FGU used to have effect points so a complex task required x effect points. reach point you rolled over target gave you an effect point. As you can roll per time period it means you could work out how long complex tasks took.
I have run several systems that work on that principle.

My Amber skill system simply gives you more information the higher your score. There is a threshold to get something.
So the party are following some bandits. the bandits have hidden their tracks but the bandit leader has "tracking" 25 so you need more than 25 to find them. The guy with 26 will get a statement "it looks like the bandits have tried to hide their tracks but you can just make out a trail heading to the west" the guy with 100 tracking will get "there were 3 bandits mounted on horses they attempted to hid their tracks in a rather crude way one of them came over here to this bush and cut down some brush they tied to the back of the their mounts. From his stride the guy is about 6 2" and weighs 200 lbs." (skills have no top end and can go as high as you like but 50 is considered genius level)  
This to me is the simplest DoS system.

I'm inclined to think 1/20 of hit chance too small for crits in BRP. For fumbles, it's close enough (to my mind) to use 99-00 for skill up to 50, just 00 after that.

A second digit of 0 or 5 gives 1/5, the 0 1/10 (both rounded down, but if your hit chance is under 10% then maybe you should not crit). If you have skills over 100%, then you probably can handle that math; but Elric! seemed to take that as more like mere competence with combat skills.

Quote from: Will;815605Peterson: yes. But it means you have to list skill/20%/5% values for the several dozen skills you might have, or do that math each time you roll.

In Runequest we had a table so players could look that up.

In practice, what we usually did was the player said what they rolled and the GM (me) told them if it was a special or a critical. The values are simple arithmetic and when I ran RQ I naturally ended up remembering them. Requiring players in general to do the math or look up the range themselves can be a bit cumbersome and is prone to them making mistakes.

I find degrees of success obtuse from the perspective as a Gamemaster, as it builds an expectation of additional success or mitigated failure on the part of the players based on an arbitrary measure. Attempting to justify  the difference between 5 degrees versus 6 degrees of success on an Athletics Test to climb a wall is just silly, particularly when the Gamemaster possesses fiat to narratively describe the results to the players.

ZWEIHÄNDER Grim & Perilous RPG is built upon the principle that you either 1) succeed at a task, 2) admirably succeed, 3) fail at the task, or 4) critically fail; that's it. It requires less numeric justification on the part of the Gamemaster to frame in a binary way YES/NO towards a task, but grants added success or critical failure. It's strikes close to the popular house-ruled D&D model of success, failure, critical failure on a natural 1 and critical success on a natural 20.

Now, I will say that degrees of success are helpful for purposes of introducing a second layer of results (multiplier method). For instance, with the example above, the Gamemaster may state that the wall can be assailed - and rather easily - with an Athletics Test. However, the degree of success may influence how many yards the climber can move over a short span of time. So, those 5 or 6 degrees have more impact direct to the results, rather than dictating the exactitude of numeric distinction within the success/failure model.

I use the Harn method of determining 20%, a 0 or a 5 on a success is a Special Success and 10% is a Critical Success.  An 05 (probably a Special and Critical if you're reasonably skilled) is the "OH YEAH!" number, like a 66 in RoleMaster.  It's easy, no math (anyone can read a 0 or 5 or chop off the second digit), and makes for good fun.

As far as degrees of success go otherwise, I like a Shadowrun type system where compared successes directly determine result rather then a nebulous "what does +3 on a d20 over what I needed mean?"  

Now, as Clash said, if you rolled enough damage to kill someone and you rolled really high, then you didn't just kill them - you chopped their head off, cleaved them in two, speared them through the left eye, etc...  Knowing the exact degree of curbstompage is important.

Quote from: Phillip;815614Chaosium's RQ had a table (using different formulas between 1st and 2nd ed.).

You can let the second digit give rounded-off probabilities down to 1/10 - but I don't find another die toss much more effort. I prefer that to poring over dice to parse them (which the desigers of Godlike seemed to think quite lovely).

Let the dice do the math! I say.

Hi Phillip! My point was that people have no trouble with a damage roll, which is a perfectly legitimate degree of success mechanic that is extremely pervasive, but never think of it as such.

-clash

Quote from: Will;815605But it means you have to list skill/20%/5% values for the several dozen skills you might have, or do that math each time you roll.

As Peterson suggests, if special/crit is based on percentage of the base chance, it is mathematically equivalent and far easier to use another die rather than calculate thresholds.

If you want to keep 5% crit, 20% special, then you can roll a d20 along with the two percentile dice. d20=20 is a crit, d20 in 17-19 is a special.


As far as degrees of success in general, I like having a handful of possible outcomes available from a given roll (roughly 4-6) rather than binary. More than 6 possible outcomes gets too fiddly for my tastes, though. Four could be done as crit/success/fail/fumble, but could also be say fail/green/yellow/red like in Marvel Superheroes, or number of successes in a dice pool.

I love TSR's Marvel Super Heroes, and also had a lot of fun with the Action Results Table in Legendary Lives (the handbooks for which suggested various uses for the target/result columns).

Talislanta has a very simple d20 table, giving general categories and some specific combat factors but mostly leaving interpretation up to the gm.

Quote from: Phillip;815640I love TSR's Marvel Super Heroes, and also had a lot of fun with the Action Results Table in Legendary Lives (the handbooks for which suggested various uses for the target/result columns).

The key thing is the the system has to codify to some extent what the degree success map to. MSH is a good example of degree of success system done right. Fate also with degree of success rolling into the damage in combat and the benefits of "Spin" effects generally being well defined.

On the other hand the more subject "Yes but, No and" sort of degrees of success I find hard work.

Unknown Armies:
Roll %ile. You want to beat your opponent while still being under your skill, else it fails anyway. Then your roll is success amount -- roll is damage for guns, or add dice together for melee.

What I find elegant there is that it neatly rewards high skill and rolling high without any math beyond <> (and maybe addition.)

Then UA does some fussy stuff with double numbers and stuff, but the core is nicely basic.

Quote from: jhkim;815636As Peterson suggests, if special/crit is based on percentage of the base chance, it is mathematically equivalent and far easier to use another die rather than calculate thresholds.

:huhsign: That's not something I suggested. Maybe you meant CRKrueger, and even then I'm not sure that's what he said.

Adding more die, for an additional indicator of result, doesn't seem 'far easier' to me than calculated thresholds.

Quote from: BedrockBrendan;815564I am fine with them but these days prefer it when they are more intuitive rather than chart based. Also don't mind if the system allows for eye-balling it.

Eclipse Phase is an excellent example of this, IMO: All you have to remember is 30, 60, and double digits are the special values.

I'm also generally a fan of systems where margin of success influences things like the amount of damage dealt (or avoided). As Will says, there are plenty of examples where you can enjoy gradated levels of success. And I also like Numenera-like systems where the dice occasionally generate a "pop" of excitement or interest or the unusual.

OTOH, let's consider FFG's Star Wars game. Using an inconsistent system of canceling dice symbols it can generate:

• Success
• Success + Advantage
• Success + Advantage + Despair
• Success + Advantage + Triumph
• Success + Despair
• Success + Triumph
• Success + Threat
• Success + Threat + Despair
• Success + Threat + Triumph

And then a matching panoply of Failure results. In theory this is almost defensible, but in practice even the designers can't figure out how to use the system. (One simple example: If you're making a Stealth check, you'll reduce the time required for the check through successes; but if it's a Streetwise check, you'll use advantages.)

I'm particularly fond of how L5R's raise system works, which is sort of a degree of success type thing, but you have to decide how much you think you can manage beforehand.

Basic idea is, roll against base Target Number to succeed at the task in the most basic way OR voluntarily raise the Target Number to get added effects, but if you miss the new target number you completely fail, even if you would have succeeded without those effects.

It doesn't work too well on some things, like knowledge checks (I'm not sure why if you try to remember more details and fail, you suddenly can't remember basic knowledge either), but with combat rolls, or for writing a poem, or something like that, I think it works well.

Quote from: Shipyard Locked;815562I have to confess, though I've played systems that use "degree of success" mechanics* extensively in their core rules, I often forget about them.

* How much your roll exceeds or fails the target number has a specific, quantified mechanical or in-universe effect.

I like them, even prefer them, but these rules have to be elegant and easy to use. That's the hard part.

The closest I've come would be combat rolls in Rune Quest. Since you have your critical numbers set before the session, and they won't change, it was pretty easy to use.

In my current D&D 5e campaign, I'm winging DCs quite a bit by just having the player roll and gauging what happens based on the result (with a few quick variables based on their background, proficiency and the current context). Seems complex but it really isn't. If I felt that the odds were stacked in their favor for some reason and they still fail, I give them a minor success at cost.

I only throw out a concrete DC number if they specifically ask me for one. Then I still do the same thing (partial success at cost etc...). It's been working well lately.

I like rulings, and luckily for my players I'm not an asshole. I'm a fan of their characters but I sure do love some peril.