Vapourware Alex's P&P System

[Alex]

My last attempt at writing my own RPG didn't pan out, but it did yield a few interesting (to me, at any rate) ideas, and I am hoping to put them into practice with this new system. However, designing is hard work, and I have to keep second guessing myself constantly, and still, lots of obvious flaws sometimes just pass me by. So, I thought maybe you guys would be interested in helping me out. I wanted to start with how the system handles rolls, so I won't go into much detail now. What I have in mind is something like a a mixture of old Warhammer with DCC RPG. Like in DCC RPG, PCs start off as lowly adventurers and carve their own place in pulpy fantasy world. But eventually, more than themselves, they start counting on others. Thy raise armies, form churches, mess with politics and may even mess with magics that can change the whole world. In fact, the envisioned "endgame" (though it is really up to the GM and the group to determine what that will be like) would probably feel like Master of Magic at times.

Anyway I want to start with the way dice are rolled. Throughout the years, some rather interesting and unusual systems have appeared. Deadlands, for instance, used playing cards instead of dice. In fact, it actually used cards as cards in spellcasting, that is, deciding how well a spell was cast was based on poker hands. Dogs in the Vineyard uses a system where you roll a dice pool, and then match values in order to solve any conflict. Being unable to match an opponent's attacks with 2 dice or less mean that you get "hurt" somehow in the conflict. Aces & Eights use a card drawing mechanic coupled with a "gun clock", a transparent overlay that is put over the shadow of your target, to determine where exactly your bullets go! Everway uses special "tarot" cards that determine hot only if you succeed or fail, but what tone the resolution takes. Amber doesn't even use randomness. Instead, the GM, aided by the game's guidelines and the character's attributes and abilities, decide what happens at any given circumstance (those guidelines are pretty clear on some stuff, though).

I am mentioning all this because these special systems seem to me to fall into two categories. Most of the stuff I mentioned help set the theme and the mood in the game. The card drawing thing in Deadlands help set the mood for the over the top western affair it is, while in Aces & Eights, the gun clock helped make the combat deadly and "realistic". Even Amber's system helped make the game feel more "storey", for the lack of a better term. All this is rather nice, but I don't want to tie in the basic system of my game with anything specific like that. Maybe when a spell uses runes, the players throw actual runes and the result are accounted for somehow, or maybe when players use their luck to influence a roll (supposing there is such a thing in the system).
The other category, represented by the Dogs in the Vineyard example, is using the dice sysem as an abstract game element. Numenera is another example of this. He, the mechanics exist not to represent something in game, but something extraneous, a layer of strategy added in to make the game, well, gamey. I don't really care about this, at least for this game, though. While I think Dogs' system is pretty interesting, I want all the reasoning and strategizing to occur on the game, not the meta game, level.

Thus, what I want out of the system is that it represents the situation surrounding an action well. As I see, I have two basic options if I am using dice. either I have a dice pool system like Shadowrun, of a fixed dice system. While dice pools are nice, I have right now an idea for a system a bit more "quirky", perhaps more suited to the pulpy fantasy kind of game I have in mind. In its base, it would work similar to D20's system. You have a basic attribute (or a derived value of some sort). You add a die (or dice) roll to it and compare it with a rising difficulty scale.

Now, as long as we aren't trying to do anything "fancy" like I mentioned above, dice rolling systems may seem to boil down to the same thing: giving you sane probabilities for the action at hand. But I think a good system could try helping this process by breaking down how different factors affect itself. Thus, I am thinking of braking the system in 5 factors: 1 - Situational Modifiers, 2 - Difficulty, 3 - Critical Range, 4 - Dice Range, 5 - Certainty.

Situational modifiers are the most basic way to change a roll. Stuff like getting a +1 while shooting a bow because the wind in on your back, or a -2 because it is dark and so on. This is really old news in RPGs, but there is a wart here that I think should be made clear. Games where lots of little things can affect a roll can see PCs going to great lengths to make every little thing before the roll perfect, as to maximize their chances. And that isn't a bad thing, necessarily, but if you use brushing your teeth, getting a snazzy haircut, waring perfume, buying a imported suit, imported pants, imported bow tie and what not to get a bigger bonus to your charm roll than your charisma score, it gets pretty silly. Because of that, I think a good solution is to have the scale on which rolls are based be "exponential". I use quotes because exponential as a concept doesn't make much sense when an attribute or skill or whatnot isn't really measuring anything. What does it mean, for instance, to say person A has twice the dexterity of person B? But at any rate, the idea is that you don't just stack modifiers. Instead, the GM should consider how the modifiers work together overall. Ten things worth a "+1" should probably, taken together, be worth a +3, maybe a +2 if they have intersections in their functionality.

Difficulty is another way players can try to shift actions in their favor, and the GM can make the gameworld seem more real. Basically, the idea is that the world isn't a zero sum game. There are easier and harder ways to do things. Trying to climb a sheer wall without equipment is really hard, while having good equipment can reduce difficulty dramatically (note that this is not the same as a bonus, equipment is really important in the case of a sheer wall, but a mountain where rocks work as handholds wouldn't be as affected by the lack of it). Jumping over the wall is probably almost impossible, unless your physiology isn't quite human. On the other hand, bringing the wall down instead will be a different difficulty altogether. It might be easier or harder, depending on what it is made and what you are using to hit it.

Critical Range is pretty much how it was used in 3E D&D, although it doesn't apply only to attack rolls. Usually, rolls are made on a D20, and both the fumble and critical success range is 1, meaning a 1 is a fumble and a 20 is a critical success. If both ranges were increased by 1, fumbles would happen on a 1 or 2 and criticals on a 19 or a 20 (that is 20% of the rolls would be either very good or very bad. Different from 3E, however, the range isn't affected by weapons, or even mundane things. Usually, increased ranges will be the effect of magic (like a curse or a potion of heroism) or be an ability possessed by larger than life figures (in other words, high level characters, but more on that later).

Dice range is something I am taking from DCC RPG. Basically, dice range is potential. Usually, the dice range for actions is 20 (meaning you roll a d20), but it can be increased or decreased by certain factors. Having a low dice range doesn't necessarilly mean a character is bad at something. It just meas that the ability is more limited. Usually dice range is increased by high level abilities and magic, much like DCC used them. Reduced dice range is usually a sign of a character who is attempting something he has little or no training for.

Certainty determines how many dice are rolled in an action. In a certainty 1 action, a d20 is rolled (supposing the range is 20). Certainty 2, however, means that a d10 and a d9 are added together (or 2D10 if you can't be bothered, the difference is minimal really), while certainty 3 uses 3D6. The idea here is that some actions can be a whole lot less certain than others, like when magic is involved, or in combat, where things can get pretty hairy. Climbing a mountain while well equipped with maps and safety gear might be a certainty 3 action, or maybe 4 if you are also familiar with the terrain. While running away naked into it would be a certainty 1 action (or maybe even less).

The thing about certainty is that it changes a lot how the other aspects influence the roll. The greater the certainty, the more significant the attributes and bonuses are. That is, having a "+3" to a roll while rolling 3D6 is a lot more meaningful (assuming the roll was challenging to begin with) than a "+3" to a d20 roll. In fact, if you needed to roll a 12 or more on a 3D6, you would have 37.5% chance of succeeding. Now needing only a 9, you have 74% chance of succeeding, almost doubling your chances! On a D20, a flat 15% more chance would be added, not matter what. Greater certainty also mean that it is much less likely to roll a critical result. If players want to attempt something that is within their normal capacity, high certainty helps them, but once they need a "miracle", they want to go for more chaotic actions. But these can, just as well, bring great ruin instead.

Well, this is it for now. What do you guys think? Do you think these five aspects would cover well the kinds of crazy stuff people do in these games? Do you think any kind of category should be added or removed? Are any of my premises flawed? Thanks in advance for everyone who decides to help.

 

[tuluse]

I like the certainty mechanic as a concept. I'd probably need to see how it does in practice as it seems like it could be hard to balance.

I think you might have over thought situational modifiers and difficulty. All the problems you mentioned seem like they would be handled by competent GMing. A lot of examples in and campaigns for DnD already include this stuff. The cliff face gives a -10, unless they have climbing gear, then it's a -2 type stuff. I don't really see a need to formalize this stuff. It can be handled on a case-by-case basis.

I'd like to see some examples on how you expect range and certainty to be utilized.

 

[Alex]

Thanks for reading all that, Tuluse!

I agree the modifiers thing doesn't really need to be formalized. A couple of paragraphs in the GMing section about how to deal with them should be enough. Still, I think it is important to figure out this stuff when designing a game. Adding modifier on top of modifier works really well in Shadowrun. But I think it only works there because you have a dicepool mechanic. The dicepools add diminishing results in a very organic way. If you are rolling 30 dice, rolling 35 instead is of course better, but not as good as rolling 6 instead of one. When you work with ranges of results, as the system proposed does, then either very big difficulty classes start becoming indistinct, or you end up with strange stuff, like I mentioned in the original post.

About dice ranges, I am thinking of maybe having a "mastery" kind of deal. People who are untrained in a skill would use a d10, maybe. This would increase as they learned more about the subject, until at professional level they would use d20s. Further training would eventually allow people to get to d30s, but that would take a whole lot of time, something not all adventurers have. Then again, at some time years will fly like turns, if I manage to piece this right.

In particular, non-spellcastng classes (although my concept of classes fr this game is a bit fluid, but I will get to this only later) would use d10s i they tried to use magic. Of course, they would need to find someone to teach them magic in first place before they could do that. This mechanic is actually lifted directly from DCC RPG, although I think it could be used in more ways than they use there. In DCC, a higher level character started getting extra action dice, meaning they would be able to do more actions per turn. But these dice are smaller than d20 (at least early on), meaning these action have a greater chance of failure.

One god example of certainty is, again, magic. I want a magic system similar to the one in DCC, and that means that spells can go very wrong (and very right to'o). Because o that, magic is usualy a certainty 1 affair, if not less! Clerical magic, however, is filtered by an extraplanar entity. In game terms, this means that their spellcasting rolls have a certainty based on their sanding with their God. Also intersting is that having a turbulent relationship with their deity can actually bring about greater miracles! But then again, i is usually in times of doubt that this stuff happens, at least traditionally. Meanwhile, magic users don't really have restrictions to their spell list, as they don't have to please some other entity with their magic, but they need to find other ways to tame their powers, or live with the consequences. Alchemy is an alternative: alchemical spells are inherently 1 step more certain than normal. But they are also less "flashy" and powerful, although being portable and storable kinda makes up for that.

 

[Misconnected]

What do you guys think?

I think a half page of keyword'ish setting info, and another half page of concrete examples of dice-based event resolution is kind of mandatory if you want more than ultra-vague "good luck" type feedback.

As for dice-based event resolution, I'm a big fan of avoiding it. Dice rolling can very easily become the game, especially if complex rolls (many dice, many types of dice, etc.) are frequent.

I'm also a big fan of transparent probabilities and Black Jack type rolls. For example, using D% rolls where the degree of success or failure depends on how near the target number the rolled result is.

And finally, I'm very much a fan of modifier hunting. I like rules to be crunchy enough to account for many factors, because it lets players exploit many factors, which boils down to the system rewarding situational awareness, and creative and intelligent play.

 

[Alex]

I was actually hoping people would answer the thread based more on a desire to discuss the theory behind RPGs. I don't really have anything in mind setting based. Ideally, you would be able to use a D&D setting if you wanted, or maybe just make something based on fantasy works. The only issue is that, as the game is planned, you will probably need some details, at least eventually. Maps, political webs, descriptions of military forces, etc.

One of the reasons I don't like the D% is because it is way too random (at least how it is normally implemented). I am more of a fan of the bell curve of stuff like GURPS, but I think the crazy randomness of a single die as they happen in games like Dark Heresy, can be fun and interesting in certain situations. Which is why I came up with the whole certainty thing.

I agree with you that modifier hunting is an important part, however I think coming up with a really good idea should be much more important than trying to come up with a way to use 10 or 20 pre-defined modifiers. Still, I agree with your that games providing good examples of this stuff can help the GM a bunch. Though, I've seen the opposite effect happen (ask Darth Roxor about Shadowrun 3e).

At any rate, thanks for the feedback. I hope that as I begin to discuss more concepts, this stuff will become clearer, and maybe more people will become interested.

 

[Telengard]

Do you think these five aspects would cover well the kinds of crazy stuff people do in these games? Do you think any kind of category should be added or removed? Are any of my premises flawed? Thanks in advance for everyone who decides to help.

For the basic question of does it satisfy everything - there's not really enough info to go on, but you can easily establish that yourself. Tasks can be arranged into four basic categories - simple, complex, competing, and contested. If your system covers all those, then you've got your main bases covered. Then you go to combat. If you've got a system where people can comfortably melee attack, ranged attack, cast magic, and defend, then you're solid there. After that, it's more a matter of fine tuning.

The categories cover the needs of the game. As for adding or removing categories, that's more of a stylistic thing. Does the pace and flow of the dice rolls that result mesh well with the themes and pace of the overall game? With only a few tidbits of info, outsiders can't really say. But, pick up a few dice and roll your system, and if it feels right, then you're on the right track.

As for the dice system itself. If I am understanding it aright, then your core concept is good, but it needs some refinement. Because - probabilities and statistics - when one needs to roll high, it is better to roll a single die than multiple dice, because the results of multiple dice bell curve to the middle. For instance, needing to roll a 12, that's your 37% chance of success. While a d20 would be 40%. And needing to roll an 18, on 3d6 your odds are half a percent, while d20 is 12%.

You can tweak the dice rolled and limit the range of difficulties to fix this issue. Or tackle it from a different direction. Such as, Certainty grants one free dice per certainty level, keep the best roll. Or go completely off the wall and have the number of dice that need to be rolled be the Certainty minus the character's expertise rank (which is separate from skill level), with all dice rolled needing to be a success.

 

[Misconnected]

I was actually hoping people would answer the thread based more on a desire to discuss the theory behind RPGs. I don't really have anything in mind setting based. Ideally, you would be able to use a D&D setting if you wanted, or maybe just make something based on fantasy works. The only issue is that, as the game is planned, you will probably need some details, at least eventually. Maps, political webs, descriptions of military forces, etc.

Hmm.... The problem with that approach is that the mechanics of your game can't avoid heavily influencing the tone and style of the fiction. And indeed, RPG mechanics are at their best when they strongly reinforce the fiction. To give an example, it's pretty much impossible to play an Indiana Jones style game using WFRP1e RAW, because almost everything that makes an Indiana Jones style protagonist is very likely to end the career of a WFRP1e character.

In the absence of setting info, I'd suggest you grab a pen & some paper, and your favourite mouldy RPG tome (or tomes), and go through each chapter, jotting down keywords on how your homebrew should handle things. Reign and Mongoose Traveller might be worth a look, by the way. Both try to tackle your late game goals, and in very different ways.

One of the reasons I don't like the D% is because it is way too random (at least how it is normally implemented). I am more of a fan of the bell curve of stuff like GURPS, but I think the crazy randomness of a single die as they happen in games like Dark Heresy, can be fun and interesting in certain situations. Which is why I came up with the whole certainty thing.

My issue with your Certainty concept is that it sounds extremely opaque, and involves far too many dice for my liking.

As for your criticism of % based systems, I completely agree with you. That's why I'm a big fan of Black Jack D% rolls. That give a curvy/spiky probability distribution just like, say, 3d6. The differences are that the curve/spike is defined by Actor ability and the possible resolutions instead of the dice, and that the chances of the various possible resolutions are immediately obvious to everyone at the table. To me it's basically a more human-friendly 3d6.

I agree with you that modifier hunting is an important part, however I think coming up with a really good idea should be much more important than trying to come up with a way to use 10 or 20 pre-defined modifiers. Still, I agree with your that games providing good examples of this stuff can help the GM a bunch. Though, I've seen the opposite effect happen (ask Darth Roxor about Shadowrun 3e).

I should say that those preferences are to a very high degree coloured by my love of fighting skirmish combats with miniatures and terrain. For something like a grid based crawl, I actually prefer a much simpler boardgame, such as Descent: Journeys in The Dark. Those games nail the teamwork, the pace, and don't have the usual RPG issues with Actors with gnat-like lifespans.

More, even RPGs that are meant for fairly elaborate tabletop skirmish battles absolutely needs some sort of narrative combat system, because combat encounters are more frequently a tool the GM uses to progress the fiction, than they are meaningful fiction in their own right. Basically, you need the option to have two guys kick in the door, guns blazing, without following it up with a 2+ hour miniature skirmish battle.

 

[Alex]

For the basic question of does it satisfy everything - there's not really enough info to go on, but you can easily establish that yourself. Tasks can be arranged into four basic categories - simple, complex, competing, and contested. If your system covers all those, then you've got your main bases covered. Then you go to combat. If you've got a system where people can comfortably melee attack, ranged attack, cast magic, and defend, then you're solid there. After that, it's more a matter of fine tuning.

Yeah, it was a dumb question, or a dumb way to phrase it anyway. What I really meant to ask is if the way I was treating basic rolls left some wide gap, if there is anything obvious missing. By the way, I will get to contested rolls next.

The categories cover the needs of the game. As for adding or removing categories, that's more of a stylistic thing. Does the pace and flow of the dice rolls that result mesh well with the themes and pace of the overall game? With only a few tidbits of info, outsiders can't really say. But, pick up a few dice and roll your system, and if it feels right, then you're on the right track.

Yeah, that kind of thing is probably better to fine tune once I have some actual systems in place.

As for the dice system itself. If I am understanding it aright, then your core concept is good, but it needs some refinement. Because - probabilities and statistics - when one needs to roll high, it is better to roll a single die than multiple dice, because the results of multiple dice bell curve to the middle. For instance, needing to roll a 12, that's your 37% chance of success. While a d20 would be 40%. And needing to roll an 18, on 3d6 your odds are half a percent, while d20 is 12%.

Sure, more dice mean you reach smaller numbers with the dice. But it also mean it is more unlikely the roll will deviate much from the mean. That means that, unless you need a "miracle", rolling more dice is safer. Specially since fumbles are also much less likely with many dice. Besides, you don't get to choose the amount of dice directly. If you want to reduce the amount of certainty of a roll, you need to take some concrete risk, which can add penalties or complications to the roll. By the way, I am expecting to use stuff similar to Rolemaster for criticals, so fumbles are really bad things.

For example, suppose we are talking about tracking. Usually, skill rolls have a base certainty 3 (as opposed to combat and magic (and whatever else might need to be inherently chaotic)). So, if you want to actually reduce your certainty, you need to come up with a way to make your own situation bit more desperate, or at least... well... uncertain. A good tracker might know ways to do this that aren't completely ridiculous. For instance, maybe he could use a dowsing rod. But using dowsing rods might be a little riskier in that they always lead you to something, even if you fail...

You can tweak the dice rolled and limit the range of difficulties to fix this issue. Or tackle it from a different direction. Such as, Certainty grants one free dice per certainty level, keep the best roll. Or go completely off the wall and have the number of dice that need to be rolled be the Certainty minus the character's expertise rank (which is separate from skill level), with all dice rolled needing to be a success.

Actually, I was thinking of coming back to the concept of positive and negative certainty, which would basically add an extra dice where you would roll extra dice but get the best result. This kind of thing would be reserved to magic or to situations where the odds are really stacked in your favor.

Hmm.... The problem with that approach is that the mechanics of your game can't avoid heavily influencing the tone and style of the fiction. And indeed, RPG mechanics are at their best when they strongly reinforce the fiction. To give an example, it's pretty much impossible to play an Indiana Jones style game using WFRP1e RAW, because almost everything that makes an Indiana Jones style protagonist is very likely to end the career of a WFRP1e character.

You have a good point. I will try to explain what kind of tone I am going for in one of my next posts. But the basic idea is very similar to Old D&D, with characters starting as rather fragile normal people, and eventually reaching absurd levels of power in some crazy pulpy fantasy world.

In the absence of setting info, I'd suggest you grab a pen & some paper, and your favourite mouldy RPG tome (or tomes), and go through each chapter, jotting down keywords on how your homebrew should handle things. Reign and Mongoose Traveller might be worth a look, by the way. Both try to tackle your late game goals, and in very different ways.

Mongoose traveller does that? I will have to check it out, thanks for the info!

My issue with your Certainty concept is that it sounds extremely opaque, and involves far too many dice for my liking.

Fair enough. Hopefully I can make what certainty should mean more clear later on.

I should say that those preferences are to a very high degree coloured by my love of fighting skirmish combats with miniatures and terrain. For something like a grid based crawl, I actually prefer a much simpler boardgame, such as Descent: Journeys in The Dark. Those games nail the teamwork, the pace, and don't have the usual RPG issues with Actors with gnat-like lifespans.

Early on, your characters die easily. To counteract this, players get to control more than one character. In fact, they get to control a whole lot later on.

More, even RPGs that are meant for fairly elaborate tabletop skirmish battles absolutely needs some sort of narrative combat system, because combat encounters are more frequently a tool the GM uses to progress the fiction, than they are meaningful fiction in their own right. Basically, you need the option to have two guys kick in the door, guns blazing, without following it up with a 2+ hour miniature skirmish battle.

This game takes an approach more like old D&D. The monsters that are in the dungeon are there. If the players decide to tackle them with combat, misdirection or whatever, it is up to them. The GM doesn't really use combat for anything, he just brings it up as a result of how the fiction evolves.

 

[Alex]

Ok, let's get to at least some specifics. The system mentioned up to now obviously is assuming some kind of difficulty scale. Now, how difficulty scales work is key to make this game really fly, but static difficulties aren't everything. Telengard mentioned four possible task categories, and I believe what I mentioned up to now would fall into what he calls simple tasks. Let's them move on to contested actions. The simplest way to go about this is, of course, simply rolling the dice and having whoever rolled the highest win. However, I don't think this kind of system is really appropriate for some, or even most, situations! I mean, sure, if you are in a bar arm-wrestling with some dude, it is probably about whoever is the strongest. But if you are trying to sneak silently past the guard, it isn't about matching your sneak skill to the guard's hearing abilities. It is about being silent enough.

Still, let's take the arm-wrestling as a starting point. At first glance, the basic roll comparison here falls like a glove. Arm wrestling is just about whoe is strongest, so, rolling a d20, or 3d6 as the case may be, and adding it to your strength to see who has the highest total seems close enough. Except that, even if we are using a 3d6 roll, there is still a whole lot of randomness here. Let's assume we are using a scale similar to D&D, that is, attributes are more or less in the same range as the rolled dice (at least for human values). Imagine instead you are in a dungeon. Suddenly, you hear a noise above you and a portcullis start to drop from the passage you just crossed. You quickly try to turn around and hold it. There is a whole lot that can go wrong. You could not manage to hold it only with an arm, you might not be able to use your back or legs, you might not manage to get a good hold on it, etc. In this kind of situation, it makes sense the random factor is as big as the attribute factor. But when people are arm-wrestling, they usually make sure everything is correctly in place. At worst, someone might pull a muscle and lose because of that, but usually, whoever is the strongest wins these things.

At first, this might look like a prime example of where using certainty might fit in. But the certainty mechanic isn't too good here. You would need to roll too many small dice for it to really affect the roll, the critical range thing would be all wonky, and besides, one can imagine someone being desperate in a arm-wrestling match. Someone might use all his willpower and luck to try to overcome a stronger adversary. The issue here isn't that the participants are being careful about arm-wrestling, but that the controlled conditions leave less room for error. What would fit here is to use the idea of a smaller potential I mentioned in the other post, except that after I posted it, I realized there were some issues with it. If I use smaller dice to represent lack of familiarity with the action, and bigger ones to represent mastery, I can fit together the critical ranges nicely with this too. By multiplying the critical success range by a number to fit the new dice, while leaving the fumble range untouched, we can make smaller die much more disaster prone, while bigger ones are more critical prone! But this doesn't fit so well if we have situations like this, where we would want the fumble rating to be smaller if anything instead. It is easy enough to solve this problem, however, by multiplying the random factor by a fraction. For instance, if we use 1/4 here, we still have a range of 20 results in a d20, but they would only go between 0.25 to 5, much smaller than the range of human attributes (3~18).

The system above is pretty good, I think, if you need to judge how well people do in a fair competition. If we had two people running on a track, using a rather small fraction could mean the difference between the first and second places to be a few tenths of seconds, much like in a real race! But suppose instead that we have a situation that is inherently not fair. Instead of a track race, we might have two sides trying to see who gets first to the buried treasure, one driving a vehicle, the other riding a magic horse of some kind (which is so fast it isn't completely outmatched by the vehicle). Now the person piloting the vehicle and the horse (and maybe its rider too) actually have very different deals. The machine has more or less static specifications. One could be really detailed or really simple here. We could just consider a "vehicle rating" to determine how fast it is. Or we could somehow take into account how well the cart handles on curves, how well it accelerates, what is its safe maxium speed, what is its unsafe maximum speed, what kind of path it is crossing, etc. But, in the end, the piloting test is really about efficiency. A really good pilot might be able to cross a tortuous and dangerous circuit almost as quick as if it was a straight line. But he can't make the cart go faster than its maximum parameters would dictate. In fact, in a easy course, the difference between a regular and a superb pilot might be mere seconds.

The horse on the other hand, has a very different deal. The horse should probably have some kind of speed score (which might be a value based on its race, dexterity, constitutions and whatever else the game has), which, together with the dice, gives a range of possible speeds. Say the scale we use is geometric, we could have a speed of 20 mean 48km/h, with the speed doubling for every 4 points. By messing with the range as above (that is, multiplying it by a fraction), using a speed score of 20 and a 1/5 as the fraction, the horse would be able to run at between 48 and 96 km/h, depending on how inspired the roll is. But even so, a critical could throw it outside the loop. The rider, if he figures in this roll at all (he certainly would on a normal horse, but magical horse, so...) could perhaps help the horse's speed with his own riding rolls. This might seem silly, but it is appropriate for this kind of game I think. Conflict between law and chaos is a common theme in sword and sorcery books, and here the unbound speed of the horse can well represent chaos, while the technological vehicle represents order.

At any rate, this problem has one extra complication. It isn't just a question of multiplying the dice by a factor, but that one roll increases the speed exponentially, while the other is an inverse exponential! This difference in scales is yet another way the system will provide differentiation between rolls on the dice level. It isn't just an issue of rolling bigger, but being more effective. Even if the cart's pilot managed to roll an absurdly high number, it doesn't matter one iota if the maximum speed of the cart is only 40 km/h. By the way, I know this probably sounds a little too detail driven. Usually, I think the GM won't need to think about any of that. These are just tools the game will provide him so he can make a context different based not on abstract numbers, but on the details of the situation, on what the different sides are doing.

Let's take a different example, that of trying to lie to someone. It is easy to think of this as rolling your bluff skill (or whatever it is called) higher than the person's willpower, or intelligence, or some other stat. But if you do just that, you miss the core of the issue, namely, the actual lie that is being told. It isn't the same thing to try to convince someone that you are afraid and running away from something, that you have a good business proposal and that cowpie actually tastes like fruit. We could solve this easily enough by simply adding a modifier. The first is reasonably easy, so maybe you have a bonus (specially if you roleplay being afraid well). The second might depend on how bad the business offer actually is, or how obviously bad it is, at least, while the third would be a big penalty. But this kinds of make having a big bluffing skill all that is important. You can convince anyone to eat shit if you have a big enough score.

So, instead of a direct contest, we can have a system where each roll decides one part of the issue. First, if you want to tell a lie, you have to roll as to how convincing you are. If your target is scrutinizing you, you would need to beat him in a contest (and he might get bonuses if he doesn't trust you, in fact, you might need to go through this kind of test even if you were telling the truth, in that case). But if he isn't, all you have to beat is whatever perception attribute he has. However, all this tells us is whether you are convincing. In the first example, of running away, this is probably all you needed. But in the second, your target might not believe you are trying to fool him, but that doesn't mean he is going to give you money. In the third, all you probably did is convince your target that you are either crazy or retarded. In these two cases, you still need your target to actually do what you want. That is usually based on some roll of your target, of trying to resist the obvious bait in the lie, usually intelligence, though it varies. In the second example, it might be based on his business savvy or whatever. The difficulty of this roll is actually determined by what lie you were telling. If your business proposal is really well made and the flaws really well concealed, then the difficulty will be high. This does mean you usually can't get non-stupid NPCs to eat shit, though, no matter how high your bluff skill is. You can, however, try to make the roll more difficulty for your foe. If you put emphasis on how surprised you were when you found out, and maybe limit it to a specific cow, it might be easier to get someone to try the cowpie. If you spend some time getting to know your target, and maybe acting so he starts to like you, it will be easier to get him to overlook details in the proposal. But all this requires planning by the part of the PCs, rather than simply dice rolling.

As a last example, consider the thief trying to sneak past a guard. Usually, this is handled simply as a question of the thief's sneak skill vs the guard's perception (or whatever). However, that is not quite how sneaking works. Suppose for this scenario that the thief used a magical item called water arrow to put of a torch nearby, making the room darker and also making the guard go check the torch. The guard has his back turned to the thief, so pretty much only sound matters in this situation. Now the thing about noise is that its scale is kind of an inverse exponential, with higher skills getting closer and closer to a minimal amount of sound (that being the minimal the body can make while moving, such as its heartbeat and so on). A good thief and a really good thief are pretty much the same on good conditions. Maybe if the floor was noisy, or the guard had some super auditory ability. But against a normal guard, as long as the thief rolls well enough, the guard might not even get a roll. Now, it would be always possible for the guard to hear the thief, even if he is as silent as a cat. If you roll a critical hit, but not yet enough to achieve a success, you get to roll again and add the result. This way, it is possible for not only the guard to hear the thief, but for the guard in the other room as well. But it is up to the GM whether such low chances merit rolling in first place, and if the guard needed a critical and more to hear the thief, he would be well justified in just letting the thief go by. On the other hand, if the thief rolled badly, but not too badly, he might make enough noise to make the guard suspicious, but not immediately aware of what is going on.

However, it isn't always the guard's ears that the thief has to face. A smart guard, or maybe one that simply has been there before, might well decide to keep his guard instead of checking the arrow. Worse yet, the thief might go check the torch to play along, only to turn back on the last second with his crossbow in his hands. Or maybe, the GM might decide to test the luck of the mansion's owner. If he is lucky, the guard might remember something and turn around right as the thief was trying to sneak by.. The point here is that the most obvious compared roll isn't always the one that should be used. Different choices here of how the GM will "defend" have different results and roll differently. Still, the GM should use the one defense he thinks applies the most, and roll with it. That is, it would be overkill to have the guard roll intelligence, then luck and the perception. Not only would this skew the probabilities, it would make dice rolling pretty boring. If a roll is risky because of several reasons, it is usually a good idea to roll all of those into modifiers for a single roll, unless the results of these risks are so different that rolling the dice more than once is actually worth it.

At any rate, my point in this post is that it is an important aspect of this game to move away from simple value comparison, at least when the results are interesting. The point of havin the defense against lies depend on the lie itself isn't to make it more realistic, but to make the core of the game more about creativity (coming up with good lies and deliveries that make them believable) instead of simply numbers. These examples probably sound pretty detail driven, and this game will probably have lots of details like that, even if just as optional rules. But a good GM won'tnitpick everything, but use them when they really mean something.

 

[J1M]

You should review the D&D 5e rules for advantage/disadvantage. Basically, if you have advantage you get to roll 2d20 and take the better result. Disadvantage => roll 2d20 and take the worst result. It doesn't matter how many advantage sources you have.

Now, I'm sure this was lifted from some indie rpg, but this isn't about giving credit. One of the reasons this mechanic has been popular is because unlike 3e bonuses, you don't need to perform a lot of calculations in order to get the result. You should keep this in mind and do a pass over your mechanic categories. One of the big problems I see right now, is that you are willing to hand out a variety of +1/+2 bonuses, and I would need to consult a table to determine what number to actually add as I shift from +1x5+2x1 to +1x5+2x2 bonus as I walk into an aura or something.

I like the idea behind your certainty mechanic, but I think you can achieve similar math with less mental arithmetic and more consistency in terms of dice rolling. One of the biggest problems with d20 has been the size the static modifiers for to-hit or skill rolls become.

Suggestions to try out:
-Make the default roll for hitting something 2d10. This makes the default distribution less chaotic, and will smooth out those early sessions so people aren't sitting around the table constantly missing.

-Go further with your idea for a mastery stat, where players can expend some of it on a turn or day to adjust their roll's certainty.

-Using 1d20, 2d10, 3d6, (4d4?,) 5d4 for certainty levels is somewhat clunky. Consider incorporating the idea of rolling an extra die like the advantage/disadvantage mechanic in D&D. For example, more certainty could mean you roll 4d10 and take the middle two values.

-You should review the distributions of 2d10, 4d10, and 6d10. I think you will find they trend very quickly towards what you are looking for so that you don't need to support more than +2 certainty in your ruleset. I would handle this by saying that certainty does not stack, and +1+1 certainty does not equal +2 certainty. Since people have been trained to consider +1 a small increment, I would instead refer to these rules as something like "certainty" and "greater certainty".

 

[Alex]

You should review the D&D 5e rules for advantage/disadvantage. Basically, if you have advantage you get to roll 2d20 and take the better result. Disadvantage => roll 2d20 and take the worst result. It doesn't matter how many advantage sources you have.

Now, I'm sure this was lifted from some indie rpg, but this isn't about giving credit. One of the reasons this mechanic has been popular is because unlike 3e bonuses, you don't need to perform a lot of calculations in order to get the result. You should keep this in mind and do a pass over your mechanic categories. One of the big problems I see right now, is that you are willing to hand out a variety of +1/+2 bonuses, and I would need to consult a table to determine what number to actually add as I shift from +1x5+2x1 to +1x5+2x2 bonus as I walk into an aura or something.

Thanks! I will check it out the rules. But I don't really want to make advantage a yes/no affair. the focus when you do this seems to shift from coming up with good plans to just coming ups with a way to bullshit the GM. About your objection of using a table to shift, that isn't really necessary. As I see things, attributes and skills should scale exponentially and this is in part the reasoning for not simply accumulating modifiers. Thus, if you want to be pedantic, you could consider the scale of the exponential. So, let's say it doubles every four points. In that case, a point of difference is the same as multiplying by 1.19. So, every "+1" makes a difference of 0.19 and 10 "+1s" a difference of 1.90, or multiplying the number by 2.9. Which is the equivalent of a +6. Or, if you wanted to be even more exact, or say, the advantage has a static value (someone helping you pull up a gate don't really multiply your strength), you could cast the attributes into a linear scale and then recast their sum into the exponential one. But of course, I don't really expect anyone to go through all this work, ever. Instead, it is GM adjudication that decides how the bonuses add up. The detail about a exponential scale is just there to help him or her get an idea of what the values mean, not to actually be used in calculations.

I like the idea behind your certainty mechanic, but I think you can achieve similar math with less mental arithmetic and more consistency in terms of dice rolling. One of the biggest problems with d20 has been the size the static modifiers for to-hit or skill rolls become.

Well, I suppose I could be more consistent by having the players roll as many d20s as the certainty score and taking the result closest to the mean. But I really like rolling 3d6s...

Suggestions to try out:
-Make the default roll for hitting something 2d10. This makes the default distribution less chaotic, and will smooth out those early sessions so people aren't sitting around the table constantly missing.

The d20 for combat is kind of by design. It is supposed to represent that combat is something inherently chaotic (which in part is related to the conflicts between law and chaos commons to some of those pulpy stories). That said, I was thinking of having "fighting styles", which would increase the certainty in combat. However, fighting stles can mismatch eachother, soyou might end up with heavy penalties for using the wrong style, and maybe even an increased fumble chance. I still need to lay some ground work before I get to combat, however.

-Go further with your idea for a mastery stat, where players can expend some of it on a turn or day to adjust their roll's certainty.

One think I am trying to avoid is purely abstract stuff. I mean, maybe the players can reduce their mastery die to increase certainty, but that wouldn't be a generic mechanic. Rather, they would need to find a way to do that regarding the skill they are using (like the dowsing rod example, maybe). By virtue of these actions not being abstract, they can have further consequences, possibly changing other parameters or even changing the "stakes" of the roll (what happens when you fail and when you succeed).

-Using 1d20, 2d10, 3d6, (4d4?,) 5d4 for certainty levels is somewhat clunky. Consider incorporating the idea of rolling an extra die like the advantage/disadvantage mechanic in D&D. For example, more certainty could mean you roll 4d10 and take the middle two values.

I will think about this. Like I said, I think the easiest solution would be to roll the number of dice, and take the result closest to the mean. Maybe to be less complicated, I could do this: Certainty 1 has you roll a d20. Certainty 2 rolls 2d10. Certainty 3 rolls 3d6. Certainty 4 rolls 4d6 and takes the three values closest to 3.5. Certainty 5 does the same, with 5 dice instead of 4. And that would be the roof.

-You should review the distributions of 2d10, 4d10, and 6d10. I think you will find they trend very quickly towards what you are looking for so that you don't need to support more than +2 certainty in your ruleset. I would handle this by saying that certainty does not stack, and +1+1 certainty does not equal +2 certainty. Since people have been trained to consider +1 a small increment, I would instead refer to these rules as something like "certainty" and "greater certainty".

I know. I think certainty 5 would be the most I would need in game. Certainty 3 is already pretty average, but I wanted to allow for people who are specially under the effect of the law spectrum.

I really like the way Twilight 2013 does their rolls. I don't know if they made them up or if they modified them from somewhere else but it is an expanded version of the D&D 5 advantage/disadvantage system that J1M mentioned.

Your skill determines how many dice you get to roll(d20s in the case of the system but any dice could work, while being unskilled means rolling two dice and taking the worst roll) while your basic ability scores(range of 1-11 in this case) and any modifiers determine what you are trying to roll under. You roll all your dice and then take the best one. Best being lowest in this case because you are trying to roll under the target. If more than one die beats the target it adds a +2 to how much you have beaten the roll by, in case how well you succeeded matters. Being more skilled allows you to more consistently get better rolls while having a naturally high stat allows you to succeed a lot even with little to no skill.

I actually have that game somewhere in my machine. I really liked the premise of that game, but I think I never finished reading. I will be sure to take a deeper look on it. Anyway, the system you describe seems like the pool system used in Deadlands. I call it a pool because how many dice you roll depends on your skill. It is a very nice system, although it might be a bit hard to take any ideas from it directly without making my game use a pool system as well.

I like how this system makes stats very important, much more than a simple bonus to rolls. I think this system accomplishes a lot of what you are trying to accomplish in your system.

I have modified my own system to use a similar die rolling mechanic. Before that I used an open ended d20 roll where stats, skills, any any equipment/situational bonuses, just all added to the roll. The new way, and allowing characters with above an average ability score to use a limited number of rerolls based on the how high the ability score is, differentiates much better between characters with different ability scores a great deal more

I think that, even if I don't end up using anything from them, this system may well be a good benchmark for my own. So thanks a bunch!

 

[J1M]

Found this today while looking at something unrelated:

http://www.redblobgames.com/articles/probability/damage-rolls.html