Matters of Luck and Skill - Reframing the Skill Check

 

A common type of resolution mechanic in tabletop role playing games is the skill check - often resolved with comparing the value of a dice role plus a skill bonus to a target number ( E.g. difficulty class/DC, actual terminology may differ).

The value of the DC tells you the difficulty of the task ahead, and the larger the character's skill bonus, the higher the chance of rolling over the DC and succeeding at the task. This framing essentially positions the roll as determining how well the character does at the task, which can feel unsatisfying when a character who should be skilled at the task fails something relatively simple (you could of course argue that relatively simple tasks should not warrant a roll, but you have to draw a line somewhere if you want to have skill checks in your game, and I like skills).

One option is of course to forego the dice roll completely - select a difficulty class and compare it to a "skill level", if the skill is higher, it's a success, and if the the difficulty is higher, failure. Of course, if the GM is free to set the difficulty levels (according to some guidelines), and the player characters' skill levels are public, this actually means that success/failure is just decided by GM fiat. Not particularly satisfying.

I think it's worth to take a step back and ask ourselves what we actually want to model with the skill roll, what is its purpose? The way I see it, it has two "basic" purposes:

1. A simple way to mechanically express character skill - a character should be more successful in something they are skilled at, a bonus to a roll accomplishes this.
   
2. A way to inject randomness (rolling dice is fun), and resolve challenges "impartially", rather than by GM fiat (though of course the GM decides the DC)
   

So essentially, the dice roll is meant to model whichever amount of luck goes into deciding whether a character is successful. A task might have some baseline difficulty, which is then affected by random chance (for a similar but distinct idea, see Oracular Odds). It could be argued that there are two sources of randomness: reasons internal to the character (are they having a good day or a bad day), and environmental factors. Having to roll twice for every challenge creates a lot of overhead, and since it is just two random factors, it makes sense to compress it into a single roll.

I think that the reason why failed skill checks can often feel bad, is in part due to their presentation: the way many games describe it, it is easy to think of it as containing only the randomness internal to the character (and in principle, the GM could be consciously making decisions on the environmental factors so there is no external randomness). Especially with the notoriously swingy d20, it can feel like a small proficiency bonus of +3 hardly makes a difference.

One solution I have is to reframe the roll as explicitly describing luck - I am not sure if it would feel most satisfying to apply it to the DC or the players skill value, but I am leaning towards the former and will use it in the example below.

When a character attempts something challenging, compare their Skill Level to the Difficulty Level, if the Skill Level equal is higher, the character succeeds, otherwise they fail.

To determine the Difficulty Level, roll a Chance Modifier and add it to the Base Difficulty.

In my mind, the Chance Modifier should be symmetrical around zero, which can be done easily with regular dice (e.g. a d20 can be easily mapped to -1–  -10 and 1 – 10), or a die for the magnitude and a coin flip for the sign. 

A nice result of this is the difficulties having the same scale as the skill levels, a base difficulty level equal to your skill level is a challenging task that you will succeed at half the time.

Another result is that if we are content to abandon the reliance on a the d20, the swinginess can be decided on a case by case basis, depending on the impact of luck versus skill: If the outcome depends a lot on luck, roll a chance die of +/- 10, if luck plays a much smaller role, roll a chance die of +/- 3 (easily done with a d6), or even forego the chance roll if it does not play a role.

 Some examples of skill check situations:

A character wishes to forage in the wilderness, they have a Skill level of 4, the base difficulty is 3 with a chance modifier of +/- 3. Under normal conditions they would succeed, but they could be unfortunate that there is no good food nearby.

A character wishes to climb a tree, they have a skill level of 0 and the base difficulty is 2 with a chance modifier of +/- 4. They might be fortunate enough that the tree branches are nicely spaced and thick, so even an unskilled climber could do it.

A character wishes to scout the area around them from the top of a tree, they have a skill level of 2 and the base difficulty is 0 with a chance modifier of +/- 6. In principle, it should be easy to see what's around, but this is greatly affected by e.g. the weather, whether other trees block the line of sight etc.

I am not 100% happy with the system right now, as obtaining the positive/negative numbers from the dice feels a bit hacky without bespoke dice, even if it is a simple one to one mapping. However, I do think that this sort of reframing could help make failed skill checks feel less frustrating, explicitly calling out the outside factors affecting the chance of success.