Dice System: Does it work?

[Lonoto]

I, as you can probably guess, am attempting to design my own tabletop RPG. Some of my friends at school, after playing D&D with me for a while, asked me if I could create my own system. They wanted something universal to use with many different settings, but also something a little more realism based than D&D, but with the ability to be used for high-powered (uber-real, if you will) characters as well. So, here I am, researching RPG design and trying to create a system based on those ideas.

One thing I have noticed about RPGs are the many different types of dice systems. Roll-under, roll-over, successes, and many different varieties within these three main systems that I have seen. For a system completely based in realism, I think maybe a roll-under system would be a good idea, but since we want to be able to get into the uber-real with this system, and since it more simple, I think a roll-over system would work (successes and realism... I don't think the two go together).

So, I have looked over the different types of roll-over systems, and there is one common problem I have found with many of them. Once you get so high in skills, there is almost no need to roll... Now, I don't know about what you guys think, but I don't think that is a good thing. At high levels in D&D, you either make it or not with most skill checks, no need to roll. Combat between two creatures becomes: Your first attacks automatically hit and your last automatically miss, unless you roll a critical. I wanted to keep the amount of variation that makes for interesting combats and tense situations when trying to do something, but also allowed for a very skilled technician, for example, to be able to automatically perform simple repairs or some such.

The system I have come up with makes the skills and the base attribute or statistic numbers add-up to give a constant bonus. Then, based on how high your attributes are, they have a dice rating which determines how many dice are rolled, which allows variance (excuse me if this is the wrong mathematical term, I am taking math in French) to continue to play a part in the skill checks at higher levels. I was just wondering what the people here think about dice systems in general and my system in specific. I can elaborate a bit more if you need me to.

Lonoto

[Ron Edwards]

Hello,

Welcome to the Forge!

I have one point and a series of related questions.

The point may be kind of disturbing to you: there is no difference between roll-under and roll-over, in mathematical terms. They are only procedural avenues to the exact same distributions.

I'll use a pretty simplistic example. Say everyone starts at 9 or 10, and the successful d10 roll must be equal to or higher to one's number.
As characters get better, their numbers go down. Or contrast, everyone starts at 1 or 2, and the successful d10 roll must be equal to or lower than one's number. As the characters get better, their numbers go up.

You can do dice pools, multiple-dice fixed rolls, single die rolls or roll X and keep Y. You can subtract one roll from another. You can use cards or colored pebbles instead of dice. You can factor in skill values. You can be trying to beat target numbers or trying to beat someone else's roll ...

... and it doesn't matter. No matter how you layer over that simple system with "stuff," the basic point remains: over or under makes no difference whatsoever. That's a big myth of gamer culture and of absolutely no consequence to play itself.

On a related point, nor is there a difference between a system in which you apply modifiers to a roll, then check vs. a target number; and a system in which you apply modifiers to a target system, then roll.

The vast "variety" among most role-playing games is therefore no variety at all. Most people, when they say they've played most systems out there, have effectively only been playing one.

Now for my questions. Let's imagine you and your fellow players all sittin' around the table or whatever, and playing your game. Tell me - how often do you think they'll roll? Once per player per hour? Ten times per hour? Less? More?

In what circumstances within the game (combat, etc) would that frequency go up? Or down?

When one person is engaged in rolling, how much are the others involved? Do they have any kind of input or response to someone else's roll, or are they in "wait my turn" mode?

How much in-game stuff do you think a typical session would get through? Just to be all combat-y for a moment, one fight? Ten fights? Twenty?

And finally, although I realize you're aiming at multiple possible settings, what would player-characters do in many of your games? Go on missions? Investigate, then strike? Wander around for a while, then develop conflicts? Or what?

Best,
Ron

[Lonoto]

Thanks for the quick response!

Yes, I do realize that mathematically roll-under and roll-over have equal chances for success, they are just different methods. The thing that I find is that when you want to have a system where the power can increase exponentially or get to very high numbers, a roll-over target number is more simple.

The way I envision play is that whenever the character's attempt to do something that their character might have difficulty doing, or where time is of the essence and they need to do something quickly, they will roll a skill check. I see this happening many times in a playing session, so probably in the upper limits of 30-40 rolls per *non-combat* hour. In combat, the numbers would rise by quite a bit.

When one person is rolling, depending on circumstance, other player's or NPCs can interrupt. Players or NPCs can either abort their next action or use an action they had been saving to interrupt each other, something that, realistically, happens a lot in combat and other types of encounters I believe.

In game, I think that per session there would be between 1-10 fights, but this game can also go an entire session without a fight. (Example: We are playing a High School-based campaign, in the session the character's are trying to sneak bear into the party and not get caught, use social skills to skillfully avoid fights, then try to overcome their massive hang-overs to play a basketball game the next day.)

I envision this game being able to cover multiple facets, so I can't really define it. The players may, in one setting, be more investigative and sneaky, and then in another be kicking the doors down and up-front about everything they do. I think that aspect of the game depends on the setting (which is why I am planning on setting specific supplements like: Mecha, Fantasy, Sci-Fi, Anime etc...).

I hope that these answers can help you better envision what my game would be like, and in turn allow you to answer my previous questions. Thanks.

Lonoto

[coxcomb]

I know this doesn't get to the meat of your question, but have you tried The Hero System, or GURPS, or both?

Both of those games (in their current forms) have similar goals of universality and scalability while modeling some form of "realism". I personally think that The Hero System would be worth study, even if you are set on creating your own system. It has dealt with some of the problems that you will run into.

As for your actual question:

Your description doesn't make it entirely clear to me what you're going for.

Can you give a concrete example of how you envision your system working. Take an in-game situation, say a character is trying to bypass security systems to get into a museum or something like that. Can you tell us how that would work in play?

[John Burdick]

My suggestion is to find some common element or feel between a wide range of possible games. Maybe the varied works of a single movie director or a single actor. Anything that allows you to establish something of a vision, while leaving all the details free.

In game, I think that per session there would be between 1-10 fights, but this game can also go an entire session without a fight. (Example: We are playing a High School-based campaign, in the session the character's are trying to sneak bear into the party and not get caught, use social skills to skillfully avoid fights, then try to overcome their massive hang-overs to play a basketball game the next day.)

One idea would be to go with the idea of young adult stories. For anime, you would have high school anime. Fantasy could be like Harry Potter. An action premise could be like the Karate Kid. That's just a quick thought.

There are games that combine a bonus based on a character trait and a number of dice determind by some trait. The ones I know of use a concept called "roll and keep". You roll a number of dice determined by some stat, and keep a number of dice. The games I know never change the number kept. The sum of the kept dice and the bonus is the result. They can work okay.

John

[Lonoto]

I've seen certain other universal systems, like FUZION, which is a combination of Hero and Interlock. I find that a lot of what is in these is sort of what I want, but not entirely. Are Hero and GURPS available for free? Because if not, there is no way I can afford to buy them just for purposes of analyses.

Anyway, here is an example of a character trying to use a skill in my game. Say the player is trying to bypass a security system as you said. They would see if they had the appropriate skill and take their Skill bonus from that (say it is +60) and add it to their REA (Reasoning) statistic, which we'll say is 20 (5 points above the average of 15, 30 is the normal human maximum). This statistic gives him a Dice Rating of 4 (Dice Rating is relevant stat/5, rounded off to the nearest whole number). For skill checks, a d10 is rolled, so this person would roll 4d10 +80 (60 skill + 20 REA) against a specific difficulty based on how hard it is to crack the security system in question. Using expanded rules from my system, degrees of success could be used and require multiple rolls and successes before you can full crack the code (I won't go into too much detail of this in this example). I hope this gives you a general idea of what I am looking at with my system.

[coxcomb]

Are Hero and GURPS available for free? Because if not, there is no way I can afford to buy them just for purposes of analyses.

For free? No. They are both in the "mainstream publishing" category.

Say the player is trying to bypass a security system as you said. They would see if they had the appropriate skill and take their Skill bonus from that (say it is +60) and add it to their REA (Reasoning) statistic, which we'll say is 20 (5 points above the average of 15, 30 is the normal human maximum). This statistic gives him a Dice Rating of 4 (Dice Rating is relevant stat/5, rounded off to the nearest whole number). For skill checks, a d10 is rolled, so this person would roll 4d10 +80 (60 skill + 20 REA) against a specific difficulty based on how hard it is to crack the security system in question. Using expanded rules from my system, degrees of success could be used and require multiple rolls and successes before you can full crack the code (I won't go into too much detail of this in this example). I hope this gives you a general idea of what I am looking at with my system.

It's still very difficult to evaluate and comment on your system without knowing all of the ranges and expectations for scores. If a normal human stat ranges from 1 - 30 (giving dice of 1 [or 0?] - 6 d10), then the skill bonus of +60 seems very large. Where does +60 fall in the range of skill bonuses? Is it exceptional?

Also, you get double bonus for your stats because you add the value *and* base the dice rolled on it. Is that intentional? I assume it is, but what design goal does it serve?

It looks as though, from the little you have shown of your system, you really want the bulk of resolution to be based on static and reliable numbers, with a (comparatively) small amount of randomization included. Is that specifically the intent? Your example yeilds a number from 84 to 120 with a reasonably reliabe average in the 100 range.

I will also point out that adding up large numbers does slow down play (although 6d10 + X isn't terrible, it will still have more handling time than the d20 + X of D20).

None of these observations is meant to indicate a "Bad Thing" about your design, I am just curious to hear more about your reasons for setting it up this way.

[Lonoto]

An 'average' guy has about a 30 in any given skill, for Rank anyway. There are two 'layers' to the skill system, rank and level of Expertise. The maximum Rank a skill can get to is 100 in a realistic campaign. The level of Expertise effects what types of things you can accomplish with the skill, as well as changing the rank bonus for things that are below your level (giving you a bonus) or above your level (giving you a penalty). This system is fairly simple, except for the case of combat skills and Expanded Sports skills*.

I am aware of the redundancy offered by stats, but more often than not at higher levels of expertise the rank bonus outweighs it and gives the skill a more 'constant' feel.

As you pointed out, I am looking for fairly reliable randomization but with the ability to still fail at tasks through bad luck or oversights, or maybe succeed at something that normally you shouldn't have through pure luck.

The reason for the attributes giving two types of bonus is simple: I wanted the attributes to give a constant bonus, but one of small value, just enough to distinguish between 'levels' within the dice rolled. (Someone with an 18 rolls 4d for checks with that stat, but someone with a 22 has a slight edge with his +4 constant. Notice the use of the word 'slight'. This edge is more with pure attribute checks because only d6's are rolled.)

I used this system because I wanted a way to keep a small amount of randomization even at the upper spectrum of skills, and not use a static 1 or 2d10, where once constant bonuses exceed 30-40 the dice become less and less useful.

*If a campaign is going to have a focus on sports for the characters, such as the high school campaign I described in my earlier post, I am providing rules to expand the sports skill. It will detail ways to play out the basketball -or other sport you plan on playing- game in the session. These, however, shouldn't be the primary focus of a campaign and shouldn't be the biggest concern, more of the second biggest concern or third maybe. If you want to focus on sports and make it primary, go out and play it in real life :P.

Lonoto

[Valamir]

Lon...there are about a bazillion systems out there that fit the bill of  "something universal to use with many different settings, but also something a little more realism based than D&D, but with the ability to be used for high-powered (uber-real, if you will) characters as well. So, here I am, researching RPG design and trying to create a system based on those ideas."

I highly recommend finding one off the shelf that meets your needs and save yourself the struggle of trying to reinvent the wheel.

[Ron Edwards]

Hiya,

Two free systems which I think you really ought to check out are Pocket Universe and JAGS. Quick Googling ought to get you to each of them.

Best,
Ron

[Lonoto]

Yeah, I know inventing my own system is a bit drastic, especially a universal system. I have kind of started to lose my steam on this project after I fully realized the scope of what it is I would have to do. For the campaign one of my friends wants to run (WWIII) I think I am just going to use JAGS (which I already had). But the question from my original post still remains: Do you think this task resolution system can work (for any other game I might try to design).

On a side note, does anyone here have a 'sample' design document to show how you should set one up? Just stuff like how much information a design document needs in each section and stuff would really help. Thanks.

[Garbanzo]

Lonoto:

Both GURPS and HERO are big, expensive books.  But Gurps Lite is a  free download.

GURPS Lite is a 32-page distillation of the basic GURPS rules. It covers the essentials of character creation, combat, success rolls, magic, adventuring, and game mastering for GURPS Fourth Edition.

The purpose of GURPS Lite is to help GMs bring new players into the game, without frightening them with the full GURPS Basic Set and a stack of worldbooks! With GURPS Lite, you can show your players just how simple GURPS really is.

This may or may not be a usefull document for you.

Suerte!
-Matt

[Walt Freitag]

Hi Lonoto,

If I'm interpreting this correctly:

The system I have come up with makes the skills and the base attribute or statistic numbers add-up to give a constant bonus. Then, based on how high your attributes are, they have a dice rating which determines how many dice are rolled, which allows variance (excuse me if this is the wrong mathematical term, I am taking math in French) to continue to play a part in the skill checks at higher levels.

... then your system is probably well-behaved in situations where the chance of success is high, such as when a skilled character is attempting a moderately difficult task under favorable conditions (where you still want a reasonable and internally-consistent chance of failure to be possible), but it will be less well-behaved when dealing with modifiers when the chance of success is already small, because then there are a small number of dice being rolled and any change to that number (such as the smallest possible modifier you can make) now makes a huge difference.

You've grasped some basic truths about resolution systems, such as that there is no fundamental difference between "base attributes" and "skill levels" and "modifiers;" they're all just numbers that affect the probability of success, and there's no reason not to treat them all the same. Your concerns are about consistency of modifiers, and about edge effects. Perhaps you've noticed that, for instance, in a d20 roll-over system, a +2 modifier has a small effect on a character's effectiveness against easy challenges and a large effect against very difficult challenges. (Perhaps you wish it could be the other way around, so that small advantages and small increments in effectiveness would be noticeable in the course of normal adventuring, but wouldn't erode the challenge of tougher obstacles so quickly?) The edge effects cause the chance of success to drop off too quickly when difficult circumstances and/or negative modifiers apply (a -1 modifier that changes an 18+ to succeed into a 19+ to succeed reduces the chance of success, not by 5% as an over-simplistic analysis would suggest, but by 33%) while favorable modifiers cause the chance of failure to drop off too quickly when circumstances already heavily favor success.

If you agree with this, then the following might be to your liking. I call this dice mechanism "Symmetry:"

Add up the character's base level, skill bonuses, and any situational bonuses to get a number. Subtract the difficulty of the challenge (which may be represented by an opposing character's base+skill+bonuses, or might be a simple difficulty factor). You'll get a net difficulty number such as 5 or 14 or 0 or -7.

Let's assume for the moment that the number is positive. Roll a d10, along with the number of d6's. So if the net difficulty is 0 you roll just the d10, and if the net difficulty is 3 you roll a d10 with three d6's.

A roll of 6 or higher on any die means success. (That means a 6 on any of the d6s, or a 6-10 on the d10). If the d10 is numbered 0-9, the 0 must be read as 10.

If the net difficulty is negative, you do the same thing, except that a roll of 6 or higher on any die means failure. So if the number is -5, you roll a d10 with five d6's, and you succeed only if every single one of them comes up 1-5.

This might sometimes require rolling of an inconvenient number of d6's, but you can add the following rules to prevent that: if there are more than four four d6's in the roll, you can remove them and replace them with another d10. You can replace two d10's, or eight d6's, with one d20. The interpretation of the roll remains the same: if any die comes up 6 or higher, you succeed (if the net difficulty number is zero or positive) or fail (if it's negative).

What this does is put all your modifying factors -- skill levels, requisite levels, bonuses, modifiers, or whatever is taken into account in your system -- on the same consistent number line. Modifiers have consistent effects with no edge effects. If the odds are against success, a +4 modifier in a character's favor will double the chance to succeed -- whether the +4 shifts the net number from -5 to -1, or from -15 to -11, or anywhere in between. If the odds are in favor of success, a +4 modifier will reduce the character's chance of failure by half -- whether the +4 shifts the net number from 0 to +4, or from +13 to +17, or anywhere in between. This property allows you to apply simple rules of thumb: 4 points of skill/requisite/modifier makes relevant tasks half as difficult. Adding 4 points to the difficulty rating of a challenge (the strength of a poison, the toughness of armor, the mental focus needed for a certain spell) makes it twice as challenging. A modifier of +1 or -1 represents, in every meaningful sense, a difference one fourth that significant. A net difficulty of zero is of course a 50-50 chance. (That means a net score of 4 is a 3 in 4 chance, a net score of 8 is a 7 in 8 chance, and so forth; while a net score of -4 is a 1 in 4 chance, a net score of -8 is a 1 in 8 chance, and so forth.)

So, have a character who's bow skill is 6 with a +3 arrow, and you want to set a target difficulty where he has about a 3 in 4 chance of making the shot? Then the difficulty should be 5.

The only thing you asked for that Symmetry doesn't do is give you specific cutoff points for guaranteed success or failure. Getting rid of the edge effects makes that impossible, since such cutoff points are edge effects. However, you can set any cutoff point you wish (say, at +/- 15) as close enough to guaranteed success or failure to be not worth rolling.

- Walt

[Lonoto]

OK, I think I understood the symmetry mechanism, but I still don't think that is what I was looking for. 1) It seems way to complicated for any of my players to understand, and 2) It is basically designed for the GM to decide how difficult they want it to be for the character then assign a proper difficulty, if I am reading it and your example correctly. I really don't like this idea, but that is just a personal preference. I'd rather have a scaled difficulty chart and the player chooses what they wanna do, I say how difficult it is and have them roll and compare stats. With your system I think it would work more like: Well, I want them to have a 3 in 4 chance to hit, so for him I will assign a 5. But this guy with more bow skill and a +6 arrow, I'll give him a higher difficulty and say the wind is blowing or something.

Anyway, as I said a few posts up, I don't think a fully universal system is a good goal for a high school student to be pursuing. However, as I am still interested in RPG design, and I have an idea for something I was going to do as a supplement to the universal system, I think I will do that. I have read Ron's post about creating a scene in your head of how you envision your game being played and stuff, so since this is my first real attempt and all, do you think it would be wise to post these things and get help from the people here from the get-go, or wait until I have quite a bit done? If I did post it, I'd start a new topic, right?

Lonoto

[Ron Edwards]

Hi Lonoto,

Long experience here has taught me, and many other folks, that starting with those questions I asked is one of the very best ways to get your game designed. So yeah, post away. I asked because I'm interested in your answers, and I'm not alone in that.

Best,
Ron