Advice?

[UristMcRandom]

Hello all, I'm new to the forums, but I've been reading OotS for a while now, and I figured this was as good of a place to ask this as any.

I'm relatively new to the field of pen-and-paper RPGs, and I am seeking to create my own system. Nothing too complicated or for commercial use, just for private use in my own groups.

I've got a general idea in my head of how I want things to work, but as I said I'm new to tabletop RPGs and I'm worried about balance. I've done research on things such as stat assignments, but I can't seem to get my own system to balance out. I keep rolling up characters with fairly horrible stats, not to mention I can't find any good resources on ideas for how race and class should affect statistic assignment and skills. If anyone could point me in the direction of such a thing, or has any advice to offer, I would appreciate it.

I can provide more information as needed, just attempting to keep this fairly brief.

Thanks in advance,
Urist McRandom

[Grod_The_Giant]

Particularly given that you're new to the hobby, is there a reason you want to design your own system, rather than use a pre-existing one?

[UristMcRandom]

My goal is to avoid magic as much as possible, as well as things such as demons or fiends. If you could recommend a system that is at least close to that then I should be able to skip the step of creating a system and start on finding/making a setting.

[Grinner]

My goal is to avoid magic as much as possible, as well as things such as demons or fiends. If you could recommend a system that is at least close to that then I should be able to skip the step of creating a system and start on finding/making a setting.

I can point you towards a couple of systems, but have you ever played an RPG? Some RPGs are more complex than others, and it's difficult to know where the line is if you've never played.

[UristMcRandom]

I have played play-by-post RPGs that function similarly to actual or virtual tabletops, but I've never really had a chance to play in person on a table (virtual or otherwise).

Edit: I have also played various computer and console RPGs, not sure if that counts though.

[Grod_The_Giant]

My goal is to avoid magic as much as possible, as well as things such as demons or fiends. If you could recommend a system that is at least close to that then I should be able to skip the step of creating a system and start on finding/making a setting.
Ok. So you want, what, fantasy with exotic monsters but no dragons? More historical middle ages stuff? A modern setting?

[Grinner]

I have played play-by-post RPGs that function similarly to actual or virtual tabletops, but I've never really had a chance to play in person on a table (virtual or otherwise).

What systems?

[UristMcRandom]

Ok. So you want, what, fantasy with exotic monsters but no dragons? More historical middle ages stuff? A modern setting?

The former is the closest to what I'm looking for (although dragons are okay, the whole arcane/divine spellcaster aspect is mostly what I'm avoiding).

Edit:


What systems?

No formal systems, per se, they functioned off of custom systems created by their authors that I do not have access nor reference to.

Edit 2:

Also, if you believe I'm just totally unqualified to do this with my experience, I'll go out and get more experience and ask again later.

[Grinner]

Edit 2:

Also, if you believe I'm just totally unqualified to do this with my experience, I'll go out and get more experience and ask again later.

Probably. Since you've got this thread going anyway, you might as well ride it out.

So from what you said in the first post, it sounds like you're emulating D&D in some fashion. You mentioned that you were rolling horrible stats, which is actually fairly common for these sort of mechanics. What dice are you rolling to make that determination?

[UristMcRandom]

I set base stats by rolling one d6 per stat and adding race bonuses. The base stats are as follows:

Strength
Intelligence
Wisdom
Agility
Endurance
Charisma

I then obtain more detailed stats by rolling one d4 and using the result as a multiplier for the base stat value that the sub-stat derives from.
The sub-stats are the following:

Attack (derivative of Strength)
Skillpoints (derivative of Intelligence)
Perception (derivative of Wisdom)
Hit Points (derivative of Endurance)


For testing purposes: here's a copy of the racial stats I settled on for a human.


Strength: 3
Intelligence: 3
Wisdom: 2
Agility: 3
Endurance: 3
Charisma: 5

[Rhynn]

To design RPGs, you need to understand statistics. (Playing RPGs can help with that, incidentally.)

You say you keep getting horrible ability scores, and the problem is obvious: you're using d6 to roll them. That's a flat distribution: equal amounts of each result 1-6.

If you use e.g. 2d6, you get a curve. 7 is the most common result, 2 and 12 are rare. If you use 3d6, you get a "stronger" curve (3-18, 10.5 average). And so on.

Your big issue, though, seems to be that you're trying to decide this stuff in a vacuum. What's your base resolution mechanic, or do you use multiple different ones?

Your main challenge, honestly, is going to be not knowing enough about RPG systems. To design a system, it helps to be familiar with a lot of them.

... not to mention that there's a pretty good chance there's a system out there that's great for you already; possibly better than what you could create for yourself, given your inexperience, which limits the things you can come up with, given that no one human can come up with a very significant portion of the set of possible ideas...

[Grinner]

*snip*

Well, the reason you're getting wildly different stats is your usage of a single d6. While your mean roll will be 3 (some say 3.5), you have an equal chance of getting any particular result.

There's a couple common fixes for this: multiple dice or point buy.

With the first option, you roll multiple dice, making it more likely that you'll receive an average result. For example, if I roll a d10, the probability that I'll receive any particular result is as follows:

{table="head"]Result|Probability|Result|Probability
1|10%|6|10%
2|10%|7|10%
3|10%|8|10%
4|10%|9|10%
5|10%|10|10%
[/table]

However, if I roll 2d10, the probabilities become weighted towards the average result*:

{table="head"]Result|Probability|Result|Probability
2|0.000000|12|9.999700
3|1.000300|13|9.005600
4|1.987100|14|8.008100
5|2.979800|15|7.015400
6|4.008900|16|6.017000
7|5.020900|17|5.027200
8|5.997900|18|4.017100
9|7.003100|19|2.989700
10|7.960200|20|1.986900
11|8.970400
[/table]

Unless you're willing to rework how your stats relate to the rest of the system, this is sort of difficult. See you could try rolling 2d6 and taking the better result, but that's not going to give you a bell curve. Rolling 3d6 and taking the middle result might work though...

Your other option is implementing a point buy system, where players would choose their own stats. You'd probably want to give each player a little better than average, so that would be maybe twenty points to spend on stats (3 points per attribute plus another two to be allocated however the players want).

The problem with point buy systems is the players. It's human nature to optimize in order to deal with problems. It's just instinct, y'know? As such, these sort of systems lend themselves towards min-maxing. Sometimes designers will counteract this by making higher stats more expensive**, but the method is of limited usefulness.

On a final note, it's worth noting that while point buy systems are easily optimized for, they're still inherently fair. Rolling for stats can produce power divides between players just by dint of one player having superior results. It's probably not a huge problem here, but it's something to keep in mind. Personally, I've toyed with the idea of awarding bonuses to players inversely proportional to the results of their stat rolls. Feel free to make use of it.

*The results don't look quite right, but I think they get the point across.
**In this case, perhaps by making a stat of 5 worth two points instead of one, and a stat of 6 worth three points instead of one

[Rhynn]

Well, the reason you're getting wildly different stats is your usage of a single d6. While your mean roll will be 3 (some say 3.5), you have an equal chance of getting any particular result.

:smallconfused: Some say?

1+2+3+4+5+6 = 21, 21/6 = 3.5. The (arithmetic) mean is 3.5, that's it. The median is likewise 3.5 (3+4/2).

All of the rest is good stuff, though, and important stuff. When I start learning a new RPG, I usually chart the base resolution mechanic probability if it isn't a flat distribution. When I started creating a Fuzion hack of Cyberpunk 2020, the first thing I did was chart a bunch of different dice combinations' resolution mechanics, and considered attribute and skill ranges and difficulty numbers to arrive at the correct combination of all those elements.

[Grinner]

You also mentioned trouble deciding on how races and classes should affect stats and skills. The thing is, there's really no definite answer here. It's up to you to determine how important those decisions are. One thing to keep in mind is how this decision will affect the rest of the system. How are you going to balance around this?

Judging by the human racial modifiers, it's pretty important here, since the modifiers appear to account for half or more of the stats' totals.

I don't know how your system treats action difficulty, but you'll want to be careful. Depending on how you balance the system, certain tasks will be much easier for some races than others. Take the humans for example. Since they receive such a hefty bonus to charisma, you can expect most players to play humans as socially-focused characters. Now, if you balance the system to keep humans from automatically succeeding on every social action, you may make it impossible for other characters of other races to succeed socially. On the other hand, if you set the bar too low, then humans will always succeed flawlessly.

Somewhere in the middle might be ideal. Then again, I've got to ask myself something: "Why is desirable to have characters fail?" The answers to that question might be beyond the scope of this discussion...


:smallconfused: Some say?

1+2+3+4+5+6 = 21, 21/6 = 3.5. The (arithmetic) mean is 3.5, that's it. The median is likewise 3.5 (3+4/2).

See, when you put it like that, it makes sense. When I'm not looking at the math, it doesn't.

[Rhynn]

Just for reference (this is obvious if you know the math, not so much if you don't):

The mean (average result) of any die with sides numbered 1...n is always (1+n)/2. d8 is 4.5, d20 is 10.5, etc.

But, as I mentioned and Grinner demonstrated, the distribution can be very different! d20 and 3d6 have the same mean (10.5), but different ranges (1-20 and 3-18), and completely different distributions (1 or 20 are a 5% chance each on d20; 3 and 18 are a 1/216 chance each on 3d6).


Somewhere in the middle might be ideal. Then again, I've got to ask myself something: "Why is desirable to have characters fail?" The answers to that question might be beyond the scope of this discussion...

I think it's a very important question for system design. There's so many ways to go.

You might not want random determination at all, but rather either a "deterministic" system (cf. Amber Diceless) or some kind of resource-management, where the PCs must choose when to succeed and when to fail.

There's also systems where failure is required for advancement, like Burning Wheel / Mouse Guard, where you need both successes and failures to advance your abilities.

You might want success to be automatic in e.g. unresisted tasks (I actually run a game, Artesia: Adventures in the Known World pretty much this way, just because I've found the PCs outstrip the common unopposed difficulty numbers quickly; it is not intended in the rules, though), and only in question in tasks that are opposed by something... but, then, you might make a system where all tests are opposed by some resistance value (like in HeroQuest, where "the world resists you", essentially).

[UristMcRandom]

Wow... you go offline for a few minutes and so much happens. Thanks for all the help so far, I will definitely be looking into implementing it (or something similar, my independent research is ongoing) without shaking everything up too badly (although that shouldn't be too difficult courtesy of my bare-bones system).

Seeing the actual numbers is especially helpful, as I'd never actually thought from the POV of adding extra dice or using a die with more sides. In retrospect it seems a fairly obvious solution. (Maybe my character's low perception score was justified...)

[Rhynn]

Seeing the actual numbers is especially helpful, as I'd never actually thought from the POV of adding extra dice or using a die with more sides. In retrospect it seems a fairly obvious solution. (Maybe my character's low perception score was justified...)

This is why experience with existing systems is critical in designing systems.

Let's say that there exist, currently, 10,000 ideas. An individual person is unlikely to come up with more than, say, 10 of them on his own, if even that. But by reading enough of other people's systems, you can learn, say, 1,000 ideas.