A simple skill resolution and a twist

[Christoffer Lernö]

I'm trying to think of some simple skill resolution for Ygg, and came up with this:

Prerequisites: Stats (or skills) have a number ranging from maybe 1-7 or something like that.

Skill resolution like this

1. GM assigns a difficulty to the task, between 1 and 12 (or more).

2. If difficulty is equal or lower than stat/skill: automatic success

3. If difficulty is higher, then you fail the task on a roll of <difference between difficulty and stat/skill> or lower on a dx

Example:

A has Move 3, B has move 5.

Difficulty to jump over the cliff is 5. B auto succeeds. A's stat is two under auto success, so A fails on 1-2 on a dx roll.

Notice how the die can be altered to accomodate different amounts of chance.

I would use a D4 usually which would cap the chance of succeeding at 3 above required. But for some instances a D6 or even a D8 might be appropriate. Do notice how the linear system works to the advantage here.

Looking at a target number of 7 for example we get

D4:
Skill 1-3: 0%
Skill 4: 25%
Skill 5: 50%
Skill 6: 75%
Skill 7+: 100%

D6:
Skill 1: 0%
Skill 2: ~17%
Skill 3: ~33%
Skill 4: 50%
Skill 5: 67%
Skill 6: 83%
Skill 7: 100%

Nothing special about that resolution other than I thought it seemed simple enough.

Now for the twist

I haven't seen this in any other game yet, but with all the games out there I'd be surprised if it was new.

Anyway, I was thinking of the old "stats vs skill" problem. The usual thing is adding stat+skill in some way or do more complex schemes. Now what about this:

You always get a stat roll, but is you have the skill you have the skill roll as well. Success on either gives you a success.

So if I have Move 5 and Jumping 7 and the target is 8 (in the system above, but the idea is applicable in any system), you get to roll for both your stat (25% of success) and your jumping (75% of success) for (of course) a chance of success amounting to 81.25%.

I think the advantages with this system should be obvious. There are disadvantages as well. But anyway, it neatly keeps the stats and skills independent. This is especially useful in the case where the stat might be changing as well as the skill.

Just a random thought for people to steal if they like it.

[Victor Gijsbers]

That twist could work, but there is an unstated presupposition there. Your skill and your stat must be completely independent; otherwise, you'd allow someone a double bonus from his skill: once in rolling for it, and once in getting his skill to a higher degree. This is possible, but it might be rather non-intuitive.

For instance, character A is weak (Strength 3, on a scale of 1 - 10), and character B is strong (Strength 7). Learning a skill should not depend on the relevant stat, so A and B should be able to learn, say, Boxing (which I assume to depend on Strength) equally well. Character A spends a lot of time on learning this skill, and he gets it at 5. Character B spend very little time, and gets it at 1.

Now imagine a Boxing match. You let both chararcters roll both skill and stat to find out how they're doing. Let's say the target number is 7, and we're using d4's. This means that character A has a 50% chance of success due to his Boxing skill, and a 0% chance of success due to his Strength-stat. Thus, he has a 50% overall chance.

Character B, on the other hand, has a 100% chance, since he automatically succeeds due to his stat - even though his Boxing skill is far inferior to that of A. I think this is a rather strange situation.


If you just added skill and stat, A and B would have the exact same chance of success, which in my honest opinion is better justified in the given situation. In your system it works best to focus on either a skill or a stat; if you have one high and one low, your chance for success is much better than if you've got both of them somewhere in between.

[Ferry Bazelmans]

This
might have some insights. It's pretty much the same.

Fer

[Christoffer Lernö]

First of all, of course skill and stat are independent. The whole system is built up so that they can be. That's the whole point of it. But maybe it needed clarification.

Secondly: Victor, you're assuming I'm using the same system for actively resisted rolls. That's incorrect. I think using a success system for resisted rolls are terminally flawed.

Resisted rolls could work in different ways.

Here's one version.

In your boxing match: A has str 3, Boxing 5; B has str 6; boxing 1. We use D4 for this test.

A rolls str+D4 and Boxing+D4. Checks the highest result.
B rolls str+D4 and Boxing+D4. Checks the highest result.
Highest result wins, equal is a standoff, neither hits.

Quickly calculating the odds here:
A result of 6: 3/16
A result of 7: 5/16
A result of 8: 1/4
A result of 9: 1/4

B result of 7/8/9/10: 1/4

Chance of B winning=~60%
Chance of A winning=~19%
Chance of standoff=~20%

You can tweak this in either direction if the results are not what you want. Especially increasing the size of the dice helps even out the odds.

(Sidenote: with both at boxing 0 and strength 5 and 6 respectively with the same system. We'd then have Str 5 win: 18.75%, Str 6 win: 62.5%, standoff 18.75%)

A simple and very useful tweak would be to weigh stat and skill differently. For example I was thinking of letting ever skill level represent between 2 and 3 stat points. So boxing 5 would tie with strength 10 or 15. On the other hand for more stat dependent things, like jumping across a chasm stat and skill might work on equal terms.

The advantage is that you separated the stat from the skill so you can do these manipulations without much problem. In some cases certain skill applications might be very stat dependent, whereas other applications doesn't depend on the stat at all. In those cases the GM can disallow the extra stat roll altogether.

[efindel]

A couple of thoughts on this:

One of my own favorite resolution methods is:

 Stat + Skill + (dX - dX)

This makes "stat + skill" be the point where the character has roughly a 50% chance of success.  Further, you can vary the size of the dice, as Pale Fire did with his system.  In this case, varying the size of the dice leaves the "50-50 mark" in the same place, but moves the upper and lower bounds of the roll.  

On the "twist" given, the Ironclaw game already does something similar, but allows for having multiple dice in one thing as well.  For example, a character could have a stat of 2d6 and a skill of 1d12.  The result of a roll would be the highest of those three dice.

The interesting part of such a system, to me, is that it allows skill or a stat to affect the low end markedly without affecting the high end.  For example, say that in System Y, a character has a Dexterity of 1d12, which is very good.  On a task of difficulty 4 he/she will succeed 75% of the time.  On a task of difficulty 9, he/she succeeds 25% of the time.

Now, if we give that character a d4 in an appropriate skill, the chance of succeeding at the difficulty 4 task becomes 13/16, or about 81%.  The task of succeeding at the difficulty 9 task, though, is still 25%, since the d4 roll is irrelevant to that.

For some sorts of skills, this makes more sense than the traditional route of adding -- one of my favorite examples is mathematics, where having learned how to do, for example, algebra doesn't necessarily increase one's chances of being able to solve a problem in trigonometry, calculus, or other higher math.  (I know... you could also handle this with a set of related skills.)

Another interesting bit is that Ironclaw uses the "all 1's means a botch" rule.  This means that even a little bit of skill, while it may not help you with achieving a difficult task, does make it much less likely that you'll botch it.  Having that first d4 for skill means you only have a quarter of the chance to botch you had without it.

--Travis

[Victor Gijsbers]

Secondly: Victor, you're assuming I'm using the same system for actively resisted rolls.

Not really. Maybe my example wasn't the best, but I'm not letting A and B fight against each other, but against a third person. To win, they have to match the number 7, simple as that. So actually, my example is a non-resisted roll. I should have made it clearer that I wasn't talking about a fight between A and B though.

My main problem with your system remains, as you have not adressed it: High stat + low skill and low stat + high skill are more effective than medium stat + medium skill. This is, imho, illogical. If a certain success depends on both stats and skill, why wouldn't these things 'reinforce' each other? It is exactly the advantage of adding stat and skill that it accounts for this reinforcing effect, in amore logical way than your system does.

Of course, if you say "I don't care about this", that's fine. It might lead to problems though if you're making a chargen-process that allows people to spend the same Currency on stats and skills.

[Christoffer Lernö]

Victor, I think the problem might lie with the non-extended dice rolls.

BUT you have to be careful to put this system where it's not supposed to go. The system is supposed to work for skill tests like "I want to jump over that cliff", "I want to remember who succeeded King Scyld after the Sabeatsan Wars".

A strong coupling gives a whole world of trouble.

For example I have move 4 (average). That cliff is mighty wide and I might just make it (7). But I also have jumping (average) at 4. Which gives me a pretty good chance. Your objection is that a low skill at jumping does not help (say having jumping at 3 would have no effect here). This can be helped by extending the dice range using the infamous "bonus die roll" ("a 4 gives you an extra die") technique.

However, I think the strong coupling of "always add skill plus stat" and resolve with that more often give ludicrous results. A well known example here is Cyberpunk where the skill rating might state that you are a world class shooter, but if your reflexes are average you're still worse in execution than the master athlete with slightly above average skills.

Usually, it's at low skill levels that stat helps the most, where skill is usually aquired to not having to rely on strength or reflexes alone.

But you are correct in the weakness you point out. I'm gonna think a little about how to handle that. But I do wish to emphasize that I don't think the pure stat+skill is any solution. It's even worse!

I think the ideal would be a system where the carry over would work a little like this:

Skill/Stat -> skill-equivalent
5/0 -> 5
0/5 -> 5
3/5 -> 6
5/3 -> 6
5/5 -> 7

[Victor Gijsbers]

You're all too right when you claim that adding stat and skill is far from ideal either. An easy solution is to make skills dependent on stats. You have to associate each skill with one or more stats, and make sure that people with high stats have an easier time learning the corresponding skills than those with low stats. Thus, in your example, someone with a high 'move'-stat could learn the skill 'jumping' far easier than someone with a low 'move'-stat. When someone has to make a jump, you simply let him roll on his jumping-score.

There are some pitfalls of course. You've got to think about questions like: Does everyone get a basic score in this skill according to his stat, or independent of his stat? Will a high stat make improving the skill easier? Will a high stat raise the maximum value a character can get the skill to? If you don't watch out, your system becomes far too difficult. For instance, in my Eternal Worlds I created a very eleborate way in which skills were based on stats - far too eleborate for easy use.

[Ron Edwards]

Hi Christoffer,

The game Shattered Dreams is the only published RPG I know that used a fully-independent attribute die and skill die. You could be successful on both (great hit), successful on the attribute die only (success, barely), successful on the skill die only (missed, barely), and failed on both (sucks).

It was astoundingly easy in play, and it always made sense. Their Currency for deriving one's values were a bit whacked and complex, but I was always impressed by the mechanic. It's one of the handful of systems that used attributes + skills successfully.

Best,
Ron

[Christoffer Lernö]

An easy solution is to make skills dependent on stats. You have to associate each skill with one or more stats, and make sure that people with high stats have an easier time learning the corresponding skills than those with low stats.

Nope, that solves about... well nothing at all really.

First up, there's the issue of a basic chance. If the character doesn't have "jump" does it mean he can't jump even though he outclasses all the other characters?

Second is the retroactive alteration of attributes. Say the "weightlifting" skill depends on strenght. My strength on a 3-18 is 11. Fine, I increase it at average pace. Suddenly I drink a potion of ogre strength. However my weightlifting skill stays the same. Shouldn't have an easier time lifting stuff tho'?

Although most games won't include skills like that it appears more or less apparent.

A third objection comes from the different dependencies on stats. Let's say I have the skill "athletics" (not too unusual). Then sometimes that skill roll should depend on strength, sometimes on dexterity and sometimes even on constitution (to use AD&D names).

Three biggies. You can eliminate some of these problems by redesigning the skill tests and minimize interdependence, but they cannot be fully eliminated. The "help increasing ability" method is the most flawed method around methinks.

Thanks for the comment Ron, that makes it sound like this method is worth investigating.

[Andrew Martin]

Second is the retroactive alteration of attributes. Say the "weightlifting" skill depends on strength. My strength on a 3-18 is 11. Fine, I increase it at average pace. Suddenly I drink a potion of ogre strength. However my weightlifting skill stays the same. Shouldn't I have an easier time lifting stuff tho'?

Not necessarily. I believe that weight lifting is a skill. There's lots of posters, TV adverts and radio ads over the years in New Zealand that exhort people to "lift with the knees and not with the back!" which seem to confirm my impression.

So in a game, for a PC to lift a great weight , I'd have the player roll Lifting skill (a skill in the Species broad skill) and Strength attribute. I'd have the conditions act as additional complications or concessions applying to either or both Lifting skill and Strength attribute. Also, concessions/complications generated by the Lifting skill would apply to the Strength attribute roll.

Drinking the potion of Ogre Strength would add several concessions to the Strength attribute roll, but not to Lifting skill.

[Christoffer Lernö]

Not necessarily. I believe that weight lifting is a skill. There's lots of posters, TV adverts and radio ads over the years in New Zealand that exhort people to "lift with the knees and not with the back!" which seem to confirm my impression.

Most definately. I was merely objecting to the solution to the stat/skill dilemma where the stat works as a modifier to improve the skill.

Basically in that case you'd much quicker improve your weight lifting skill if you're strong than if you'd be weak. (Actually I find in many cases the reverse tends to be true)

Your resolution seems to work Andrew, but then again it wasn't what I was arguing against in the first place ;)

[Andrew Martin]

Your resolution seems to work Andrew, but then again it wasn't what I was arguing against in the first place ;)

I don't worry too much about things that don't work. I just use/do things that do work. :)

[Bob McNamee]

you could use a different die type each for attribute roll, and skill roll... to tip the scales slightly in favor of "skill and experience" over raw talent.

enough "raw talent, youth" etc can still beat "old but experienced", but with a skew... skilled has more bearing... someone skilled at jumping will do better than a circus strongman, or a thief with lightning reflex agility, but either attribute could have a bearing.

Bob McNamee

[Mike Holmes]

I think you're dismissing Victor's suggestion a bit to hastily.

First up, there's the issue of a basic chance. If the character doesn't have "jump" does it mean he can't jump even though he outclasses all the other characters?

This is assuming one were to make the system badly. For example (and I'm not suggesting that this be used exactly, but just to illustrate), one could begin all skills at the level of the attribute. Increasing the skill simply moves it up from there. Problem solved.

Second is the retroactive alteration of attributes. Say the "weightlifting" skill depends on strength. My strength on a 3-18 is 11. Fine, I increase it at average pace. Suddenly I drink a potion of ogre strength. However my weightlifting skill stays the same. Shouldn't have an easier time lifting stuff tho'?

Again, assumes bad design. Just use something like the above system, and make the potion a direct addition to either stat of skill or both, whatever you like.

A third objection comes from the different dependencies on stats. Let's say I have the skill "athletics" (not too unusual). Then sometimes that skill roll should depend on strength, sometimes on dexterity and sometimes even on constitution (to use AD&D names).

Fine, roll whatever stat you like. Just develop off of a single stat. That way, you actually can get the affects of two stats in a roll. Which is cool. So if you say that athletics develops off of strength, using the system above, it would start at strength base, and go up from there. But in an actual roll, you could roll the skill and, say, dexterity.

Mike