image by Pexels via pixabay.com CC0

Every time there’s a fifth Saturday in the month (with the occasional exception), I run my Dr Who campaign. This uses a simple home-brew game system that I’ve been tinkering with for more than a decade. As the game proceeded (a good time was had by both GM and Player), I noticed the way that I was setting difficulty targets for the occasional skill roll required by the PC, and realized that if I described it, and extended the logic to fit other systems like D&D (3.x variety) and the Hero System – because the processes that I use in those cases is very similar – then it would make a good article. So that’s what’s on today’s menu.

The Lovecraft’s Legacy Skill Mechanics

The Lovecraft’s Legacy campaign uses a mechanic whereby the character gets so many dice from his stat, 1 die for each point in an Application Method, and 1 die for each point in a skill. You can assume that the average untrained person will have 2-4 dice to roll, that the typical NPC with a normal skill level has 4-6 dice, and that a typical PC will have 5 or more, possibly considerably more if it’s something that they are good at.

If the total rolled on these dice is greater than or equal to the target, the character succeeds in whatever he was trying to do. Since the average result on a d6 is 3.5, it’s easy to translate average rolls into “50-50 target levels”:

• 2 dice: 3.5 x 2 = 7
• 3 dice: 3.5 Î 3 = 10.5 rounds to 10
• 4 dice: 3.5 Î 4 = 14
• 5 dice: 3.5 Î 5 = 17.5 rounds to 17
• 6 dice: 3.5 Î 6 = 21
• 7 dice: 3.5 Î 7 = 24.5 rounds to 24
• 8 dice: 3.5 Î 8 = 28
• 9 dice: 3.5 Î 9 = 31.5 rounds to 31
• 10 dice: 3.5 Î 10 = 35

… and so on.

Foundation Question

The first half of foundation of any difficulty target is the simple question, “how hard is it to do [x]”, but after that comes any number of often-unwritten assumptions.

“…for a character with typical expertise” is one – alternatives include “…for a character with no training”, “…for an expert”, and “…for a standard character of ability y with a stat of z”.

“…under ideal conditions” is another, vs. “…under typical conditions”, vs. “…under current conditions”, vs. “…under adverse conditions”.

There are others as well, but they all provide context for the initial part of the question, the important part, defining the standards against which the question must be resolved.

Translation

Once you have a sense of the answer to the foundation question, the next step is to translate that into a base difficulty target. This is quite often done almost instinctively, using guidelines provided by the rules and the GM’s experience in running the game as factors. It’s usually unnecessary to actually articulate a specific answer to the foundation question; you can simply use the translation as your solution.

Modifiers

The next thing to do is to take into account any circumstantial modifiers. If the goal under the game mechanics is to roll higher than a target number, then + modifiers added to the die roll represent advantages to the character, while + modifiers added to the target represent disadvantages; if the goal is to roll less than a target number, then these are the other way around. And, of course, any negative modifiers work in the opposite way.

It’s most commonly the case that only one value is adjusted, either up or down, because that makes interpretation simpler, but the exact way of phrasing and applying modifiers in the game system determines the interpretation that attaches to a modifier of “plus-whatever”. In the Dr Who campaign-mechanics, the difficulty target is the value adjusted, so a “+” makes things harder for the character and a “-” makes things easier.

This is for two reasons:

1. As a process, this more accurately reflects the concept of a “Difficulty Modifier”, in my opinion.
2. By modifying a number that the GM doesn’t have to reveal, it means that the GM can also keep the modifier secret. This will soon be seen to be more important than it might seem right now.

The normal process is to consider each of the environmental and circumstantial factors that the GM considers relevant one at a time, compounding them until you get an aggregated total or run out of factors to take into consideration.

Total and Test

The final step is to apply the modifier in the manner proscribed by the game system and have the character make his roll, then interpret the results.

….At least in theory: The Metagame

That’s all well and good, but it isn’t quite the way that I do it in the real world. Instead, I look at a whole range of Metagame considerations. It’s critical that I do so, because when you have only one Player and one PC in the campaign, there is no safety net. When you have multiple characters, one can backstop another.

The Foundation Question, revisited

The interpretational context used for the foundation question is different, for a start, and relies on my understanding of the PC as a character within the game world and his capabilities. The foundation question I ask is “How difficult should it be for this character to do this, under the current circumstances” – actually, strike that – what I ask is “What chance of success do I want this character to have?”

Answering that question automatically takes into account a whole bunch of variables, many of them indefinable or completely unacknowledged by game mechanics. The consequences of failure to the character, how success or failure will impact the plotline, whether or not the character can get a second attempt, whether or not the character can somehow recover from an error before it’s too late, whether or not the time is right for “high drama”, whether or not there is an alternate strategy (that perhaps the player isn’t currently even aware of), whether or not success would be appropriate for the character being depicted, and yes, the difficulty of the challenge.

The interplay of multiple modifiers

Let’s say there are a number of different modifiers that apply to a given check, and that it’s extremely unusual for any one to be a bigger adjustment than What happens when you put these together?

Well, a lot of the time, one will cancel out another, in whole or in part. In effect, you can simulate the range of a single modifier as “d7-4”, and therefore, compound N of these “die rolls” to map out the normal consequences.

Graph of 1, 2, 3, 4, and 5 modifiers compounded, produced using anydice.com – I realize this is only barely legible; if you can’t see it clearly, you can open a larger version by clicking on the image.

This, to me, is a very illuminating graph. It says that most of the time, no matter how many modifiers you have, the net effect of ±1-to-3 will be ±1-to-4-point-something. In fact, it’s a fair statement to say that except in extreme cases, the whole assessment can be be reduced to a single modifier – “will the preponderance of modifiers favor or hinder the character, and to what extent?”

You can also state that if there are N modifiers to consider of equal scale, it’s unlikely (even in extreme circumstances) that the net modifier will be more than 4/3 of the maximum of any single one of those modifiers. Also, even if one of the modifiers has a bigger range, that range will be compromised – evened out – in most cases by other modifiers. In fact, 4/3 of the average will be the boundary value most – say, around 70% – of the time.

So, why not skip the hard work and simply assess where on that ±1-to-4/3-of-the-average-of-maximums range the overall situation belongs? Or even take the next step and decide how hard the check should be based on the plot needs of the moment? Not only do you have a lot less work to perform in determining what a given skill target should be, but the entire basis of the question has been changed from a mechanics-oriented question to a storytelling-oriented basis. The aim of the approach is to permit the PC to succeed – when he has to, and be able to fail when he doesn’t, but always for it to be close enough to be exciting – or trivially easy when the PC is in his element and in a position to show off superior expertise.

Application to Other Game Systems

The same basic approach can be applied to ANY game system that employs a skill check to determine success or failure. My co-GM and I use this basic approach all the time in the Adventurer’s Club campaign; the basic question is always, “What chance should this PC have to succeed at task [x]?” We then consult a copy of the character sheet to determine what the character’s base chance is, and from that, determine what the modifier to his skill check should be in order to achieve the desired outcome, i.e. likelihood of success.

This technique works for 3.x / Pathfinder as well, and I have no doubt that it would function equally-well regardless of the game system. The only difference is in how you approach the problem.

In 3.x, for example, you set a DC that the character has to beat. The number of dice is fixed (one d20), modifiers get applied to that roll, in the form of skill ranks and a stat modifier. Everything else is factored into the DC “target” by the GM. The theory is that this ranges from 5 for the trivially-easy up to 25 for the almost-impossible; but, as I showed in “How Hard Can It Be?” – Skill Checks under the microscope, A fourteenth level character can succeed at a DC 40 task more than 50% of the time. Under reasonable circumstances, he is more likely to succeed than to fail.

Clearly, the scales are out of whack. By the time the character reaches his upper teens in level, it will take a DC of 55 before the character has even a 50-50 chance of failure at any skill he has been improving consistently. In fact, to achieve chances of success that are reasonably accurate to the labels attached to them, you need to apply the corrective calculation, New = 10 + 1.2 x (Old – 10).

I revisited the issue, and the solutions, in Six Impossible Things Before Breakfast, looking at the question of the standard of expertise that should be considered when assessing how difficult any given task should be, and acknowledged along the way that Pathfinder does a better job of it than 3.x did, because for every skill, it provides examples of what applications a particular DC represents. The problem is that this leaves the adventure hostage to the game mechanics.

Simulation vs Abstraction… again

This is yet another case of the debate that’s been going on ever since OD&D was published – should you be a simulationist or an abstractor?

Simulationist-styles attempt to quantify every possible variable and take them all into account. Abstractionists toss realism into the waste-bin whenever they deem it necessary or desirable to do so.

The approach I take in the Dr Who campaign, and the Adventurer’s Club Campaign (pulp), and – for that matter – the Zenith-3 Campaign (superheros) – is relatively abstractionist, with varying degrees of pretense to simulationism. Most game systems are so written that the style you choose can be employed equally-well. I choose this approach in those campaigns because they are all cases in which the campaign’s genre sets the desired tone as being larger-than-life.

My fantasy campaigns have been more variable. The Fumanor campaigns and Rings Of Time campaigns were unashamedly High Fantasy, and so benefited from the larger-than-life approach; the Shards Of Divinity campaign was low fantasy most of the time, but high-fantasy in the right environment, a more unusual choice. The Zener Gate campaign is at a mid-point between the two extremes, leaning toward the larger-than-life later in the adventures while being more gritty and realistic in tone early in an adventure. This is all about adventure Pacing, a subject that I’ve discussed here on multiple occasions.

But, as a general rule, slavish adherence to game mechanics can and should be set aside in the interests of making the game more exciting, more fun, for everyone, in my opinion. Make it look harder than it really is – then let the players succeed anyway, if it improves the experience for everyone involved.