skill check, he fails to make a jump he tries to make.

Second, it helps me to get a relation to the world and to my character. When I roll the dice, there is a symbolic aspect of it as well: My dice-rolling is my character jumping.

A well designed game approaches the rules from both these functions. It seems, however, that it's not as easy as it sounds. I have encountered a number of games that focus to much on either of these two.

One extreme is a game that focuses purely on a good statistical simulation. In this particular context I mean that the world says that something should happen, and the rules actually supports this. Someone who is extremely strong actually does more damage and lifts heavier weights. However, if this is the only aspect of the game, the rules and the character can be quite dry and fail to convey the sense of a strong character. A small example: In the game Heroes Unlimited, the "Strength"-attribute is called P.A (Physical Affinity). It's perfectly valid, and it doesn't take you long to learn what it means, but it's still P.A, it's not "Strength" or a similar straightforward english word. In other words, it's technically correct, but fails to relate directly what it means.

The other extreme could be a game that is very good on giving out an impression of someone who is strong. This game could use a dice pool system, having more dice to roll gives you a tactile sensation that the character is more powerful. However, if the statistical outcome of the rules doesn't support the description, the suspense of disbeliefe get's harder. One example for me would be the old Marvel RPG (The old version known as FASERIP). Since it used descriptive words and clear steps of their attributes, you quickly got a good sense of how strong or fast someone was supposed to be. If I read that a character had Incredible strength, I knew that he was about as strong as Spiderman or Iron Man.

The problem with this game was that, to some extent, the results of the rolls failed to support the descriptions of the attributes. In the real game, at least in my experience, my character with Incredible strength and Amazing Agility couldn't do what Spiderman could do in the comics.

This to me is an example of where the rules support the sense and impression of the character, but fails to live up to it's "promises".

In earlier posts, I've looked at these parts as Layers of a gaming system, where the core is about the statistics and the outer layers are about the interface. After looking at some different games through this layer system I had to abandon that idea, at least partly. This is because in some cases, the core statistics and core mechanic is closely related to the "Human Interface"-layer, and to make a distinct separation between what parts are Interface and what parts are Engine is basically impossible.

Slight interface change in Mutants and Masterminds

In the Green Ronin game Mutants and Masterminds, you buy your stats and modifiers with points. This includes your Attack Bonus, which is bought independently of your stats.

Each level of Attack Bonus costs 2 points, and encompases both Melee and Ranged combat.

Later on, when you are buying feats, it's also possible to buy a particular feat that lets you add 1 to either melee och ranged combat. This feat can be bought multiple times, adding one to either combat type per level. Each feat costs one point.

From a user standpoint, this means that if I, for instance, want to build an archer I might reason that an archer needs to have a good Attack bonus, since he is a good shooter. So, I spend points on getting a high general attack bonus.

What I should have done instead is get a low attack bonus and then increase the ranged attack bonus with the feat. This way of thinking, however, isn't clear in the book.

Consider this instead:

Attack bonus is bought for Melee and Ranged attacks separately. Each level costs 1 point. The feat that lets you specialize is removed entirely (since it's no longer needed).

The effect of this gamewise is exactly the same as the original rules, but it's much easier for me as a player (especially as an inexperienced player) to think about how to build my character. Using the example above, I it feels natural to get a slightly lower melee attack bonus and a higher ranged attack bonus.

This changed doesn't change the behaviour or the implementation of the system, but it does change my perception of it, making this a change at the User Interface/ Presentation level that I would say makes the game easier.

Layers and D20

This is an attempt to describe the different layers using the D20 System. The reason I chose this is that I can use the information in the SRD, which is open to use. Other possibile systems to analyse in this way would be Fate or D6.

Apart from the core rule, each game system has several individal parts, which all can be analysed in this way. For instance, the magic system in the D20 System can be considered a system in itself, and therefore should be analyzed as a separate part.

The System

As mentioned in the previous post, all actions are measured against how difficult the task is. The solving mechanism is random and linear. This means that extreme results are fairly common (the average roll is just as common as an extreme high or an extreme low).

This means, for instance, that adverse or positive circumstances will affect all participants in a contest the same. If the result-curve is bell-shaped, then a skill increase will have different statistical effects depending on the skill level of the character.

The system measures degree of success. Even though it's not always used, it's possible to calculate not even if someone succeeds or fails, but also how well they succeed (or fail).

Each task has a set level that says how difficult it is. In contests between two characters, this level is dynamic and decided by the opponent. Once again, since all results are linear, adverse or positive circumstances affect all participants equal.

The Engine

This is where we look at the implementations of the relations and behavior described above.

A classic way to get a linear random result is using a single die. We can chose to make a high roll good or a low roll good. This won't affect the statistical outcome, but there is a difference in usability and understanding of the system. Saying that a high roll is good is most often more intuitive.

A good skill gives a better roll. There are different ways to implement this:

• Addition: Skill + roll = result
• Multiplication: Skill * roll = result
• Random addition: SkillRoll+roll = result
• and many others

This choice could be said to be made at the system level, since they affect the basic behaviour of the game. There are two linear methods, addition or multiplication. The most easy and straighforward of this is a straight addition.

Since higher is said to be better, we say that if the result >= Difficulty, the task is a success.

The Game

This is where the rules meet the user. As of now there is basically one important choice to make, and that is what kind of die to use. The above rules are usable no matter what kind of dice we are using. No matter if we are using 1D6 or 1D100, the above rules and descriptions still apply.

The choice of die affects a couple of things though. Ease of use (lower numbers are easier and faster to add up) and granularity (a span of 1-100 gives more detail than 1-6) are the two more important. In this particular case, a D20 is used. This means that each skill-level gives a 5% chance of success. It's a span that gives a fair amount of fine tuning, as well as keeping the number relatively low.

Notes

This was a first attempt att analysing a system with these three levels. As said before, this modell is a work in progress, and it will be more fine tuned and defined as it is used more. This may also mean that the descriptions above could change.

RPG System layers

Just as software systems has different layers (algorithms, implementations, GUI etc) RPG systems could be described in the same way. One rough setup is to describe a system in three layers:

1. The System The relations between and the behavior of artefacts/ entities within the system. Defines the behaviour of the game engine.
   
• Example: All actions are measured against how difficult the task is, and solved with a random element. This difficulty could be either static or dynamic. Extremes should be just as common as regular rolls, so we want a linear curve.
  
2. The Engine The implementation of the game system. This is where we decide on which form an equation in the rules should be written (as an example).
   
• Getting a high result is considered better. We wanted a linear spread, so we want just one die. So, rolling a die and trying to get a high number is where we want to go. Dynamic contests are opposed by a similar mechanic, static contests have a fixed number to beat. The final basic rule is written down as
  
• Roll a die and add how good you are what you are trying to do. If you roll equal to or higher a certain number, you succeed
3. The Game The way these implementations are prestented to the player. Also small details that make the system more intuitive and usable. For instance: Should damage taken be deducted from a certain amount or added up from zero. In the former case you don't want to get to zero points left, in the latter you don't want to reach a certain level.
   
• "How good you are" is called a Skill, or in case of combat: Attack Bonus. Rolling and adding percentiles (1-100) can be time-consuming, but to few steps make the game to "rough" and not nuanced enough. A 20-sided die is chosen since it gives a fair amount of fine-tuning without giving to high numbers to add.
• Damage for each weapon is described in table XX. Damage is added up for each hit and when damage
  

Statement: To make a successful game, all these levels has to be considered.

The above names are just descriptive, and should not be taken as definitions. Also, the borders between those layers are fuzzy at best, and this model is not to be taken literally, but more as a guideline to point out that things happen at different levels within a system.

For instance, a roll-under-system and a roll-over-system could both said to be rooted in the same System (using my own nomenclature), but they use different Engines, that have different psychological and (to some extent) mechanical consequenses. The statistical behaviour of the two, however, remains the same.

Powered by ScribeFire.