Rationale: The Core Mechanic

This is the first Rationale post.  In these posts we’ll take the mechanics that I’m moving forward with and we’ll scrutinize the rationale behind them.

Today, we’re going to look at the Core Mechanic.  By the Core Mechanic I mean the in game mechanism for determining success, and those other pieces that are tightly bound to it.  For example, GURPS Core Mechanic is “roll 3d6 under skill”, D20’s Core Mechanic is “Roll 1d20 and add Skill”, and Fates Core Mechanic is “Roll 4dF and add Skill”.  In each of those cases I’m over simplifying, but those are at least the core of the core mechanic.

The Search for the Ideal Mechanic

Previously I had lamented the fact that the very property of dice pool mechanics that makes it so appealing was also the property that prevented them from “scaling”.  I faced the problem that linear mechanics and dice pool mechanics each had attractive properties that were mutually exclusive.  We can visually compare each one by using a difficulty matrix; each of the following charts is normalized against each other using the standard deviations from the dice mechanics.

In each of the following you should think of the color coding as representing an Objective target number, roughly representing the following:

Blue – Generally below the well defines mechanical line

Teal – Easy

Green – Challenging

Yellow – Difficult

Orange – Formidable

Red – Above the well defined mechanical line

Diceless

This represents a diceless system where ability is compared to difficulty, and if the ability is higher the Character succeeds.  Here the numbers are arbitrary, we count by 5 but could easily count by 1.  As you can see this “scales” perfectly.  If person A is better than person B, person A always wins.  Scaling in this sense can be thought of as the slope of the colors, and in this case the slope is 0, so we might say the scaling is “perfect”.  However, we don’t actually want perfect.  We want the better to substantially outclass the lesser, but not perfectly.

Dice Pools

Now let’s look at two dice pool systems.  I present two of them here so that you can see that the difference in how the pool is measured doesn’t substantially change the outcome; no matter what sort of dice pool you use you always get similar scaling.  While they do not scale, they are useful in that each extra dice provides a diminishing return from the previous dice.

New World of Darkness

Here we have a count successes system.  You roll a number of 10 sided dice indicated by your ability, and you count the number that exceed 7, and must have equal to or more than the Target Number.  For example, if you have 5 dice, and the target number is 3, you must have 3 of your 5 dice come up as 7 or more.

As we  can see, our “Expert” only finds “Easy” what someone abysmal at the task would find “Formidable”.  At our very best no task is ever trivial, and what was “Formidable” for the “Apprentice” is still “Challenging”.

If we changed this mechanic to use d6’s and measure above 5’s it would still not have a huge impact on this distribution.  It would still fundamentally be unscalable.

Shadowrun

Here’s Shadowrun.  Shadowrun uses a “Take the Highest” from an exploding d6 pool.  This highest dice is then compared to a target number.  This system has lots of things going for it – You can decide between height and width for dice beyond the target number, adding dice provides diminishing returns, it’s kind of fun…  However it actually scales less well than World of Darkness!  To get something that is Statistically “Difficult” for someone Poor at a Skill to be Easy, you must go all the way up the scale to “Super Human”!

Consider this… as a professional you would on a daily basis, without thought, perform tasks which the average person finds difficult.  Right?  If you’ve ever worked with a Junior in your profession (which I’ll assume you’re an expert in), you know simply grasping concepts can be daunting for this neophyte.  So there is something fundamentally wrong with a matrix like this, at least if you are looking for a generic mechanic that in any way matches reality.

Before we move on, I will note that a matrix like this is often represented in media.  You do see this sort of things in some Super Hero Comics, and in shows like Buffy: The Vampire Slayer – in these cases the narrative device allows anyone to overpower anyone else no matter how well established the better of them is.  So this sort of mechanic might apply really well to those sorts of games.

Linear Mechanics

I define Linear Mechanics as any sort of mechanic that involves Rolling Dice and add the total to a score – or equivalent (I’m looking at you GURPS).  These systems pretty much universally scale well.  They easily allow a game to establish who is better, and the Players can typically rely upon their characters to outperform their lessers on a given task.

Again, I’m going to look at two different systems to demonstrate that the specifics of the dice mechanic are less important than the fact that the mechanic is Linear.  Lots of thought goes into what size dice to use and how many to roll – and while these do have some effect on the probability distribution, after a certain point of analysis it is irrelevant.  The dice used changes the central tendency of the challenges, but it does not substantially effect the how it scales or whether the returns are diminishing.

Dungeons and Dragons

Dungeons and Dragons uses a d20 which is added to a skill and compared against a Difficulty Class.  The system is quite simple.  Because it uses a d20, the deviation on the dice is really high, a high deviation means that the many more points are required to achieve the next rank in mastery than using a different system, but nonetheless it is quite plain that given a roughly six point difference the better outclasses the lesser.  Here we can see what we want to see in a game, a character that is Skilled can consistently (“easy”) do what untrained would find “difficult”.  D&D 3.5 did suffer from the fact that the characters pretty much maxed out at what I call “Master” without the aid of equipment, but that’s completely fine.

However, like other linear mechanics D&D must put limiters on how much a character can focus in one thing.  If for example you had a “point buy” system in D&D where your could just buy anything, you would quickly find everyone trying to max out one thing, because there is no diminishing return.  This is why Dungeons and Dragons uses Levels; it’s not because levels are a simple measure of power (though that may be a side effect), it’s because levels allow one to define mechanism for exactly how a character is limited.

Fate

In Fate you roll four Fudge (or Fate) dice, which is a three sided dice with a +1, 0 and -1.  You add these dice together and add it to your score. Here we see very different measures of central tendency than with D20.  I placed the types of difficulty (Trivial, Easy, etc…) on these charts on Standard Deviations; since Fate uses a dice mechanic with a Standard Distribution (unlike D20’s flat distribution), it creates crisper diagonal lines.

This here is what I would describe as the correct slope for scaling in an RPG.  Each 2 steps up the Ladder puts your Target number for a given challenge level in a whole new class.  The Expert finds trivial that which the Untrained find Formidable.  It aligns with our feeling about what it means to master something.

Also note that Fate maxes out right around Master just like Dungeons and Dragons – this is a pretty common sense of minimum and maximum ability that one finds in games.  They max out where the “Master” readily defeats someone skilled, or almost to “laps” around the probability distribution twice.

Again we see that because a +1 in Fate does not offer Diminishing returns, Fate had to define another way of limiting the ability for a character to specialize.  In the case of Fate, rather than relying on Levels it relies on a “Skill Pyramid”, where the best of your skills will be 5, and you must have five skills at 1, etc…  Advancement works from there in columns, where it takes a substantial number of adventure to rank a skill up to 8 – probably about equivalent to ‘leveling to 20’ in Dungeons and Dragons.

Having Our Cake and Eating it Too

The big push pull between these two dice mechanics is how do we provide both Scaling and Diminishing Returns?  That is: How do we create a sense that the better significantly outclasses the lesser, but also retain the flexibility afforded by mechanics with diminishing returns?

The answer to that is to hybridize.

If you look carefully at the mechanics detailed above, you will notice that the mechanism that provides scaling is adding a fixed number to a roll of the dice – also note that you retain this property no matter how many dice you roll.

If you look carefully at the dice pool mechanics, it is the addition of extra dice that provide diminishing returns.

Most of you are familiar with games providing two output axis, games like One Roll Engine where you can measure success by Width and by Height.  There also exists games with two input Axis, for example Legend of the Five Rings where Attribute determine number of dice rolled, and skill determines how many your keep.

Well, here we are going to have our cake and eat it too.  Where going to do this by having two input axis – one axis will define how skilled you are, and the other axis will define how “advantaged” you are.

The Living Myth Dice Mechanic

Every action will be based upon an Ability.  This stat will define how good you are at performing a certain type of actions.  It might be things like “Melee” which is how good you are at close combat, or “Cop” which is how good you are at being a Cop, or whatever.

The Ability will have an escalating cost.  Specifically the cost of buying the next rank is equal to the value of the next rank.  For example buying an Ability up to 4 would cost (4+3+2+1 =) 10 points, and buying it up to 6 would cost (6+5+4+3+2+1 =) 21 points.  But here is the important part, nothing in game (short of things that effectively hand over points), and no other character options will boost this.  The only way to increase an ability is to spend more points to raise its rank.

So to make a test, each side will add the Ability (for non-character tasks the difficulty is equivalent to the ability) to a roll of d10’s.  Assuming each side is only rolling 1d10, the entire test can be described as PC Ability + 1d10 vs Opposition Ability + 1d10.  Which is pretty straightforward and provides the expected scaling on Axis 1.

Living Myth Axis 1 Mechanic

We can see here that we’re getting a Scaling somewhere between Dungeons and Dragons and Fate – Don’t be fooled by the color, the definition here is arbitrary – if I we to make Living Myth similar to Fate and D20 in terms of Maximum ability I would max out at 21.  However I want Living Myth to specify itself into higher ranges so I recoded the colors to reflect this.

Now we have a linear scaling mechanic with diminishing returns.  But we’re missing another piece – the accelerating cost prevents any other part of the system from interacting with it.  Consider a Perk (Advantage, Feat, whatever) that gave a +1; this would provide 22 points worth of value to a Master, and 3 points of value to an Apprentice.  What that means is that there is no effective way to price these things with a Linear Scaling Mechanic with Diminishing returns.

Which brings us to Axis 2.

Living Myth Axis 2 Mechanic

In Axis 2 we aren’t measuring how good a Character is at a skill, instead we are measuring how Advantaged that character is in this situation.  We could have used a Mechanic like Shadowruns and added to that, but I wanted a mechanic that was as Simple but still felt like adding so that the mindset doesn’t switch.  I could have just added dice and had you add, but that would not only involve adding a number of dice, it would have also been another form of linear addition.

So, the dice mechanic works as follow, you add any 10’s, plus this next highest dice.  If for example you rolled a [4][0][2], you would count that as 14, if you rolled [0][1][6][0], you would count that as 26.  This is not only really easy to do, it also means that with 10 dice it is possible for an Apprentice level character to beat a Mythic level character.

Also, it defines Advantage explicitly as “Increasing the risk for the opposition”.  See Advantage in this case is not so much about being better it is instead about increasing risk for the Opposition – and this jives with our conceptions about advantage.  If you are Flanking someone, you are not that much more likely to hit, but you are increasing the risk to that Character.  On the low end, the Easy column, you barely move the number with increased dice (with the exception of moving from 1 dice to 2).  It is the high end, the Formidable tasks that the extra dice really help.

The mechanic is still PC Ability + Nd10 vs Opposition Ability + Md10, but N and M measure each sides respective Advantage in the situation and the Ability of each side measures exactly what it says – Ability!

Deriving the Rest

From this Mechanical Kernel we can then derive the rest of the core mechanics.  Advantages come from Perks, Situational Modifiers, Bold Actions, and Emotionally Intense Roleplaying.

We also define a way of taking dice away called Disadvantage.  Disadvantages come from Impediments, Careful Planning, and Emotionally Indifferent Roleplaying.

Finally, it gives us a way to mechanically reflect increased (or decreased) intensity in Stakes.  We define Higher Stakes as both sides being Advantaged, and decreased Stakes from both sides being Disadvantaged.  Stakes can be set by the GM based on all sorts of factors, but for the most part it just reflects an change in Story Tension.

Describing Ability Level, Difficulties, Degree of Success, etc… these can all directly fall out of the mechanics.  The systems lends itself easily to counting by tens, which is a nice, since 10’s are very easy for people to count.

From the problem of “How do we have something that scales and also provide diminishing returns” I have come up with a system that as a side effect provided really interesting and dynamic Narrative Hooks for any sort of game.