PDA

View Full Version : Realistic skill system.



Alfryd
2008-01-15, 06:57 PM
Alright, he's a more clarified version:

A few notes:

All dice are 6-sided, and go from 0 to 5. (i.e, If you roll a 6, take 0.)
All divisions are rounded down (i.e, negative numbers increase in magnitude.)
Actor = PC
Extra = NPC
Director = GM

Skills, parent skills, and XP:
Every skill has a base level, together with circumstantial modifiers which may impair or improve performance in a given task- these sum to determine effective level. Skills gain XP (Experience Points) through practice, which then raises their base level.

Most skills will have a parent skill that mediates performance and development of the skill. A child skill's effective level is cumulative with it's parent's. (Thus, a parent may substitute it's own effective level for any child, and even descendant, skills'.)

1/2 XP you gain in a child skill grants matching XP to the parent skill.

Skill checks:
The difficulty of a given task is measured by Stress. To perform an appropriate skill check, roll two dice, and multiply the total by your effective level in the skill. If the result exceeds or equals the Stress, you succeed in the task. Otherwise, you fail.

Check modifiers come in 3 different forms:

Skill modifiers alter your effective skill during a given check. They represent subjective factors that affect the Actor only.
Roll modifiers raise or lower your dice total during a check. They represent factors beyond assessment or control, such as random events.
Stress modifiers raise or lower the Stress. They represent objective factors that would affect anyone attempting the task.

Your degree of success or failure will often hinge upon the check margin. Your check margin matches check result minus Stress (and is thus negative for failed checks.) If you succeed on a check, with check margin no greater than Stress, the skill gains 2 XP.

Support and antagonist skills:
Certain skills enhance eachother's performance under special conditions. These are known as support skills, and grant 1/5th their effective level as a skill bonus to checks made. 1/2 XP you gain in a given skill grants matching XP to all applicable support skills.

Conversely, certain skills actively inhibit eachother's performance and development, and are known as antagonist skills. Antagonist skills impose 1/2 their level as a skill penalty to checks made, and if the check grants XP, any antagonist skills lose 2 XP.

Raising, learning, and forgetting skills:
In order to learn a new skill, you must have at least 2 base levels in the parent skill. The child skill begins at level 0. In general, you need not learn a skill unless desired, but neglecting to learn a skill you make frequent use of (by substituting ancestor skills) may be deemed poor role-play by the Director. Moreover, once you take a skill, any XP you gain in that skill must be used to determine base level when the opportunity presents itself.

Every skill has a base XP cost that determines how hard the skill is to learn. Whenever the Actor Rests, you may determine a skill's base level in accordance with it's current XP, as follows:

{table=head]current XP (in base XP): | x1 or less | x2 | x3 | x5 | x8 | x13 | x21 | x35 | x55 | x90 | x150 | x250 | x400 | x600 | x1000 | x1600 | x2500 | x4000 | x6500 | x10000 or more
Skill level: | 0 | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 | 11 | 12 | 13 | 14 | 15 | 16 | 17 | 18 | 19
[/table]

Actors may not, in general, raise skill levels during Rest in this fashion more than once per day.

The only way to discard a skill permanently (aside from antagonist skills,) is through XP decay. To account for XP decay, simply deduct the current skill level in XP, (which can cause the skill to drop a level.) Skills with higher base XP costs suffer XP decay less frequently:

{table=head]retention level | Decay interval
0 - 9 | 2 days or Rest periods
10-19 | 1 week
20-29 | 1 month
30-39 | 3 months
40-49 | 1 year or equivalent
50-59 | several years
50+ | decades or more (Director discretion)[/table]

A skill's retention level = base XP cost + retention modifier. Thus, skills that were slow to learn are typically slow to forget. Certain skills may have retention modifiers which offset their base XP cost for purposes of determining decay interval, these will be noted as appropriate.


Common skills and aspects:

In general, a player may develop and employ any combination of skills and attributes he or she wishes, but every player should be at least roughly familiar with the following. Even if they never gain levels as such, almost every adventurer will have some occasion to make use of these faculties on occasion.


Life[] (XP 45)
Body[Life] (XP 10)
Brawn[Body] (XP 5) -brute force and muscle power
Vigour[Body] (XP 10) -fitness and endurance
Fortitude(type)[Body] (varies) -take punishment
(general) (XP 15) -support all Fortitude checks
(injury) (XP 10) -reduces physical damage
(fatigue) (XP 5) -exhaustion
(exposure)(type) (XP 5) -heat/ice/glare/salt, etc.
(disease)(type) (XP 5) -options by setting
(poison)(type) (XP 5) -options by setting
Mind[Life] (XP 25)
Artifice[Mind] (XP 20) -visual/spatial ability
Logic[Mind] (XP 25) -formal grasp of cause and effect
Reflex(type)[Mind] (varies) -react to change
(general) (XP 30) -supports all reflex skills
(visual) (XP 25) -sight
(tactile) (XP 25) -touch and poise
(aural) (XP 20) -hearing
(scent) (XP 20) -smell and taste
Soul[Life] (XP 40)
Willpower[Soul] (XP 35) -resist distraction
Muse[Soul] (XP 40) -creative ability
Relation(type)[Soul] (varies) -establish rapport
(general) (XP 45) -supports all relation skills
(genus)(type) (XP 40) -group of similar species, by setting
(culture)(type) (XP 35) -single sex/race/guild/nation etc.
(personal)(type) (XP 35) -one individual acquaintance


Original post below:
There's an idea- couple of 'em, actually- that's been rattling about in my head for a while, with regards to coming up with a purely skill-based RPG framework. I'm fairly well-acquainted with the d20 system, familiar with GURPS and have a passing acquaintance with a half-dozen other RPG systems, but I haven't been able to find a precise match with what I had in mind and could thus adapt to my purposes. I'd like to outline the notions here, explain the difficulties raised, possible solutions, and hopefully solicit some advice.


The first major feature I want to include is that XP gain should be, first and foremost, by skill. That is, XP gain is mediated directly through skill checks against appropriate DCs, and XP gain by a given skill is reserved for that skill's advancement. ( GURPS features this to some extent by reccomending that the GM permit Xhundred hours of study to go toward CP in the chosen areas, but there's no particular reason why it shouldn't be a basic game mechanic.)

The system I have in mind at the moment is that every successful skill check, (provided you roll more than 2, and don't beat twice the DC, to impose CR limits,) grants 2 XP, + 1 XP to support skills (similar to synergy.) In addition, every 2nd XP point you gain grants 1 XP to the skill's parent skill (which I will come to momentarily).

The reason why D&D bundles certain skills, BAB progression and saves, etc. into a single unwieldy package known as a 'class' is because they're supposed to be mutually supportive, or relate to the same basic tasks. GURPS, and many other RPGs, permit you to stash your hard-earned XP/CP/Karma/etc. more freely, and rely on players' self-interest to ensure sensible and realistic development. But to me, predicting future use of a skill and developing it accordingly is putting the cart before the horse- use triggers development, not vice versa.

So anyways. I was thinking that the solution to this problem would be to build up skills into an elaborate tree structure, stretching all the way from basic attributes like Brawn/Logic/Willpower and so forth up on down to trees for related martial maneuvers, weapon specialisations, favoured enemy subtypes, metamagic, or what have you.
When a given skill gains XP, half that amount will spill over to it's parent skill, which, in turn, can trigger a chain reaction of XP gain right down to the root skill(s) at the base of the tree. This way, you can get freeform specialisation, while ensuring that it's most efficient to develop related skills.
For example, the Brawn and Vigour skills both have the Body skill as their parents (which is also the parent skill for Fortitude, and a child skill of Life,) so adding 8 XP to Brawn grants 4 XP to Body and 2 XP to Life. Thus, improving Brawn indirectly improves your Fortitude (albeit less efficiently than applying Fortitude checks would do directly, and Life, which cannot be used directly.) Body is parent skill to Brawn, Vigour and Fortitude because it's very hard to improve your strength without increasing fitness and resilience, and vice versa. (It's a little similar to a generalisation of 'key attributes' to allow other skills to mediate eachother's performance. Like in reality.)

This may seem like a lot of paperwork, but in practice I've found it works out OK, since the XP system is reasonably simple and instantly gratifying. You make a check, you make two ticks next to your skill listing, 1 tick for each support skill, and 1 or more ticks for ancestor skills that trip the threshold for spillover. That is, it's an O(log n) operation.

What I've had difficulty working out is the precise system for performing skill checks, what 'child skills' should default to, and how skill advancment is performed.

For example, let's say a hypothetical (fairly advanced) warrior has the following skills, with parent skill in brackets:


life[] 2
body[life] 3
| brawn[body] 4
| vigour[body] 5
| fortitude(general)[body] 4
| fortitude(injury)[fortitude(general)] 7
|
mind[life] 3
artifice[mind] 6
| combat[artifice] 8
| melee[combat] 14
| | parry[melee] 18
| | fencing[parry] 22
| |
| sword and dagger[combat] 10
| rapier[sword and dagger] 15
|
knowledge(general)[mind] 5
knowledge(humans)[knowledge(general)] 10

Among many others. When wielding a particular weapon, knowledge of that weapon acts as a support skill to all combat checks, as does knowledge of your opponent(s). A support skill grants 1/5th it's levels as a skill bonus, so the above warrior fencing with a rapier against humans has an effective skill of:
22 + 3 + 2 = 27.

You see, because child skills allow a lot of specialisation relatively cheaply for large performance gains in a particular area, I'm worried that rolling n dice and adding the result can't really keep up with the performance gaps that emerge at higher levels, or will compel every player to specialise to an unhealthy degree. (A similar problem is that support skills really allow you to pile on the synergies, and I really like support skills. These aren't the wussy +1/+2 bonuses you get from Greater Weapon Specialisation, these scale indefinitely. Of course, that's exactly why WoTC won't allow that.)

Again, none fo the above is set in stone, but it should give you a rough idea of what I have in mind. What I've actually been using for skill checks is multiply skill level by 2d6-2. It grants the benefit that proportionate opposed skills have equal odds of success/failure at all levels. e.g, skill 20 vs. skill 40 has the same odds as skill 5 vs. skill 10. It also means I can add scaling synergies from parent/support skills without wreaking bloody havoc on game balance at higher levels.

There are, I find, two drawbacks- the math is a little more tedious, but worse, you have to find ways to curb those ever-scaling check margins. For instance- you perform a melee check against someone else's dodge result. The margin of success (all else equal) determines damage dealt, but at double/triple the opposed skill levels, that will, on average, be twice/thrice as big. Which means you need double/treble the hit points and/or double/treble the armour to compensate.

I don't even like hit points, and if you're rolling on anatomy tables to determine mortal blows or injuries based on damage threshold, and imposing check penalties based on hit point deficit... well, it all gets quite fiddly. In theory, if you take the same opponents and double all their skills, you should get the same odds of survival. That's not trivial.

The other hurdle is skill advancement. Clearly, especially valuable or hard-to-develop skills, (especially those with many well-used descendants to milk for XP,) should have a higher base XP cost. A skill should be easy to raise to the level of it's parent skill (or even begin matching it,) and becoming increasingly difficult thereafter, while a parent skill should be able to substitute for children in a pinch (much like skill defaults/development in GURPS. I was thinking that the fibionacci sequence could provide good XP cost multipliers after a skill outpaces it's parent, since it approximates exponential costs- the idea being to curb uneven skill growth and prevent overspecialisation.)
.

Fiery Diamond
2008-01-15, 07:17 PM
I just have to note that, while this may or may not be a good idea, compared to the way D&D works, it is mindbogglingly complicated. That's not usually a good thing. You lost me part of the way through. I'm sure if I reread it several times I would get it, but generally I prefer to be able to understand rules through the first reading, possibly even if I just skim.

-Fiery Diamond

Alfryd
2008-01-15, 07:32 PM
I just have to note that, while this may or may not be a good idea, compared to the way D&D works, it is mindbogglingly complicated.
It's possible that's just because I'm waffling on about comparisons with other RPGs and what was motivating me to make a given change. Could you point out where you got lost, and I'll try to streamline the intro?

To be honest, I don't think it's especially complex compared with D&D. You're still making skill checks (which is basically what BAB and saves represent,) against a given DC. Only you multiply, instead of add. If you succeed- but not too effortlessly- add 2 XP, and 1 XP to anything that gave you synergy. For every 2 XP you gain, give 1 XP to the parent skill. Level up skills, using the XP they gained, when you rest.

EDIT: Edited initial post to give more streamlined exposition.

Lakoda
2008-01-18, 10:50 AM
OK, first of all this is almost a system...add rules for movement and damage and you have a fledgling system on your hands.

This is a lot more complicated then d20, there is too much math (fractions are bad; taking half of stuff is annoying, taking a fifth of something is catastrophic if you want people to use the system.

The raising and lowering of skills section needs a table, the two lists are difficult to follow. I'd scrap most of the test there, as a table would be more telling; to me is seems like each skill levels "costs" an amount of XP, but it is hard to tell that or what that amount is if I'm correct. Also the x# notation for the base XP is confusing, it seems to imply a multiplier, which doesn't make sense to me.

The skill decay table is hard to follow as you have base how frequent a skill "delevels" based on the skills base XP level; the table goes from 0-50 while the base XP list a little higher up on the post goes from 1 to 10000.

Overall you have a good starting point, now simplify it. Find a way to use whole numbers as much as possible and use easy progressions (for example, base XP goes up at weird and inconsistent intervals).

Anyways, I am sorry if this sounds harsh, you are tackling a very hard thing. I have seem a lot of attempts at realistic skill systems (I've tried a few times myself) and they are usually just as (or more) complicated as yours here.

Have you seen FUDGE? For some reason this reminded me a little bit of FUDGE. linkage (http://www.fudgerpg.com/fudge.html)

Good luck!
~~Lakoda~~

Alfryd
2008-01-19, 10:50 AM
This is a lot more complicated then d20, there is too much math (fractions are bad; taking half of stuff is annoying, taking a fifth of something is catastrophic if you want people to use the system.
(I thought it was similar to the 'attribute bonus' you'd get in D&D, which is actually more complicated, since you have to subtract 10, and deal with negative modifiers.) But on reflection, support/antagonist skills need reworking anyways. (If support skills are too closely related and far up the skill tree, then having lots of mutual ancestor skills means that you're essentially adding 20% to the supported skill for free.)

I was hoping that a lot of the complexity would be relatively cheap, since it would only apply once or twice per session- such as rules for XP distribution, which can be handled (mostly) during Rest- or only change infrequently, such as bonuses from support skill. I don't know. It may need to test the mechanics more throroughly.

What I have been worried about is (A.) the notation needed to ensure complexity can be handled in a fairly transparent way, and (B.) individual check complexity. Multiplying check results can get tedious when dealing with high skill levels-
"9 x 27 = ...? You have 2 seonds to answer."

I've thought that a different system would be to allow players to roll and add bigger dice, (or multiply the rolls,) at higher skill levels. Like so:
{table=head]skill level | dice roll
0-6 | 2d6
7-8 | 2d8
9-10 | 2d10
11-12 | 2d6 x2
13-16 | 2d8 x2
17-20 | 2d10 x2
21-24 | 2d6 x4
25-32 | 2d8 x4
33-40 | 2d10 x4
41-48 | 2d6 x8
49-64 | 2d8 x8
65-80 | 2d10 x8
[/table]
...etc.
Or, possibly, just 2d6 x(skill level / 4). That way, you still get proportional results (so that skill 20 vs skill 10 is similar to skill 80 vs skill 40.)

The raising and lowering of skills section needs a table, the two lists are difficult to follow.
Will do.

The skill decay table is hard to follow as you have base how frequent a skill "delevels" based on the skills base XP level; the table goes from 0-50 while the base XP list a little higher up on the post goes from 1 to 10000...
(for example, base XP goes up at weird and inconsistent intervals).
The decay interval works off base XP (which is static,) not current XP.
If you're talking about level progression, that's based (roughly) off the (or a) Fibonacci series (http://en.wikipedia.org/wiki/Fibonacci_number), to approximate exponential growth. For base XP by skill, that's meant to reflect how easy learning the skill should be (it's far easier (or at least faster) to improve your fitness and muscle power than your logic or creative ability.) At the same time, of course, it's easier in the long term to raise high-XP skills to high levels, since they suffer much less frequent XP decay.

Also the x# notation for the base XP is confusing, it seems to imply a multiplier, which doesn't make sense to me.
No, that's correct. Raising a skill to base level 5 takes 8 times the base XP. But no, it doesn't eat XP, it just sets current level based on current XP. So, if you have 40 XP in Brawn, then your base Brawn is 5. I'll try to clarify the whole thing.


Have you seen FUDGE? For some reason this reminded me a little bit of FUDGE.
I can see the resemblance, actually, in that the system seems to have freeform skill trees and skill difficulties. I'll take a closer look in a while.

Anyways, I am sorry if this sounds harsh, you are tackling a very hard thing. I have seem a lot of attempts at realistic skill systems (I've tried a few times myself) and they are usually just as (or more) complicated as yours here.
Not a problem :)- harsh, I can work with. Lack of info, not so much.


EDIT: Tables added to initial post. Will work on other mechanics later.