So, confessions time.
I've been ...
Rick on
100 Post Retrospective
Dicing on the Cake
Not too long ago, I wrote a bit about designing SystemX – primarily about my history with fucking up already established systems and what got me into gaming to begin with. Today I want to go a little deeper. Today I want to juggle talking about designing the math for my system… without giving too much away.
There seem to be two types of systems: those that are very simply and broadly cover everything, and those with tons of statistics that have a system for any occurrence. Keeping with an example I used last time, the Amber Diceless Roleplaying system is very simple. The game has four attributes that range for -25 to +25, and then rates special powers on a grading scale, topping off with exalted levels. Beyond that there are two more traits – I kid you not – Good Stuff and Bad Stuff. With this handful of stats, conflict is resolved without every rolling dice and the system is broad enough to cover an amazing range. It almost perfectly captures the heroic tension and action that Roger Zelazny presents in the Chronicles of Amber. The system disappears, encouraging the players to let their imagination lead the way.
On the flip side, you end up with games like (you had to see this coming) Exalted by White Wolf Games(or Publishing… or Studio.. or CCP… or White Wolf CCP LP… or White Tiger… or Tyler Perry Presents Some Guys that Write Words About Rolling Dice in ‘da Hood). The Storyteller system (you know, that one from before the Word of Darkness Shattering Kaboom) is fairly robust, and Exalted’s modified version (based loosely off the version for Trinity and Aberrant) seeks to bust the system open at the seams, jacking more power into that most players thought possible back in the day. Charms allow characters to twist and warp the math in a surprising amount of ways, and it requires a lot of fine tuning – almost like patching a video game. I have never seen a game with more errata than Exalted.
Then you have games like Rolemaster, which might as well be a fucking exercise in reading spreadsheets.
As I said in the earlier post, one of my big design points is “character customization” – I want players to have access to a lot of choices so that they can feel perfectly unique, both in personality and in build. Even if characters choose similar vocations, I want the choices they make to vary so that people don’t feel that creating two warriors for a group is going to drag them down. That lends to a statistic-heavy system.
I started this entire process by sitting down and creating a design document for the system, basically something that is a combination of mission statement, outline, and data dump. I am going to admit it, I am taking a huge cue from video games. Around seven years or so, I pursued a film degree at Full Sail University, and while their film degree may not be top notch, their video game design course is amazingly complex. For the first few months, I learned about video game design alongside film stuff, and two of my roommates were in the video game program. I am not saying I am in any way qualified to make games, but it gave me a launching point to approach designing this system from. I can’t imagine that designing a dice system is really all that different from designing a video game anyway, just the format is different and less coding.
I started with one primary item: the basic roll to resolve conflict. This is a very important choice for the system; it is what the rest of the game will hang off of. Think of it in terms of a blank wall that you drive one giant nail into. Everything else you do with your system is more like thumbtacks that you then tie to that nail with bits of string. If that one roll doesn’t cut it, the rest of the system will fall apart – you’ve either cracked the drywall or some other evocative metaphor. Every game has approached this differently, though many systems are very similar. Dungeons and Dragons has always done it by manipulating a d20 for a single target number, whether it was the -10AC of yesteryear or the modern +6 melee basic attack bonus. Older editions of Shadowrun run off tests where you try to achieve multiple successes against an ever-increasing target number.
This is where writing this post gets difficult – I don’t want to give too much away, but I want to talk about it at the same time. I decided my conflict resolution would come from combine a few statistics (four in total) against one of two things: a static target number or a dynamic one, depending on if a character is being passively or actively resisted.
Four statistics? Four? Cuatro? That’s right fuckers – but don’t worry, it isn’t like Rifts, I promise.
My system takes the things into account: natural ability, trained aptitude, situational modifiers, and other applicable modifiers. Throughout designing the system, I make it very clear where each type of statistic comes from and where it applies. Some have hard caps: I have a limit on how much a situational modifier can either add or detract from a roll. Some are more open ended: a lot of the special tweaks a character chooses falls into that nebulous “other modifiers” range and that can be dangerous. A math savvy player can find ways to really compound that section if they want, and as the kids these days say, rape the system. Part of the challenge is designing built-in modes of capping and limiting what can be done.
Finding ways to apply that basic conflict roll is the heart of design. Every power, ability, trait, and vaginal fart should give the characters new ways to do something with that conflict roll. The next post will go into more details about my bits and pieces, but this is sorta applicable to the math section. Every mechanical choice the character makes should either:
- Modify the conflict roll.
- Apply new results to a conflict roll.
- Create new ways to use a conflict roll.
- Create a dramatic unrolled event within the game.
This is just my loose list, but it’s what I’ve been working with and what I use to judge whether or not a power works. You decide which of those four things it does, then hang modifiers off it as you want. You can use this also to set up caps – if you design the math with this in mind, keeping it “balanced” becomes a little bit easier.
Another big thing I am keeping a close eye on is multiples. I decided I want to keep to d6’s (you can buy these fuckers without making a special trip), and with that decision in mind all of my statistics that give a hard bonus add +1, +2, or +3. When I am designing things, I can keep that variance in mind – most single sources will add only half a die’s worth of result. As a character becomes more experienced, I can expect more “half a die” worth of hard bonuses to be taken into account (so, an experienced character can be assumed to have a constant +6 bonus in combat, where a really experienced might have +18). With that math in mind we hit one of the trickiest parts of system design: appropriate challenges.
This just seems to be the hardest damn thing to nail. I think the closest I’ve ever seen a game come to an accurate metric for determining challenge is with the current edition of Dungeons and Dragons. Exalted attempted some threat system that really didn’t work, DnD 3rd Edition challenge ratings were really just spit balls, and Shadowrun doesn’t even both trying – suck it up, chummer. So one of the things I am trying to do (in an open, non-level system) is use those half-die units to achieve an appropriate danger level from threats, both combat-related and not (mostly combat related, because combat is the most explosive and dynamic form of applying the conflict roll).
Alright, I’ve talked a little bit about other systems and a little about how I approached the math. If you have any thoughts or anything to add, please comment away. I am always looking for new ways to approach designing a system. This will wrap up next time as I go into the bits and pieces of system design.
Also! Like dice? You must check out “The Bones: Us and Our Dice” when it hits. Here is a link to the page on Gameplaywright about this forthcoming tome of awesome.
I find it interesting the bias you put into your descriptions above. There is no question which type of system you prefer. As a big fan of Rolemaster and d20, let me just say that there are merits on that side of the fence, too.
Out of curiosity, how does your description of your conflict resolution differ from, well, the standard definition? Pretty much every conflict resolution system I can think of combines those factors. Well, okay, a lot of them don’t split the aptitude and training out. But stat + skill + special + situational is pretty standard. The differences mostly come in how many dice you roll, how those modifiers get applied, how you compare the result to the target number, and then how the success/failure is applied.
One of the really interesting things being done with core mechanics over the last few years is finding ways to squeeze more and more information out of a roll. The king of this is ORE (One Roll Engine, as seen in Godlike). You get a big pool of dice, and roll them. You then try to make matching sets. The number on the die that you match is the “height” and the number of dice in the set is the “width.” This gives you two numbers to play with from one roll, allowing you to read out things like attack and damage simultaneously. Don’t Rest Your Head has a neat mechanic where you can choose to risk your sanity to add to a roll. You add a special colored die to the result. But, if that dies comes up 1 (or maybe 6, I’m confusing it with my own rip-off now), you lose a point of sanity, even if the roll was a success. Unknown Armies gets a bit of it by giving “cherries” for rolling doubles (it uses a percentile system).
Being able to read multiple pieces of information out of a die roll also means that your system can have multiple ways to modify a die roll. In Spycraft (a d20 mod), your basic d20 roll has simple modifiers, threat and error ranges (numbers on either end of the possible die results that can be activated for critical results), and the ability to take 10 for very reliable results. When you pick a gun from the Modern Arms Guide, it has ways to alter all of these factors (in addition to modifying the damage roll). They have a dizzying array of feats that all interlock with this basic roll in really neat ways.
I wish you much luck with the challenge rating concept. That is a very tough nut to crack. Especially because the effectiveness of a party and the effectiveness of a challenge is often more about how they’re played than simply how the numbers add up.
No, no bias actually – just what I was going for in this system prefers a more open system. Level-based systems are totally fitting for several games (I can’t imagine D&D without it). I decided I wanted something with more direct player control, sacrificing the ease and direct progression for flexibility. This post was written with that decision already made – believe me, I went through several iterations of how to tackle it before I settled on one.
One exception: Rolemaster. That system and I have never gotten along.
My description of my conflict resolution was less about “this is new” as opposed to spelling out why I was doing it like I did. It was showing, from what I see using the way I’ve set up my dice, how I want to tackle that particular mode. For me it’s roll a certain amount of dice for a result (aptitude), add reliable amount from training (skill), apply effects of the environments (sit mods), and apply any remaining modifiers (other stuff). If you look at other systems, they tackle those factors but do them in different ways, sometimes combining them all into one set roll. D20 is probably the best example of that – and before I am making it sound simple or elementary – it does it really well. White Wolf typically uses varying dice with increasing target numbers or success thresholds. Marvel Superheroes broke it all down to percentages, Palladium ripped on old D&D with “innovations” (whether you consider that a true statement or a joke is up to your own taste).
The point I am going through here is that there is probably nothing new under the sun when looking at a design standpoint. Every game out now builds on the games that came before it – even Chainmail was built using the ideas from other types of games, all the way back to when Captain Caveman decided to resolve conflict abstractly. And with all due respect to Gygax and Arneson, the old D&D engine was solid but faulty, yet it set the launching point for systems (including future editions of D&D) to revise, adapt, or break apart – sometimes well and sometimes badly.
I completely agree with how neat some of the more “out there” systems are. I like them, I really do, but primarily from an intellectual point. When the dice get really funky, it’s been my experience that they end up detracting from gameplay. What I want to do is make a solid frame with mechanics that are easy to grasp from which to hang amazing stories and gameplay. That was one of my problems with the old Deadlands game – I adored it on so many levels, but the system made it really bulky and (after the gimick-cool wore off) felt really restrictive.
So don’t get me wrong, I am absolutely not trying to reinvent the wheel here. Part of this blogging attempt is just sharing my thoughts on creating this system and what I’ve gone through mentally.
On to challenges – I am really struggling to find a happy medium between combat-heavy games and more social games. I think what I will end up doing is designing it with the idea that every character is going to have a medium range of combat scores, hopefully account for more social characters or combat-heavy characters. It won’t be perfect, but hopefully it will accurate enough that individual groups can tighten the precision based on their group composition and what style of game they are going to be playing.
Thanks for the thoughts
Today on, “I Read That As”: The Dicking on the Cake.
The funny part is that I wasn’t surprised when I read that. It seemed completely in-character for you.
Yeah, I can see that