Help me design a better non-combat skills system for Cold Iron

[ffilz]

I've never been very happy with non-combat skills in Cold Iron. The original designer has only ever shared much detail about one sort of non-combat skill - Riding...

Let's start with some basic stuff. Cold Iron resolution uses the cumulative standard normal distribution rendered in chart form along with dice as a randomizer. So the cumulative normal distribution shows the cumulative probability of landing somewhere on the curve or lower. The steps are modifiers called a Chance Adjustment (or CA). A Chance Adjustment less than zero happens 50% of the time. The way a roll is made is you roll digits of a decimal number between 0 and 1, normally two at a time using D100. So a 50 is 0.50. To get a non-negative CA you need to roll a 0.50 or better, 0.56 scores a +1 while a 0.44 scores a -1. The math behind a CA is that a +20/6 is one standard deviation above average.

Here is a Google Sheets rendering of the Chance Adjustment Chart (at least the non-negative half):



For combat, basically you take an attack rating, add a Chance Adjustment and compare to a defense rating.

Attack ratings are generally Fighter Level + 1/2 Skill Level + STR modifier + DEX modifier
Dodge is Fighter Level / 2 + Skill Level + DEX modifier
Weapon Parry is Fighter Level / 2 + Skill Level + DEX modifier + MIN(STR modifier, Skill Level)
Shield Parry is Fighter Level / 2 + Skill Level + DEX modifier + Shield Size (1 to 5 for most characters) + MIN(STR modifier, Skill Level * 2 + 4)

Defenses are penalized for size (and shields have a size requirement that means that for large creatures, the largest shield size modified by size modifier is always +5). You can also never use more than a 6 more STR modifier in attack than the defender gets to use.

Margin of success factors in for combat in that if you net 7 or more than the defender you do double damage (i.e. approximately one standard deviation above). For each 2 more you get another multiple, but armor has a Critical Protection factor that reduces your margin of success when checking for triple or better damage.

Now I don't think non-combat skills need quite that level of complexity. I'm inclined to have skill ratings be 2 * Skill Level (no class level) + 1 or 2 attribute modifiers.

I started off PCs in Cold Iron Samurai Adventures with Expertise Level 3 which gave them enough skill points for 5 level 3 skills and 5 level 1 skills. Skills cost (L * (L + 1) / 2) skill points (or L points to go from L-1 to L) and Expertise Level gives you 5 + 5*((L * (L + 1)) / 2) skill points. I expect most starting characters to take Expertise Level 3 though if one REALLY wanted to focus on that at the expense of fighting or magic, Expertise Level 4 could be had. I don't expect characters to advance as far in Expertise Level as Fighting Level, Magic Level, or Cleric Level.

But then the question is what should margins of success look like and what do static "target numbers" look like to make things reasonable.

Social skills should probably mostly be PC rating + CA vs NPC rating. A simple success should be enough to get something basic, with higher margin of success being necessary for more extreme goals.

Crafting skills probably mostly wouldn't require a roll, and rather would qualify the character for employment though opportunities to craft while on adventure might require a roll.

Athletic skills should as much as possible require a single roll unless the challenge is truly an extended challenge.

But beyond those kinds of ideas, I'm pretty stumped...

 

[xanther]

If you like the combat approach, would use it as well for all else. Why not? You may want to pre-calculate a lot of things so it goes faster.

Could adapt the concepts of combat to other skills, double damage may become half time to do something, "parry" may be a counter argument or some security feature on a lock.

The static concept would be fairly simple, in that you could take a certain skill level then decide how often do you want it to succeed at doing tasks of a certain difficulty.

I'd first set out a spectrum of task difficulties. Simple, Basic, Difficult, Hard, Expert, etc. (using whatever adjectives you like) then determine the chance you want a character of "basic" level to have against a Basic task. It may also be where it is not so much pass/fail overall but pass in a given period of time that is relevant.

It can come down to setting two calibration points, and then working on the scaling in between.

In the final rules you can ditch the adjectival descriptions if you want and just go with numerical, like a Level 3 lock instead of Basic Lock.

Personally I'd set at least a 50% chance for a level 1 or 2 PC to succeed at a Basic task, and then try to keep that 50% mark for harder tasks when get more experienced. So a veteran PC level 3 or 4 might have a 50% chance at a Difficult task.

I would look at it first this way and then look to "reality" to see what a task difficulty constitutes in real life.

That is, what kind of lock can a person with basic lock picking skills pick 50% of the time in the time frame you want to set, and whatever that kind of lock is defines what a basic lock is in game. Of course you may want to change the adjectives or just go with a numerical rating approach, if that "basic" lock if very complex mechanically etc.

The problem oft comes with getting the scaling the way you want it without a lot of math involved; but here if one can accept the fine detail and math in combat, you really have a larger amount of room to get complex.

In sum, I will admit not sure how these mechanics work. My impression is they use a linear roll, d100 and then want to map that to standard deviations on a bell curve. It seems like extra steps when one could use a mechanic that already is on a bell-like curve (e.g. 3D6). Not here to change the mechanics though, but to get these ideas to fit the mechanics you have.

 

[robertsconley]

My recommendation

• Decide what odds an apprentice with minimal training has at succeeding at a task with limited time and/or resources.
• Do the same for Journeyman, Master, Grandmaster, Nobel Laureate, Legendary (Sun Tzu, Einstein, etc.) . These are your developmental waypoints.
• Decide on the balance between natural talent (attribute) learned skills (experience).
• Build a character at each waypoint don't worry about progression yet.
• Figure out the expected progression between each waypoint. Could be linear or a curve of some type.
• Given the time or resources what can an Apprentice, Journeyman, Master, Grandmaster, Nobel Laureate, and a Legend can do with 95% to 99% certainty?

This is your skill system whether it uses a d100, d20, or 3d6. Each of these is a subjective judgment call based on whatever science you feel is applicable or in many cases genre/setting logic. How people learn and develop is not settled when it comes to the nuances of a specific individual. So you have to sort between the different options.

 

[ffilz]

My recommendation

• Decide what odds an apprentice with minimal training has at succeeding at a task with limited time and/or resources.
• Do the same for Journeyman, Master, Grandmaster, Nobel Laureate, Legendary (Sun Tzu, Einstein, etc.) . These are your developmental waypoints.
• Decide on the balance between natural talent (attribute) learned skills (experience).
• Build a character at each waypoint don't worry about progression yet.
• Figure out the expected progression between each waypoint. Could be linear or a curve of some type.
• Given the time or resources what can an Apprentice, Journeyman, Master, Grandmaster, Nobel Laureate, and a Legend can do with 95% to 99% certainty?

This is your skill system whether it uses a d100, d20, or 3d6. Each of these is a subjective judgment call based on whatever science you feel is applicable or in many cases genre/setting logic. How people learn and develop is not settled when it comes to the nuances of a specific individual. So you have to sort between the different options.

Yea, that all makes sense... Now to find time to do the work... :-)

 

[ffilz]

I'm particularly interested in systems folks feel actually work well for some of the more relevant adventuring tasks, particularly tracking. How do you handle long term tracking across multiple terrains? Roll until you fail tracking doesn't seem right.

On the subject of things like jumping, certainly a good thing to look at would be records, on the other hand, a record long jump MIGHT have only been a 5% chance of success for a legendary athlete. So how would you take those records and scale them back to what might be a 95% success chance for a legendary athlete?

Rob, do you have a cheat sheet you use beyond what you're written up for skills in your Majestic Fantasy rules?

 

[Telok]

On the subject of things like jumping, certainly a good thing to look at would be records, on the other hand, a record long jump MIGHT have only been a 5% chance of success for a legendary athlete. So how would you take those records and scale them back to what might be a 95% success chance for a legendary athlete?

I tackled a bit of this one time. Basically it was figuring out what a nirmal person and their results looked like in the system, then repeat for common successful high end stuff for the professionals.

So don't worry to hard about stuff like olympic records except to make sure it's doable at all. But use things like qualifiers and competion averages to figure different levels. So for... I think I was looking at pole vaulting one time, I looked up the qualifiers to try out for olympic teams as what an expert should hit about half the time. Because if you regularly beat that then you're top tier. Then looked at high school athletic numbers. That was giid because there was bunches of stuff aimed at coaches and trying for realistic expectations of various beginners and first year athletes. That made for a good 'talented + basic training' row, plus it had some ranges for very early beginners that could benchmark for just talent. With all 4, talent - basic - expert - world record, it gave a good sense of scale.

For stuff like tracking I went with what it would take to make a living off the skill. You need the ability to get numbers the untrained can't at get all, at a reliable rate. Then scale system effects/results off that sort of stuff. For the systems I've done this in its OK for heavily invested PCs to push fall into world champ/low end supernatural effects. Other systems will vary of course.