Wednesday, 5 August 2020


I don't think anyone reads these sorts of posts by I feel things need to be recorded for posterity. Here I talk about how ability checks and skills interact with one another in a classless OSR-based game.

You may want to take a look at this 11 year-old post which pops up in OSR circles from time-to-time:


I have a new red sharpie
So no messing around, let's cut to the quick: here's how we do it in PARIAH, which is available at  Drive Thru RPG and


So this is basic, but only roll when:

  • There's a consequence to failure.
  • There is an element of time pressure.
If it's not risky and the player isn't bothered about when the goal is achieved, it can be assumed that the action is successful given sufficient time.
  • Don't roll for knowledge: make a judgement based on the character's background etc. and either tell the players or let them find out information through questions.
  • Investigation & perception checks: don't bother, again provide information based on the actions of the characters as described by their players
  • Exception: sometimes it's fun to do listening rolls. You may have a different opinion regarding this.


  • High is always good.
Whatever the roll, the higher the number the better. Nothing against roll low/under mechanics, I just like consistency: one or the other, not both.


  • An attribute check will use the WHOLE STAT, not just the modifier.
  • There's a qualitative difference between a strength score of 9 & 12.


These two preconditions present us with a paradox given that:
  • Roll high systems usually rely on a 1d20 roll modified by a stat bonus to hit a target number (TN determined by task difficulty).
  • Whole stat systems are usually roll under meaning low numbers are better.


  • Add the whole stat to the roll to hit a target number.
  • Use a smaller die (D6, because D4s are evil).
  • Base difficulty on ability scores (3 = everyone can do this, 10 = about 50% of people can do this; 18 = fewer than 1 in 200 people can do this)


Don't worry, I have two solutions for you:
  1. Roll 1d6, imagine that's your bonus... now look at your stat... that's your "D20" roll! ADD THEM TOGETHER YOU CAN DO IT! I BELIEVE IN YOU!
  2. Alternatively, get your DM to do some maths. They can deduct your attribute from the difficulty, resulting in a target number for you to roll.


"My character, Huffer,  is going to scale that wall... with no ropes or safety equipment... Huffer isn't going to take their time and pick out a route... they're going to scramble up as fast as they can without rushing, if you catch my drift."
"I reckon there's quite a few handholds for you to grab hold of and the rocks aren't too loose... but it's a fair old climb, and most people would find that very challenging to do as you describe. You'd need to be pretty strong to do that without rolling, maybe strength 16?"

"Well Huffer has strength 12 so—"

"So you need 4 or more to accomplish this climb."

Huffer's play rolls 1d6... and scores 3. The GM rules that Huffer gets stuck about halfway up. Maybe a one would have meant falling from the same height...


You don't need to roll, the action is accomplished as you describe it.


This is borderline, the GM might make you roll but the target number is "2". They might not.


Yep, sorry, sometimes there are times when you just can't do things because you're too weak or unhealthy or slow.
Re-evaluate your plan: how can I make overcoming this obstacle easier? Is there another way?

In desperate times, the GM may allow you to spend  ability scores or hit dice to improve their chances of success, and to represent a surge of adrenaline to overcome a life-threatening obstacle 
  • leaping a ravine while being pursued by rabid dogs
  • lifting a boulder from the crushed body of one's loved one.
  • making a saving throw against a mind control effect impelling a character to harm themselves or those they care about. 


Characters are allowed to better at things than other characters. In a classless system this is how they get to develop specialities and build the identity of their avatar before it is ruthlessly eaten by an aberrant creature from The Beyond. Before that happens, they get to be marginally better at ropemaking or swimming, for example.

This is represented by using a larger die when making ability checks in specific circumstances.

A lot of old school players hate skill checks because they think it encourages people to think that those without that skill are unable to perform that action.

This just makes them a little better.

If Huffer were a skilled climber, they would roll a D8 rather than a D6. 

Die size progresses to D10 and eventually D12 for exceptional mastery over a specific area.

This raises the ceiling of achievement without adjusting the floor, and also provides a simple framework for character progression.


I've been doing this for a long time now and it seems to work okay, though I'm interested in what you think. Two things I should about PARIAH in its current form:
  • D20s are still used for combat and saving throws.
  • Ability score modifiers are still used for those rolls and for reaction rolls

The only thing stopping me from altering this is perhaps a thinly veiled desire for my game to still "feel" like D&D.

  • What if combat rolls were stat+d6 like any other ability score roll? Perhaps attack bonuses could advance in the same way as skills (D6, D8, D10, D12...)
  • What if your starting "attack die" was the same as your hit die? And the amount over the opponent's defence score is the damage inflicted?
  • Why do I have five different saving throws when in a typical session no-one makes any?


I want to draw attention to this adventure by Arnold Kemp:'s free and excellent but (and pertinent to this post) contains an update to Arnold's GLOG (and there are many other GLOGs out there to try so shop around!) which has a roll+stat mechanic that works a little differently to this one.

Also worth checking out this:

Again it is excellent and all the other superlatives but (pertinent to this post) author Luka Rejec has yet ANOTHER roll+whole stat mechanic in his SEACAT system.


  1. I have not been getting enough sleep lately so perhaps my brain is breaking and what I'm about to say is obviously incorrect and brings shame upon me for even asking, but doesn't this system significantly change the probability distributions of rolls compared to regular roll under or d20 + mod mechanics? Nothing wrong with that, but if true, seems worth noting. Also, careful moderation of the difficulty means the overall probabilities would remain more or less the same, but still, the distribution.

    So for instance, if I have a 16 STR and roll under, there is a 16/20 = 4/5 chance I succeed. If we add an extra circumstantial modifier so it's more difficult by 1, then there is a 15/20 = 3/4 chance to succeed.

    But for your system, let's say difficulty is 18, that's 16+1d6, where I have to roll 2 or greater, so 5/6 chance to succeed, but then let's try difficulty 19, so 3 or greater, or 4/6 = 2/3 chance to succeed. In other words, the affect of modifiers in your system is much greater and there is less fine grain control. Likewise, the maximum difficulty a task can be over a stat before a task becomes impossible in your system is 6, whereas in a roll under or d20+mod system it's 20.

    There's nothing wrong with this, but I believe it's important to acknowledge. Assuming I'm not missing anything obvious here.

    1. Ok my brain broke after all and I somehow did not integrate any of the stuff about changing the size of the die based on skills 0.o. That still makes the probabilities less intuitive, but I actually think systems like this, where skills affect not just the probability of success in a static sense, but actually change the distribution of the probability of success, is very cool. I've played around with this a bit as well in various ways, none of which I can remember off hand.

    2. >but doesn't this system significantly change the probability distributions of rolls compared to regular roll under or d20 + mod mechanics?


      >Likewise, the maximum difficulty a task can be over a stat before a task becomes impossible in your system is 6, whereas in a roll under or d20+mod system it's 20.

      This is the most important part of it, the idea being that this will perhaps drive players to come up with more inventive solutions to obstacles by placing more definite, hard limits in their way.

      In theory, anyway.

  2. I love the idea that you have here—thinking hard about some concrete goals (high is good, use the whole ability) and then designing mechanics around that. Very smart.

    1. Thanks. Probably should have also explained where these goals came from:

      1. Roll high is just about consistency (actually I do sort of allude to that).
      2. The whole stat is relevant as attributes can be depleted and have to be treated like a resource to be managed (similar to HP)