Tuesday, April 10, 2007

Funky mechanics

In case the two people who read this have been wondering, I've been popping in and out of town a lot lately on business, which has cut into my posting rather dramatically. Unfortunately, this isn't going to change anytime soon, so I will have to adapt.

(Actually, Statcounter tells me that I have significantly more than two regular readers.)

Anyway, I was up in McHenry County, Illinois last week, and I was thinking about how skill systems tend to work in rpgs. One thing that I find odd is that, in real life, one of the benefits of becoming more practiced in a skill is that you become more consistent. In rpgs, the variability of outcome often either stays the same (d20) or increases (Storyteller). There are some exceptions, like Savage Worlds, but these exceptions often have some other issues... such as discarding the possibility of highly skilled individuals accomplishing feats that are impossible for the less skilled.

I think both of these are important. As one increases in skill, one should become (1) more consistent and (2) able to accomplish more impressive feats.

How to model this?

Here's one way. Let's say we have a system that uses d6s. Skills have three levels: Apprentice, Journeyman, and Master (or Beginner, Intermediate, and Expert... or whatever). Each level has a numerical rank. In order to have a rank in Journeyman or Master level, you need a minimum rank (say, 3) in the level below it.

For each level, you roll a number of dice equal to your rank and take the highest roll, then you add the results of each level together. So, for an Apprentice you get a result of 1-6, a Journeyman 2-12, and a Master 3-18. Alternately, you can choose not to roll at all and simply take your rank in the level as your result. For example, if you are a Master (2) with 3 ranks in both Apprentice and Journeyman, you can automatically get a result of 8 without rolling.

Also, you can move ranks down. The above individual could roll Apprentice (3) Journeyman (4) and Master (1) if she wanted to - the reason why she might isn't apparent yet. I'll explain that in a minute.

Easy enough so far?

Modifiers can work well in this system. It allows you to impose modifiers that have a more significant effect on less skilled. Each modifier would have a single level and rank - which in the case of a penalty would be negative. The ranks are added to/subtracted from the appropriate levels of the skill. This can result in a negative number.

If there's a negative number, then that level's die is subtracted rather than added. This is why you might want to sacrifice a higher level rank for a lower level one - to remove penalties.

This is still a work in progress, obviously. I'm using this space primarily as a brain dump at the moment.

13 comments:

longcoat000 said...

I'm finding it hard to "get" what you're getting at. I've got how you can add your skill levels and ranks together, but the actual rolling portion is fuzzy.

As written ("For each level, you roll a number of dice equal to your rank and take the highest roll, then you add the results of each level together"), it looks like you roll the appropriate number of dice (based on skill level) for each rank in said skill level, then add the results together. Thus, in your example, your Master (2) would roll 3d6 for their apprentice level, 3d6 for their Journeyman level, and another 2d6 for their Master level, or 8d6 vs a target number.

I think that there may be a disconnect in your example, because you show a master as being able to generate a result of 3-18, which (as written) assumes that they have one rank per level, but your example uses a minimum rank of 3 to progress to the next level. This could be taken to mean (or might be exactly what you meant, but it's vague) that the number of dice rolled per rank depends on the level being rolled. If that's the case, then it could be interpreted to mean you would roll 3d6 for your beginner roll (1d6 per rank), 6d6 for the journeyman roll (2d6 per rank), and another 6d6 for your master roll (3d6 per rank), with three shots at beating a set target number (or maybe adding all 15d6 together to try and beat one massive target number).

Depending on how target numbers and degrees of success are defined (see below), it could also be rolled as 1d6, 1d6, 1d6 (beginner), 2d6, 2d6, 2d6 (journeyman), 3d6, 3d6 (master), with each individual result being a chance to beat a target number. This jibes better with your example of taking your total ranks as the result of your roll, so you get a minimum of one 8 rather than trying to beat a set target number with multiple combinations of dice.

I think you need to define how successes work. Is the goal to beat a straight target number (with degree of success measured by how much you beat the target by), like d20, or to simply generate a set number of successes, like the Storyteller system?

Either way, I don't see how sacrificing a higher level to add to a lower rank would be beneficial. Everything I outlined above seems to indicate that it's either better to roll at higher levels (because this gives you more dice), or doesn't matter because you roll the same number of dice (with bonuses/penalties) no matter how you shuffle the ranks around.

Or am I missing something?

szilard said...

For each level, you roll a number of dice equal to your rank and take the highest roll, then you add the results of each level together.

...so for each level you take a single die result (the highest of them).

longcoat000 said...

Ah, each level is a seperate pool of dice, and the highest individual die from each pool is added together. It was sitting right in front of my face, but I couldn't see it. Now the die rolling mechanic makes sense.

With this, I could see why you would want to swap out a rank of a higher level of a skill for a lower one, but it only makes sense to do so in order to "balance out" the differences in two or more ranks. For instance, our Master (2) tries to do something with a -3 Beginner rank penalty, giving a Beginner pool of 0 dice, Journeyman pool of 3 dice, and a Master pool of 2 dice. He can drop a rank of Journeyman (bringing it down to 2) and add it to Beginner (giving a 1 die pool), but there's no point in reducing any of the pools any further, because all of the pools are within 1 die of each other.

Now, if this Master's ranks were different (B3, J3, M5) and he were faced with the same task (-3 B pool), he'd have pools of 0 B, 3J, and 5M. In that case, it would make sense to drop two or three dice from your J pool and add it to the B pool, then drop two or three dice from your M pool and add it to the J pool in order to "even out" your die pools. This assumes that you can only drop from existing ranks to add to the level below. Otherwise, there is no incentive to increase your rank in lower levels past the bare minimum required for the next one because you can just keep dropping dice from your highest level to the lowest.

You can get around this minimax approach by creating a rule stating that a skill rank cannot be greater than the rank of the level below it, which would force players to spend XP on lower level skill ranks to reflect competency. Thus, the Master (2) could buy up to Master (3) without any worries, but would have to buy Beginner (4), then Journeyman (4) before they could buy Master (4).

If this approach were used, then I could see changing the minimum rank needed to advance in skill level from a static number to be the minimum rank one level below the level you're trying to advance to. So, in order to buy Journeyman (1), you'd have to have a minimum of Beginner (2), because Journeyman is the 2nd level of skill competency and you're advancing from the 1st to 2nd level of competency. To advance to Master (1), you'd have to have Journeyman (3), because Master is the 2nd level of competency and you're going from 2nd to 3rd.

If XP costs for higher level skill ranks increase (i.e. XP cost = desired rank multiplied by level, so B would have a multiplier of 1, J of 2, and M of 3), then this rule for skill costs would also better simulate real-world examples. For instance, there are a lot of people who are really good at doing simple things (high Beginner rank), but don't want to put in the time or effort (spend the higher-cost XP) to do things better (advance to the next level of competency). This could also allow higher levels of competency than 3rd (Master), which would reflect true devotion to a skill or godlike abilities.

Now this makes a lot more sense and is a reasonably fast rolling mechanic (just toss X number of different colors of dice at the same time and pick out the highest). Now, the only problem is figuring out target numbers and establishing bonuses / penalties. Would they be based on the final target number (target 1 - 6 is a Beginner mod, 7 - 12 a Journeyman mod, and so on)? And is the degree of success (amount above / below the target number) a factor in anything?

Also, would raw abilities (str, int, etc) also be counted as one of these die pools? Doing so would allow characters to get by doing the easiest things on natural talent, but having an actual skill rank, even the lowliest Beginner (1), would have a profound impact on their competency. It would also allow for the ability/skill mixing-and-matching that I was particularly fond of in the Storyteller system. For instance, climbing a tree might be your dexterity-equivalent plus athletics-type skill, but picking up a heavy weight could use strength-equivalent ability plus athletics-type skill.

longcoat000 said...

Another couple of thoughts (as if they weren't long enough already!):

1) Each full 6 points the roll exceeds the target increases the degree of success. So rolling an 8 to climb a fence (difficulty 4) would get you up and over it without snagging your jeans or breaking an ankle. Rolling a 14 to do the same thing lets you do so effortlessly, maybe making very little sound. Rolling a 24 on the same task could let you vault the thing one-handed ninja-style and land in absolute silence.

2) Allowing an ability pool in addition to the skill pool also allows for universal conditional modifiers to be applied which normally wouldn't affect a skill, but have the potential to depending on the character's raw ability. For instance, having a headache or losing blood from being wounded don't actually affect your knowledge of a skill, but they would impede your ability to use a skill. Conversely, not having the right tools (or having the perfect tools) for a job wouldn't affect your raw ability, but they would impact the skill itself.

However, if the ability penalty exceeded the character's actual ability, then the penalty would spill over to the skill and be applied against it, starting with the lowest level of competency and working its way up.


For instance, having a broken leg could be good for a -4 penalty to most of the things a character could try and do physically. A very high strength character (strength 4+) could still use their arms to pull themselves up a tree (not suffering a penalty to their skill because their ability is equal to or greater than the conditional penalty) with raw upper body strength, but a lower-strength character wouldn't have that luxury. Any "leftover" penalty would be applied against their skill beginning with the Beginner die pool and working up from there when all skill ranks from a level were "used up".

szilard said...

Some good thoughts there. I hadn't really gotten to the point of developing this to a full mechanic. I was mostly playing around with something I thought was a cool idea.

I'd probably peg target numbers at multiples of three rather than six. I want there to be some tasks that most beginners can get fairly easily... and that competent Journeymen don't even have to roll for.

Attributes are trickier.

How's this:

Extraordinary attributes alter the number of dice of a character's Beginner level.

In addition to lowering dice from a higher level to a lower one, you can raise dice from a lower level to a higher one, but at a significant penalty (maybe a 4 to 1 rate). So, if you have B (3) and your attribute gives you an additional 2 dice, you could roll either B(5) - or B(1) and J(1).

I like the idea of having some status effects change the attribute modifier.

longcoat000 said...

I was thinking minimums and maximums for each competency level rather than averages for the whole multiple of 6 thing. I figure that each level of competency averages 3.5 on a single die roll, so an average unskilled task (attribute only) would have a target number of 4, beginner would be 7, journeyman 11, master 14, and so on.

I like using attributes because then the system uses a character's natural talents, and a naturally gifted character is able to perform simple tasks easier, but training plays a much more important role later in the game (unlike the Storyteller system, where a character's natural attributes were equally important, if not more so, then the abilities they were paired with).

If attributes aren't used, then how would you simulate a character trying to do something they had no skill at? Would they get a "phantom" B1 skill if they had the corresponding extraordinary attribute?

I don't know about using a lower level of competency to simulate a higher level, even an a 4-1 trade ratio. It makes more sense to me for a highly competent person to be able to use their knowledge to offset problems they may have faced or learned about earlier in their training or experience than for someone not as competent to try and create lightning in a bottle.

What about a "rule of maximums", where if you roll the maximum amount on your roll (each pool coming up with at least one 6), you can roll another die and add it to that roll. If that die is a 6 also, they get to keep rolling and adding until the die comes up as something other than a 6. This would still allow lesser skilled characters a chance to do something amazing, but the chances of these "brilliant" insights decrease as a character becomes more skilled because they learn these same "amazing" insights as part of their training.

szilard said...

Hmmm...

It seems like you are proposing another level:

Untrained/Attribute + Beginner + Journeyman + Master

I like that in principle, but it adds another level of complexity to a mechanic that is already a bit complex.

I'll have to give it some thought.

It makes more sense to me for a highly competent person to be able to use their knowledge to offset problems they may have faced or learned about earlier in their training or experience than for someone not as competent to try and create lightning in a bottle.

In general, yes... but there are cases in which people with a small amount of skill but a lot of natural talent can accomplish tasks that someone of equal skill is simply incapable of accomplishing.

I think that the 4-1 trade (or something vaguely like it) could work, with the recognition that some tasks would have skill minimums.

The rule of maximums has some odd statistical effects: someone with, say B(3)(J1) is significantly more likely to get a 13-18 result than someone with B(3) J(3). That seems weird.

longcoat000 said...

It seems like you are proposing another level:

Untrained/Attribute + Beginner + Journeyman + Master

I like that in principle, but it adds another level of complexity to a mechanic that is already a bit complex.


Correct. I thought about how this would be adding more complexity, but after a lot of internal debate decided that it really doesn't. As I've said before, it's already a very simple and elegant mechanic. Count the numbers, use a different colored die for each competency level and roll. Almost like a color-coded Yahtzee, only you're adding the highest die of each color group.

I think that the 4-1 trade (or something vaguely like it) could work, with the recognition that some tasks would have skill minimums.

I've never really liked skill minimums, because to me they've always smacked of limiting a character's options when they really want to do something neat. I was going to write something about trading dice pools is a tactical decision, but then argued myself out of it. Something still bugs me about it, but I can't quite put my finger on it.

...but there are cases in which people with a small amount of skill but a lot of natural talent can accomplish tasks that someone of equal skill is simply incapable of accomplishing.

This was bugging me yesterday, but I thought back to a couple of game systems (Shadowrun, Earthdawn) and came up with the idea of larger dice pools generating better results. Thus the rule of maximums, where a character with a lot of natural talent and little training has a good chance to do something better than a different character with equivalent training and less natural talent, or even someone who is better trained. But a person who is better trained will perform a task consistantly better than a character who is relying on talent to see them through. The chance of a talented character outperforming a more skilled character is close when their skill levels are close (say Beginning and Journeyman), but better training (Beginner vs. Master) will consistantly trump talent.

The rule of maximums has some odd statistical effects: someone with, say B(3)(J1) is significantly more likely to get a 13-18 result than someone with B(3) J(3). That seems weird.

With the rule of maximums, the entire pool doesn't have to be 6's, just one die from each pool (since you only count the highest die from each pool). So someone rolling two pools of three dice has a greater chance of getting at least one six per pool than someone rolling a pool of three dice and a pool of one.

I also just realized that I keep calling each competency level a "pool" of dice. Does this terminology work for you (for now)?

szilard said...

The mechanic is relatively simple, but it does require sets of differently-colored dice (albeit d6s) in order to retain that simplicity. I'd like to keep the number of levels as small as possible. I'll have to give it some thought.

Hmmm... what if instead of a separate attribute pool (fine word, for now at least) you could substitute your attribute for your numerically lowest pool (of at least one die)?

Let me know if you figure out what bugs you about the die-trading. I think the potential for tactical decision-making with it is sort of cool.

On skill minimums - I'm not a fan of them being used broadly... but there are some things that take a minimum amount of knowledge before you have a chance of success (brain surgery, computer programming, translating something from Mandarin Chinese into Arabic, etc.).

Something about the rule of maximums still isn't sitting well with me. Maybe it is just too much dice-rolling. I'll give it some thought.

longcoat000 said...

Hmmm... what if instead of a separate attribute pool (fine word, for now at least) you could substitute your attribute for your numerically lowest pool (of at least one die)?

I like the idea of augmenting knowledge with talent (attribute + skill) more than replacing knowledge with talent (swapping attribute for skill). I think that it would put too much emphasis high attributes and not enough on skills.

By swapping an attribute pool for a skill rank, you can stick characters at artificial static levels of development. For example, if your character had a skill level of B1, but an attribute of 4 dice, they would have to spend an insane amount of XP to raise their skill higher than their natural ability already compensates them for, or just keep grinding until they can progress to the next level of competency. Once the next level of competency is reached, suddenly their J1 skill level becomes J4, and their B4 (with attribute) drops down to whatever minimum skill level they had.

With all the time and effort that they spend improving their skill, the actual dice rolled don't increase for a long time. From a player's standpoint, spending XP and not getting anything back (more dice to roll) is discouraging.

BUT, the idea of using individual dice from your attribute pool to increase your various skill dice pools is appealing. It would reflect how exceptional attributes help at lower skill levels and become less of a factor at higher skill levels. A character with a 3-die attribute pool could add them all to one pool or spread them amongst several different pools. But then you're hit with the problem of players dumping them all into one low pool, which brings us back to the "XP dump" problem I pointed out earlier.

Let me know if you figure out what bugs you about the die-trading. I think the potential for tactical decision-making with it is sort of cool.

Logically, "trading down" makes sense, because (as I've said) a more competent character is able to draw on their experiences and training to compensate for skill penalties. It doesn't make logical sense to me to go the opposite way, because the character hasn't encountered these types of problems or hasn't been trained in what to do with them.

But I also don't like limiting what a character can and can't do by their skill level. I tried fixing this with the rule of maximums, and while it allows for the possibility for a character to "capture lightning in a bottle", it doesn't address the problem of a character deliberately trying to reach beyond their grasp.

What about limiting "uptrades" to only come from a character's attribute pool, and each level of competency purchased "costs" one die? So if a character with a 4-die attribute wanted to make an intricately engraved rocking chair worth some $$$ (normally a journeymenan level target of...say...12), but they only had a B2 skill in carpentry. They could allocate three dice from their ability pool (reducing it from 4 to 1) and apply one die as a "phantom" journeyman skill level (3 dice allocated, less one for the beginner level and one for the journeyman level, leaving one die). Or they could roll what they had and hope for two 6's and the rule of max to kick in.

If this sort of ability trading were allowed, I'd say that you could either add to any existing skill level or add up to one "phantom" level above what you currently have, and not allowing it at all for skills the character doesn't have.

On skill minimums - I'm not a fan of them being used broadly... but there are some things that take a minimum amount of knowledge before you have a chance of success (brain surgery, computer programming, translating something from Mandarin Chinese into Arabic, etc.).

I think that the problem could be solved with the target number. Things like brain surgery would have a sufficiently high target number for untrained people to think twice about trying to cure a headache with a power drill and shunt. Midieval(?) surgeries often hurt more than they helped, but sometimes they did work...

And if the rule of maximums were in effect, the average target number for a beginning task would be 7, which means that an unskilled character would have to roll a 6 with their attribute pool to be able to do it.

It could also be fixed with characters attempting something that they weren't skilled at to suffer an attribute penalty, but then you'd have to figure out which skills would get what kind of penalty unskilled use. Trying to outrun someone may carry no penalty, trying to throw a spear might be good for a 1-die penalty, and brain surgery might be a 4-die penalty. Not quite a dealbreaker, but it does involve a lot more work. But it would give characters a chance to do things that they weren't skilled at.

szilard said...

I'm going to be AWOL from the blog for a few days, but I'm thinking...

What about if there are the three levels... and attribute dice can add to those levels at differing rates:
- adding a die to pool B: 1 for 1
- adding a die to pool J: 2 for 1
- adding a die to pool M: 3 for 1

longcoat000 said...

That works. I gave it some more thought on the ride home yesterday, and started wondering why the up-trade value couldn't be relative depending on the attribute or skill you're shifting from. The progression would be something like this:

Adding a die to the pool one level above (Attribute to B, B to J, or J to M): 2 for 1

Adding a die to the pool two levels above (Attribute to J or B to M): 3 for 1

Adding a die to the pool three levels above (Attribute to M): 4 for 1

Now, you don't have to take all of the dice from one pool (you can spread the hit out over several pools), but the ratio is determined by the furthest level of skill you're taking dice from. Take a character with the following skill:

A4, B3, J3, M1

To add one die to their M level, they could take two dice from J, three dice from B & J, or four dice from A, B, & J. I've worked out the numbers, and it looks like this holds up to whether a player does everything at once (four dice from A, B, & J) or tries to do it in 1-level shifts (2 for 1 A to B [net loss 2 dice], 2 for 1 B to J [net loss 1 die], 2 for 1 J to M [net loss 1 die]).

The problem with this is that it makes upshifting dice pools a lot more complex. I think that your solution may be better in game terms, but I would bump up the trade ratios to 2, 3, and 4 to 1 respectively.

Incidentally, I'm also thinking that attributes and skill ranks go from 0 to 6 (3 being average), and making 6 skill levels available rather than 3 (but XP costs above 3 would be horribly expensive and reserved for true grand masters and godlike beings).

Hope you're having fun, wherever you end up...

szilard said...

Hotel has free computer/Internets. yay!

I was thinking levels would max out at 6, with 3 ranks being necessary before you buy into the next highest level... so few people would be likely to get beyond 3 in a level.

My last post, by the way, was meant to suggest that attributes are only used for adding dice to other level - that there is no rolled attribute pool.