Echelon, being derived primarily from the *Revised System Reference Document*, uses a d20-based task resolution system.

Pretty much all character-related actions that can pass or fail have the decision made with a single die roll:

d20 + modifiers >= Difficulty Class (DC)

This is about as simple as it gets, really. A consistent mechanic and easily remembered.

## Degrees of Success and Failure

Most of the base game runs pretty handily on a pass/fail mechanic. There a few places that consider degrees of success and failure, but not many. Off the top of my head,

- Critical hits (natural 20 — sometimes lower — confirmed by a second attack roll);
- Climb checks (fail by 5 or more points and you fall);
- Craft checks (fail by 5 or more points and you spoil half the materials and have to pay for them again in order to continue).

There are also some skills and checks where you can modify the DC or adjust your roll for better or worse effect. I do not consider these ‘degrees of success’ because it is still pass/fail, you have just accepted some greater risk for potentially greater result.

Using Climb and Craft checks as a base, you have two degrees of failure: no progress made, and bad effect (falling or ruining materials respectively), based on how much below the DC you rolled. This can be extended for greater degrees of failure (fail a Craft check by 10 and you have ruined the work entirely, and must start over), and inverted for degrees of success (every five points you roll above the DC gives some greater effect such as faster speed climbing or more progress made toward completing the item you are crafting).

This requires that you roll, determine success or failure, then subtract the DC (or subtract from the DC) to determine degree of failure.

There are times in the course of any night’s ~~drinking~~ playing that subtraction itself becomes difficult, but thankfully there is an easier mechanism.

## Degrees of Success by Examination

Instead of determining degree of success or failure by subtraction, look to the natural die roll. The lower the value on the die, the better the result (bigger success or smaller failure), and the higher the value on the die, the worse the result (smaller success or bigger failure).

I might divide the d20 natural values into the following ranges:

Roll |
Success |
Failure |

1..5 | Supreme Success | Normal Failure |

6..10 | Great Success | Improved Failure |

11..15 | Improved Success | Great Failure |

16..20 | Normal Success | Supreme Failure |

Or, being Echelon where I like four-point ranges (and skill training gives a +4 bonus):

Roll |
Success |
Failure |

1..4 | Supreme Success | Marginal Failure |

5..8 | Great Success | Normal Failure |

9..12 | Improved Success | Improved Failure |

13..16 | Normal Success | Great Failure |

17..20 | Marginal Success | Supreme Failure |

‘Marginal Success’ is one where you *almost completely* succeed, but not quite. You did achieve your goal, but there may be something not quite ideal. You might suffer a minor (-2) penalty on the next related check or suffer some minor inconvenience or loss (you did clear the pit, but you dropped something).

‘Marginal Failure’ means you *almost completely* failed, but there is some means of salvage. You might gain a minor (+2) bonus on your next related check (you missed your opponent but are in a better position) or enjoy some minor convenience or gain (you didn’t clear the pit, but you did hit the edge and may climb out, or you cleared the bit but landed poorly and take a bit of damage, or you fell in the bit but find the nice thing the last person who tried this dropped).

### Automatic Success and Failure

In many checks, 1 and 20 are special cases, where 20 always succeeds and 1 always fails. If you need a 20 to succeed, you will never succeed by more than 0. As such, it would be reasonable that this is the minimal success.

Roll |
Success |
Failure |

1 | — | Automatic, Marginal |

2-7 | Great Success | Normal Failure |

8-13 | Improved Success | Improved Failure |

14-19 | Normal success | Great Failure |

20 | Automatic, Marginal | — |

### Why this Works

The chance of rolling a 20 when you need 15+ is 5%, as is the chance of rolling 15. The chance of rolling 19-20 when you need 14+ is 10%, as it the chance of rolling 14-15. The chance of rolling 18-20 when you need 13+ is 15%, as is the chance of rolling 13-15. I won’t demonstrate a proof here, but the pattern does hold true for other values and ranges.

Degrees of success by examination provides the same chance of each degree of achievable success (or failure) as by finding the difference, as long as the target DC is no more than 20 points away from the modifiers available.

### Why This Won’t Work

People love the ‘natural 20 woohoo!’ feeling.

## Closing Comments

Basic task resolution is just that: basic. It is entirely serviceable, but many people like to see a little more differentiation in degrees of success.

To do that, the obvious mechanism is to compare the total check result against the Difficulty Class of the check. Easily explained, but I find it annoying in practice because common cases, obvious success and obvious failure, still require a lot of work to be done. Depending on the mental state (fatigue and inebriation) this could be more trouble than people are willing to do, and error prone.

Instead, going to an examination method allows the obvious success and failure cases to be handled very quickly using a standardized table. The only time it is necessary to work out the exactly check result is when it is a borderline case… in which case the additional effort determines whether it is a greater success or greater failure.

This does take away from the “20 is always best!” idea, and I admit I’m not really excited about that because I like rolling natural 20s as much as the next guy. I think in the end that if I want to include degrees of success in a convenient and functional way, I can live without “20 is always best!” pretty easily.

Pingback: Simplification and Complexity | Echelon d20 - An RPG framework based on the d20 system.

So would a “great” success/failure replace the idea of a critical success/failure? I guess it would be impractical to use natural-1 and natural-20 for the latter at the same time as your above approach.

I just threw out arbitrary ranges and descriptions. I don’t think criticals, as used by D&D 3.x, can really apply here.

Critical hits in D&D 3.x are typically first natural 20, followed by a confirmation roll (second attack check against the same AC). It means basically that 5% of all successful hits are criticals. That applies to all combatants, even those that have very little chance of hitting at all.

Moving to a model like this means the improved effects only happen if you are more skilled and removes the straight luck from it. I’m okay with something like “improved success gives an extra die of damage” or something. You don’t have to confirm, but you have to be good enough to hit your target on a low roll.

Actually, this suggests the grades above are too strict. normal success on 8-13 means that a “reasonably challenge” with 50% success will only get normal success… about 15% of the time. That can’t be right.

This clearly needs some tuning — the principle is sound, the application is a little whacked.

I just changed the table. The new one is simpler and I think it’ll work better. Success is success and you just get better from there. It might be worth softening failure a bit, with 1-5 being “barely failed”. This covers cases where you don’t achieve your actual goal but might be able to recover. Climb checks, you just can’t make progress this round (but don’t fall), Jump checks may mean you didn’t quite reach your goal but are close enough to catch the edge with your hands and pull yourself up, Craft checks mean you didn’t ake progress but didn’t ruin it, and so on. This is mostly a matter of the label, but a better label may make it easier to remember to treat this way.

Success could be the same. “Barely successful” still is successful, with normal effect, but it’s not particularly special.

I must be brain-dead today because I don’t understand your “simpler” system at all. Don’t take this as a big deal, but it might mean you need to rethink how you explain it.

I can see one minor objection, or maybe not an objection but an observation: you are assuming that the PCs are almost always facing on-level or near-level competition. Can’t my heroic fighter have fun carving his way through an army of wimps? “I succeed by 27 and kill three adjacent goblins in one blow”

One more example of success degree: Knowledge rolls, with each +5 giving “one more piece of useful knowledge.”

I was somewhat distracted by the election last night. I’ll look the post over later and see if I can explain better.

However, I’ll see if I can explain now.

Assuming the game doesn’t fall off the RNG (Random Number Generator), when checks are needed it should be necessary to roll somewhere between 1 and 20 on the d20. As long as you can’t “always fail” or “always succeed”, this’ll work.

I am basically inverting the interpretation of the roll. This is more easily seen for the failure case, but the same principle works for success.

In the normal “degrees of failure” model, if I need at least a 12 on the die (12+mods = DC) then rolling an 11 is “failed by one”. Rolling a 1 is “failed by 11”. If I know I need a natural 12 or better then this isn’t so bad, but there are a couple of complications. First, everyone is likely to have a different target natural roll. Second, we probably don’t know what that is yet, until we have a bit of practice against this task. Until I have figured out my target natural roll I have to (roll d20, add modifiers, subtract DC) each time until I figure it out… and as soon as something changes the modifiers or DC, I get to start over. Everyone trying to do the task gets to do the same thing — especially when the natural roll is toward the outside of the range, 1 and 20 are where you can expect to find the biggest degrees of failure and success, even when you often know easily they are a failure or success.

Now, considering the failure, where I rolled too low. I can only ever fail from 1..11 points. I can subtract my total roll from the DC to find the degree of failure, subtract my natural roll from 12 (since I know that’s what I need, eventually) to find out how much I failed by, or simply look at the natural roll, since that is in the range of 1..11.

If the natural roll required was 10 instead (DC two points lower or modifier two points higher), I can only fail by 1..9 points. I cannot fail by 10. The natural roll will again fall in the same range, so why not use it?

The same principle, but with different numbers, applies for success. If I need a natural 20 to hit, I can only ever succeed by 0. If I can hit on natural 14 or higher I can succeed by 0..6. I can do all the math (d20+mods-DC; nat roll – nat needed) or examine (16-20 is a standard success, 14-15 is a better success). Whether you do all the math each time or not, in this case you have a 25% chance of standard success, 10% chance of increased success, 15% chance of great failure (“failed by more than 10”), 25% of increased failure, 25% of normal failure.

Why do more work than you need to?

It seems to me that you’re wanting to compare roll > DC-skill-mods instead of roll+skill+mods > DC. If so, I have no objection. I suspect most people want to know “what do I roll to hit?” anyway.

It’s a bit more immediate that way, if you know what die roll you need ahead of time. In the case of asynchronous play, such as on a forum, it’s extremely valuable because you can describe your success/failure without having to wait for a round-trip through the GM.

… I had not thought to express it that way, but you have a very good point.

However, the modifiers can change from character to character, the DC doesn’t. I would think it easier to keep track of as “d20+mods >= DC”.

Which is disappointing, because “d20 >= DC – mods” is easier to explain, it just doesn’t look like it would work in the face of different modifier totals.

What about just giving the DC and having each player modify it himself? It often ends up being done just like that anyway, right? Saves the GM trying to remember everybody’s list of possible modifiers.