This article is the seventh in a twelve part series that talks about the choices/decisions that need to be made when designing a role-playing game system. This time I talk about mechanics for making attacks; next time will be the defense mechanics, so this is only half the story.
About a year ago (mid November 2012), I started work on my 3D RPG. It was NaGa DeMon (NAtional GAme DEsign MONth), I was driving in my car (stopped at a light), and was thinking about how I could change/redesign my first attempt at a RPG which I also called “3D” to actually make it work. My first attempt was overly complex, slow, clunky, and not very narrative. But it had taught me a lot about what NOT to do, and being able to spot flaws is the first step towards fixing or avoiding them. The real motivation of the do-over was that “3D” is a cool name, and 3 dice are a good number to hold in your hand, roll, interpret, and pretty much the minimum needed to have interesting statistics (I got my PhD doing uncertainty quantification so interesting statistics were important to me).
I’d been hearing a lot of good things about Marvel Heroic Roleplaying which used a dice pool system where the size of a die (between a d4 and d12), rather than an additive bonus to a roll, represented how good a character was at various tasks. I wanted my core attempt resolution mechanic to be similar to that of Marvel Heroic Roleplaying, but to always use exactly 3 dice in the pool. I also wanted armor to provide damage reduction rather make you harder to hit, and weapons to have the same damage dice as in D&D, specifically a long sword would do d8 for damage. I also wanted there to be a “Finesse” rating/die for weapons, so for example a knife with a damage rating of d4 and a Finesse rating of d8 could conceivable be just as deadly as a long sword (which has a damage rating of d8 when wielded in one hand) depending on what other dice you had in your pool.
So what’s the point I’m trying to make with this introduction? The implementation of attack rolls was the seed that the core mechanic of and indeed the whole of my 3D RPG grew from. And if I asked you to describe the d20 system’s core mechanic, I bet that most of you would say something like “Roll a d20, if it hits then roll damage, if the d20 came up as a natural 20 you do double damage.” The point is that the game mechanic for making attacks is THE central mechanic in most RPGs. So if you’re thinking about designing a RPG, that’s probably the mechanic you should start with, and you really should design it in concert with your defense mechanic.
Not so long ago I wrote an article for the GSA entitled “Die, d20 die!” which caused a bit of controversy. Specifically I said that “Using a d20 roll as the central mechanic was the biggest mistake that Gary Gygax, may he rest in peace, made when he invented D&D.” And in response, some readers replied that they liked the d20 system and didn’t take kindly to me bashing it. Given the title of the article and the picture that got selected, that’s a very reasonable misinterpretation of what I was saying. I wasn’t bashing the d20 system as a whole, just the use of the d20 die, and I was doing so for statistical reasons. The d20 system actually has some really good mechanics that are worth salvaging, I just don’t think the use of the d20 die is one of them. If you’re interested in what I’d do if I were the lead designer for D&D Next, you can read about it in an article entitled “Extreme Makeover, d20-ish Edition”.
But that’s neither here nor there. The purpose of this article is to highlight things that you need to consider when developing a game mechanic for attacking.
You should start by working out the dice roll statistics for a melee attack with a non-reach weapon in-the-context-of/in-combination-with your defense mechanic. And by “working out the dice roll statistics” I don’t mean “here’s the complete distribution” (that can come later); I mean a semi-qualitative semi-quantitative understanding of what the typical outcome of a roll will be.
When the Star Wars Saga Edition RPG first came out, its designers made a big deal about a revolutionary new design philosophy they had used, namely, to design for the desired effect not to design for cause. Earlier games such as the Revised Core Rules (RCR) edition of the Star Wars RPG had been designed to model causes faithfully and “let the chips fall where they may” in terms of the effect. I’ll give you an example.
The RCR Star Wars RPG was mostly “D&D 3.5 in space” but with some significant changes to the attack and defense mechanics. Armor provided Damage Reduction (DR) and capped your dexterity bonus to defense. Characters had wound points that represented their physical toughness and vitality points which represented turning what would be a serious hit into a near miss or grazing hit that didn’t hurt you physically. Critical hits dealt your damage minus their DR directly to the target’s wounds. But lightsabers ignored DR and could deal obscene amounts of damage in the hands of a high level Jedi. To make the system a little less lethal, you had to roll a second d20 to confirm a critical hit (if you didn’t confirm the crit, the damage is dealt to VP instead of WP). But really, whenever there was a fight between heroic characters, whoever was lucky enough to threaten a critical first would just spend a force point to make sure they confirmed it. Ostensibly these rules made for a realistic simulation of combat in the Star Wars universe, but in truth the system was incredibly lethal, too lethal. There were a lot of threads in the WotC forums which crunched the numbers, and basically it was nigh impossible for there to be any high level characters. Anyone with that much experience should have gotten punked by a mook who got lucky many times over. And that is a classic example of a “design for cause” only philosophy.
In Saga with the “new design for effect” philosophy, things got a bit closer to D&D 3.5 (yes I’m saying it, D&D 3.5 was a mechanically better system than RCR Star Wars). In Saga, armor provided a boost to Reflex Defense, characters had “hit points” plus a condition track instead of vitality points and wound points, and critical hits did double damage again. How is this design for effect? Well for starters the Reflex defense and hit points of storm troopers were carefully chosen so that they went down with the “right” frequency for a “movie simulator” (as opposed to a universe simulator). They also ditched skill points for a one half level plus a training bonus, and characters got a single attack per turn unless they invested feats to get more. All in all, game play in Saga was a lot faster, smoother, and better balanced than in RCR. Saga has been widely hailed as the best d20 system ever because it got the effects/outcomes of attempts pretty much spot on. But I know at least one “RCR grognard” who was extremely unhappy and refused to play Saga because its representation of causes was less realistic than RCR.
If you have to choose between designing for effect and designing for cause, then history shows it’s much better to design a RPG for effect; but really you don’t have to choose one at the expense of the other. The effect that I designed my attack mechanic for was for each attack to typically deal a few points of damage against an evenly matched opponent and to rarely do significantly more; also, most attacks should “hit.” I figured that would be a lot more-fun/less-frustrating than only rarely hitting an opponent and usually taking them out when you actually did hit.
So with 3 dice how could I achieve the desired effect? Well the die with the largest rolled result (call it the “first die”) should usually be large enough to “hit.” The die with the second largest (i.e. middle) rolled result (call it the “second die”) should generally be very close to its average value, so if we set a character’s damage reduction close to the average value of a die roll, then the DR should cancel the with the second die plus or minus a point or two. And that leaves us with the die with the smallest rolled result (call it the “third die”) for damage plus or minus a point or two leftover from the second die. So to summarize the “designed for cause AND effect” attack mechanic in my 3D RPG: the first die determines success/failure; the damage you deal equals the sum of the second die plus the third die; and the damage the target takes equals the damage you deal minus their damage reduction.
That’s the touchy feely semi-qualitative, semi-quantitative math I did for my central mechanic while sitting in my car at stop lights in mid November 2012. Of course having gotten my PhD doing uncertainty quantification, it was simple matter for me to write a short throw away program to calculate the exact true statistics (assuming fair dice), and I used those results to set the value of defense scores (Reflex Defense and Fortitude Defense/Damage Reduction). It’s a good thing I did calculate the exact stats because the initial guess for the defense score values that I came up with in the car were a couple points too high to get the “right” statistics for the desired effect.
Because I did the actual math, rather than just the touchy feely qualitative math, my first actual play test ran fast, smooth, and pretty much like I intended it to. Based on the first play test, I: revised the critical hit damage; increased the number of chips in the Karma pool; included grabbing someone as part of the action needed to throw them rather than have it be a separate action; and revised the character sheets to make them more functional. But honestly, not much needed to change.
The second aspect of the attack roll mechanic that you should work out is how to handle critical hits.
From the design for effect viewpoint, there are 2 key choices you need to make about critical hits: how frequently they occur, and how much damage the target will take because of them. And again these decisions need to be made in the context of the defenses of the target. I chose to work on the damage first and the frequency second.
The “defense” features that I needed to worry about were Reflex Defense, Fortitude Defense/DR, and hit points. And really I could just ignore Reflex Defense because in my opinion a “critical hit” should hit whether or not the maximum possible value on the first die is as large as the target’s Reflex Defense. That left Fortitude Defense/DR and hit points. And because of the “cause” part of my design for cause AND effect philosophy, I was inclined to have critical hits ignore DR, so that was my tentative decision, conditional on how the numbers worked out. That meant I really only needed to worry about hit points.
I was thinking that mooks/non-heroic NPC would have about 10 hp (or more if the specific NPC needed them), and that starting hp for heroic characters would be 20, and that 30-40 hp would be typical for “low level” to “mid level” PCs and important NPCs, while “high level” PCs and BBEG type NPCs might have about 60 hp. Yes those numbers are low compared to d20 systems (a 20th level heroic character could easily have 200+ hit points in the Star Wars Saga Edition RPG), but in general I think it is better to use smaller numbers in a RPG to make the math simpler.
I was thinking that mooks should go down with one critical hit, as should a heroic character with the base 20 hp, but “low level” to “mid level” PCs should be able to take at least one critical hit without going down, and that “high level” characters should be able to take 2 critical hits without going down, and go down on the third critical hit. That was the effect I was designing for; I just had to make sure that the numbers worked out right.
My first take on critical hit damage was that it would ignore Damage Reduction and be the sum of all three dice. I figured optimized PCs would probably have attack dice pools of 3d12 or close to it, and the average sum of 3d12 is 19.5, round that up a few points because PCs get to choose when to score critical hits via Karma (i.e. on a good roll), and it achieves the desired effect, or so I thought.
Then it was time to deal with the second part of critical hits, i.e. how often they come up. There are two things to worry about when choosing the frequency of critical hits. First, you don’t want critical hits to be so frequent that PCs die because they get punked by mooks (accidental character death really sucks). Second, you want PCs to have a decent chance of scoring critical hits when it matters. Those are competing objectives. My solution was to make rolling a critical hit very unlikely but provide 2 mechanics that let PCs either stack the odds in their favor or just straight out say “I scored a critical hit” (following the model of Destiny Points in Saga Edition). So critical hits occur when you roll the max value on all three dice, which had probability 1/64 for 3d4 and 1/1000 for 3d10, or you spend a Karma chip. Karma chips let you straight out say “I scored a critical hit;” my destiny pool of dice mechanic lets you stack the odds (for WHEN a critical hit occurs) in your favor. You can read more about my destiny mechanic in part five of this series, but basically you have a d12, d10, d8, d6, and d4 in your destiny pool whether you’re the GM or a PC and you can swap dice you roll with dice of the same size in your destiny pool.
To keep GM’s from abusing the destiny pool to score critical hits on PCs I made it a rule that, whenever a NPC scores a critical hit on a PC, one Karma chip flips from black to white. Basically, the GM can NOT score a critical hit on the PCs without giving the party white Karma (unless the Karma pool is already all white, in which case a NPC rolling a critical hit on a PC can still happen and does not incur a Karma debt).
So I wrote a short little program, where you tell it the dice for the attack pool, the defenses of the target, and it tells you how many attacks it takes for the target to die (accounting for critical hits and the destiny pool but not Karma, because “choice” is extremely hard to model). But since how long it takes to die is a random thing, I had the program repeat the process 1,000,000 times and give me a histogram (rectangle/bar) plot that approximates the “true” time to die distribution. And the statistics looked very reasonable. Then came the first play test.
I had one PC for my first play test, a total newb to RPGs (total newbs are actually the best/most informative play testers when the goal is a super simple system that runs fast and smooth). By “total newb” I meant he repeatedly asked “Which one is the dX again?” where dX is a stand in for d8, d12, etc. He’s a comic book fan so I statted out “iron guy” and “guy of steel” (actual names are not used to avoid potential intellectual property rights issues) knock offs for him to choose from. He chose the “guy of steel” knock off; the setting was a post apocalyptic mega future. In an hour and a half, I explained the rules, set up the battle map (with 1 inch squares representing “flying squares” see part 6 of this series for a discussion of position tracking in my 3D RPG), chose minis, ran two combat encounters, put everything away and took notes on his feedback. The system ran fast and smooth; also the “guy of steel” knock off character felt authentic (to me anyway) in terms of how he played yet did not wreck game balance. All and all, the first play test validated the design by the numbers math I had done as being spot on, with one exception: critical hits.
Concerning critical hits, I didn’t do the math wrong; I did the wrong math. I had incorrectly assumed that either the destiny pool OR Karma would be typically used for critical hits. The reality, that even the total RPG newb player figured out and took advantage of, is that the destiny pool AND Karma would typically be used together for critical hits and that made very deadly crits a lot more common. On the plus side, that the game is simple enough for a total RPG newb to find the optimal way to score a critical hit is a very positive sign. On the negative side, I was a little too focused on the math to predict this ahead of time.
Just to be clear, the sum of three dice critical damage did NOT “break” the game, but it was definitely sub optimal and I’m shooting for as close to a perfect system as I can get. The “average” numbers I had calculated for the sum of three dice were near optimal (actually a point or two to below the “golden” target), but combining Karma and destiny meant you could CHOOSE to do significantly greater than the “average” damage. So… I changed critical hit damage to be double the first die. This substantially decreased the maximum (and typical considering destiny) damage you could do with a crit while increasing the “average” damage a few points to be spot on the golden target. It also made the PCs’ decision of when is a good time to spend Karma (you only need a good roll on one of three dice) a lot easier, which in turn made it a lot easier to model the player’s “choice” about when to spend Karma. And it made the “typical” crit with Karma and destiny a lot closer to the golden target
For those readers who are not well versed in statistics, “typical” (most common) and “average” (i.e. the mean, i.e. the sum of outcomes divided by the number of outcomes) outcomes can be very different things, particularly if the distribution is multi-modal or otherwise has a large standard deviation. For example: a d20 die roll has a uniform distribution so no possible result is more “typical” than any other from the point of view of probability. And since a d20 has a large-standard-deviation/lot-of-variability, its average isn’t even “close” to all possible results. Basically, “average” does not mean anything close to “typical” for a d20; this is one of the reasons I wrote the “Die, d20 die!” article.
The double the first die damage for critical hits rule also made crits scored by spending Karma nearly identical to crits scored by rolling the max result on all three dice, which is an independently good thing. That combining Karma and destiny was typical also meant I needed a slightly larger Karma pool, so I increased it by 2 chips with the justification of it representing the NPCs under the GM’s control which had useful side benefits.
Note that whenever I say something like “optimal,” I mean something like “optimal in the context of the other design decisions that I made.” So when I say the new critical hits deal double the first die of damage (and ignore damage reduction) is “golden” or “optimal,” I mean optimal assuming that “low level” heroic NPCs have about 20 hit points, “low level” PCs have between 30 and 40 hit points, “high level” heroic characters have roughly 60 to 65 hit points, PCs are rolling 3d12 or 2d12&d10 or d12&d10&d8 for their attack dice, and they spend a Karma chip to trigger a critical hit after rolling at least one 11 or 12 (including use of the destiny pool). This means that critical hits triggered would do 22 or 24 points of damage, which meets my stated design goals for how many critical hits are needed to bring various types/power-levels of characters down.
Opinion: in terms of player perception, i.e. rare enough to be very impressive when it happens, but common enough for players to track its frequency semi-reliably without resorting to a written log of dice rolls, the ideal frequency of a rolled critical hit is between 1 in 12 rolls and 1 in 14 rolls. One in 20 is just too rare for players to track reliably (another reason for me to bash the d20 die, not the d20 system); I’ve got anecdotal evidence to support this, but I’ll leave it as opinion for now.
When using destiny but not Karma, critical hits will occur most frequently when you use 3 dice of different sizes (so you can store maximal rolls in the destiny pool as they occur and pull them out as soon as you have 3 maximal results stored in the pool). If your attack dice pool is d12&d6&d4, then using destiny you can roll a critical hit once every roughly 13.75 rolls, which falls within the (in my opinion) optimal frequency bands.
If your attack dice pool is d12&d10&d8, then using destiny you could roll a critical hit once every roughly 17.5 rolls. This is a little rarer than the ideal frequency, but the average damage for non-critical hits is greater so it evens out. The touchy-feely math says that average non critical damage of d12&d10&d8 is about 4 points higher than d12&d6&d4; but that doesn’t account for interaction with DR or destiny. I haven’t yet written a program to calculate the distribution of non critical hit damage assuming optimal use of the destiny pool.
How to calculate statistics
In response to earlier articles in this series, I had gotten a request to explain how to calculate the statistics for various dice rolls. This is very context specific (it depends on your rule set) but I can give you a few generally applicable pointers.
- Write a program to calculate the stats for you; doing the math on paper is prone to error even if you know how to do it; the math is very complicated/involved/intricate but not terribly difficult (no calculus or even algebra). Fortunately, you can avoid all those mathematical intricacies by making the computer brute force the solution. For this approach, you only need to implement the “relatively simple” logic for rolling and interpreting dice (does it hit? how much damage does it do?).
- The logic for optimal decision making based on dice rolls (for example when to switch which dice into and out of my destiny pool) can be rather involved, but it also makes the analytical math nigh impossible, so you’re stuck with writing a computer program. Also, working out the logic for optimal decision making is a LOT easier than modeling human behavior (humans aren’t always logical).
- Calculate the frequency of occurrence of events rather than the mean and standard deviation of the events; for example to calculate damage statistics, use an array whose index represents the value of the damage and the corresponding element of the array is a counter of how many time that event comes up. When you get done looping over the events, you can calculate the probability of each type of event (i.e. the different possible values for damage) by dividing how many times it came up by the total number of events; this gives you a histogram representing the distribution of damage rather than just the mean and standard deviation. You can “easily” calculate almost any other statistic from the histogram representation of the distribution. Also the histogram/distribution will show you what outcomes are “typical.” What’s “typical” affects player perception/experience/enjoyment to a much greater degree than what’s “average.”
- If you can exercise every possible combination of rolls in 6 or fewer nested loops, then use this exhaustive approach to create the histogram; it will give exact statistics assuming fair dice (real dice aren’t perfectly fair but they tend to be pretty close). Since my 3D RPG uses 3 dice for each roll, that means I can calculate exact statistics for any action that can be resolved in 2 or fewer rolls (if I make an assumption about what the destiny pool holds at the beginning of those two rolls).
- If you can’t exercise every possible combination in 6 or fewer nested loops, then use a random number generator to simulate dice rolling inside a single loop that repeats the roll roughly one million or more times. The “million or more times” comes from the central limit theorem which says the standard deviation of the error in a computed mean of a function equals the function’s standard deviation divided by the square root of the number of samples; thus one million rolls gives you roughly 3 significant figures of accuracy (the square root of 1 million is 1 thousand, i.e. 10^3, i.e. 3 significant figures). This is because the standard deviation of the probability, P, of an event is stddev(P)=sqrt(P-P^2); for you math geeks out there, this follows directly from calculating probability as the mean of the indicator function (the indicator function is 1 when the event occurs and zero otherwise). Something that occurs only 1% of the time (P=0.01) isn’t in-the-head-trackable (it’s so small it doesn’t matter), and stderr(P)=sqrt(0.01-0.01^2)/1000 ~ 0.0001, which means you’ve got 2 significant figures of accuracy for something too small to track. For events that occur with probability P=0.5, you have exactly 3 significant figures of accuracy; i.e. stderr(P)=sqrt(0.5-0.5^2)/1000=0.0005. That’s more than enough accuracy for the purpose of designing a RPG.
- The 6 or less vs. more than 6 nested loops criteria is based on the current speed of computers. Currently personal computer speeds are measured in “gigahertz” which means 10^9 floating point operations per second. Since typical programs will be able to get 5% or less efficiency, we’re talking on the order of 5*10^7 operations per second. Each roll involve multiple floating point operations which means a million rolls takes on the order of seconds to evaluate. Ten to the 6th power is a million. So six nested loops has comparable computational cost to one million dice rolls (which would give you 2 to 3 significant figures of accuracy for any probability large enough to matter to a RPG). For 9 nested loops we’re talking about 10 minutes to a few hours of number crunching for the computer; iterating that process with rules changes is just not practical.
You need to determine how double/multiple attacks work. For example: do character’s gain extra attacks, possibly at a lower attack bonus, as they increase in level; do they have to spend feats to gain extra attacks; can all characters make 2 attacks? More attacks is not always better; it could require extra dice rolls which would slow down game play, and that’s something that you’d want to either avoid entirely or at least discourage.
In RCR Star Wars and D&D 3.5 all high level characters could make multiple attacks (the number depended on their base attack bonus), but to get them you had to spend a full action which meant you gave up your movement other than a 1 square step. In my opinion, this was a bad design decision because it encouraged characters to stand next to and wail on each other until one of them dropped… that isn’t very cinematic. If you watch any swashbuckler movie (and any Star Wars movie except “A New Hope”), there’s a lot of movement, dodging, swinging from ropes, etc. and in my opinion a good RPG system will encourage, or at least not discourage movement.
In my 3D RPG, all characters can take double actions. A double action is completely resolved with one roll of the dice; the first die is used to determine success for one action and the second die is used to determine the success of the other action. Either, neither, or both of the actions in a double action can be an attack. Any attack made as part of a double action deals damage equal to double the third die and damage reduction is applied against each attack separately. Since the third die, i.e. the smallest die, counts twice when determining the damage dealt by a double attack, this discourages characters from taking double actions. Allowing but discouraging double attacks is a good thing because narratively speaking they should be something special not the default go to, no-brainer decision. If a double attack results in a critical hit, then both attacks deal damage equal to double the first die and ignore DR. Yes that makes spending Karma chips to score critical hits more powerful, but spending a Karma chip to take out two mooks is still very reasonable, and if a PC does this against a BBEG, then the GM should seriously consider flipping a Karma chip from white to black to make the double attack miss.
You need to decide how “reach” melee weapons work. First of all you need to decide whether to even have a mechanic for “reach” weapons, it’s perfectly acceptable not to, because most reach weapons are large weapons which means they generally deal a large amount of damage. If you do choose to implement a reach mechanic, your available options strongly depend on how you implemented position tracking; you can see part 6 of this series for a detailed discussion about position tracking, distance, and movement. But basically, if you’re using a gridded battle map then attackers wielding a reach weapon don’t have to be adjacent to their target. Another option that doesn’t depend on your position tracking system is to have reach weapons grant a “first strike” ability similar to that Pikemen from Wizards of the Coast’s “Magic the Gathering” card game; this could translate to a bonus on the initiative check at the start of combat.
You need to decide how ranged attacks work. Copying your melee attack mechanic is a good place to start but you should also account for how range affects the difficulty of hitting a target.
In many RPGs from ancient times (read as “circa 1990”), any target within a weapon’s range was equally hard to hit; this is still true of the Palladium system.
Some personal opinions about the Palladium system: It’s long overdue for a complete redesign. “Complex” and “advanced” are NOT synonyms. The “Rifts” RPG created by Kevin Siembieda has an absolutely amazing setting and makes buying the books worthwhile despite the archaic Palladium system. If you’re a fan of “Rifts” but prefer simpler, faster, better balanced, and more narrative game systems, then you might be interested to know that I initially designed my 3D RPG with the intention of using it for a home game in the “Rifts” setting. Also, I’m releasing my 3D RPG rules-set under the open gaming license; I’m looking for recognition/gaming-fame not financial compensation, so if some established game company wanted to adopt my system for their own setting I’d be very happy about that. But that’s neither here nor there.
Modern d20 systems typically have a range dependent penalty. For instance, in Star Wars Saga Edition, you suffer no penalty for attacking targets in point blank range, a -2 penalty for attacking targets at short range, a -5 penalty for attacking targets at medium range, and a -10 penalty for attacking targets at long range. For a d20 system, that’s about as good as you can get; but dice pool systems provide other, and in my opinion better, options.
For example in Fantasy Flight Games’ Edge of the Empire RPG you add extra difficulty dice as distance increase, this is similar to a range penalty but better because the narrative dice system has “threat” in addition to “failure” symbols on the dice.
But my personal preference, and what I implemented in my 3D RPG, is that each ranged weapon can have a different Finesse die size for different range increments (short, medium, long, extreme). You might be wondering how this differs from a fixed penalty based on range increment. Well there are 2 differences:
- Different weapons can have different “penalties” for a particular range increment.
- Some weapons, such as a guided missile, can become more accurate as the distance between the attacker and target increases.
Another example of the second benefit is that Rifles increase in accuracy going from short to medium range, and then get less accurate as they approach extreme range. This lets pistols be the ultimate short range weapons and rifles be the ultimate middle range weapons. The point is that this is a lot more realistic than uniformly fixed penalties to ranged attacks based on distance and it is still very simple.
- characters never run out of ammo;
- tracking how many shots you have left before you need to reload and/or buy more ammo,
- having dice rolls determine when you need to reload a weapon and/or buy more ammo (which can be different).
I chose the third option for my 3D RPG. For example: revolvers need to be reloaded whenever the second die comes up as a 2 or less, and they need to have more ammo purchased whenever the second die comes up as a 1. Here “purchasing ammo” does not meaning running to the store; it means the character pays 1 character point on the spot to have retroactively brought enough ammo. In my opinion, it is not worth the mental effort to track shots at the gaming table.
You need to determine how aiming works. In the Star Wars Saga Edition RPG, aiming costs 2 swift actions and lets you ignore cover. If you have the careful shot feat it also gives you a +1 to your attack roll. In my 3D RPG, aiming requires half your squares of movement for a round, and gives you +1f favorable circumstances. You can read about how I implemented favorable/unfavorable circumstances in part 5 of this series. In addition to the +1f favorable circumstances, if your ranged weapon has a scope, then aiming lets you use a different/larger die for finesse (listed in the weapon’s stats).
You need to determine how area attacks work. You have a lot of options here; I’ll mention a couple of modern implementations.
Star Wars Saga Edition and D&D 4e are both d20 systems, and were designed by the same company at the same time, but they handle area attacks very differently. I greatly prefer Saga’s method because it can be resolved in much less time.
In Star Wars Saga Edition you roll once for an area attack and that single d20 roll (plus modifiers) is compared to the reflex defense of every target in the area of effect; damage is also rolled once for all targets. If the attack hits a target you deal normal damage to that target; if the attack misses but would hit a reflex defense of 10 you deal half damage to the target; if the attack misses and would not hit a reflex defense of 10, you deal no damage to that target. Also, area attacks can NOT score critical hits, and the evasion talent reduces the damage to half damage on an area attack that hits and no damage on a miss.
In D&D 4e you roll a separate d20 for every target in the area of effect; if the attack hits, it deals full damage, if the attack misses it deals no damage.
Area attacks in my 3D RPG are simpler than either Saga or D&D 4e. A single attack roll applies against all targets in the area of effect. This is a standard attack roll, you don’t do half damage on a miss, and it can still critical as normal. There are a few differences from normal attacks, e.g. explosions in confined spaces are always treated as critical hits. There are also a couple other variations on area attacks such as: “splash” attacks which target a single character and if successful then the roll is also compared against adjacent characters; and the magical “chain lighting” spell which lets you choose which characters in the area of effect to attack and to not attack.
You need to determine how flanking works. In d20 systems, you can “flank” an opponent by having two allied characters occupying squares on “opposite” sides of opponent. And flanking an opponent can trigger all sorts of special abilities like for example sneak attack; it might also give you a bonus to attack. I kept flanking simple in my 3D RPG, all it does is give you +1f favorable circumstances. I proposed something a bit more complicated in the Extreme Makeover, d20-ish Edition article, basically I made it one way to spend a “tactic action.”
You need to determine how grappling works. Historically, grappling has been one of the more complicated parts of d20 systems. Saga greatly simplified it, required feats like pin or trip for a character to “grapple” but introduced the weaker option of “grabbing” that anyone can do.
In my 3D RPG: anyone can grapple; the required skill is “Brawl;” all characters engaged in a grapple take their turns at the same time; the character that won the last grapple check gets favorable circumstances on the next one and gets to choose who goes next. The only character(s) who can leave a grapple is/are the one(s) who won the last grapple check, and yes that means multiple characters can “cooperate” (contribute to the dice pool) on a grapple check like they can on any other action. A grapple can also be part of a double action such as a tackle (running + grappling), or a damaging attack such as a bear hug or opposing characters wrestling over a pistol, in which case only the winner of the grapple can deal damage against the loser with the secondary action. Semi related to grappling, if you want to throw a character martial arts style the two required skills are “Brawl” and “W.P. Thrown Weapons;” if you throw one character into another it’s treated as a standard double attack.
You need to determine how touch attacks work. This is pretty straight forward; touch attacks target reflex defense, possibly with a bonus, and just touching someone doesn’t cause damage unless there’s a rider effect like a contact poison. Like I said it’s pretty straight forward, but you need to make sure your rules cover it explicitly because it is going to come up.
You need to determine how attacking an item works.
If you look at epic works of fiction, for example “The Lion, the Witch, and the Wardrobe” (where Edmund destroys the White Witch’s Wand), or “The Lord of the Rings” (where 3 books are devoted to the destruction of the one ring), the destruction of an item is a narrative event. If you look at most RPGs, attacking an item falls into either of two categories
- an easily abusable I win button (what is King Arthur without his sword Excalibur?)
- incredibly difficult for PCs to pull off at the NARRATIVELY APPROPRIATE MOMENT.
Both are bad options; if I had to choose between them I’d choose option 2, but fortunately there is another, better option: make the destruction of an item a narrative event, i.e. use a mechanic for narrative fiat. I talked about narrative fiat mechanics in part 5 of this series, and recommended for RPGs to have at least 2 of them, a weaker and a stronger version. Personally I prefer having 3: a “make it so” mechanic, a “tip the odds in your favor before the roll” mechanic, and a “make the rolls less random by altering them after the fact” mechanic. In my 3D RPG, these are “Karma,” “Themes,” and “Destiny” respectively. You can read more about them in part 5 of this series, but basically they all involve paying some kind of cost from a limited “resource” to achieve a desired effect (here I’m using the word “resource” in a non-technical sense because “destiny” does not carry over from one session to the next in my 3D RPG). I believe that the destruction of a held item should require the use of the “make it so” mechanic, so in my 3D RPG this involves spending Karma. Also destroying a magical item always requires spending Karma whether or not it is held, and “artifacts” can only be destroyed at the place they were created.
Do you want to have attacks of opportunity and readied actions? As a generally rule of thumb I vote “no” because they really slow down game play and choosing a simpler and more flexible initiative system makes them unnecessary.
Play Testers Wanted
Rodney Thompson, lead designer of the Star Wars Saga Edition RPG, has repeatedly said that “no game system survives contact with the players.” I suppose that depends on how “survive” is defined; Saga was widely hailed as the best d20 system ever, but errata was issued based on “contact with the players”; Fantasy Flight Games did such an amazing job on Edge of the Empire that the core book had only 3 pages of errata/clarifications. Personally I’d call both of those very survivable injuries but Rodney had a valid point nonetheless.
Knowing “how to math” well gives me an advantage over most other game designers, but I’m not arrogant enough to think that I haven’t missed anything. The first play test confirmed that I’m not clairvoyant/infallible and the resulting changes (mentioned above) really improved my system. From my most recent play test, I learned that a “perfect” RPG (my admittedly impossible goal) really needs simple scalable mass combat rules, so during the writing of this article I expanded my Organizations mechanic to cover mass combat (I’ll talk about Organizations, Mass Combat, Mental Combat, and Social Combat in part 9 of this series). I expect further play testing will identity other holes for me to plug.
If you’re reading this, I’d like YOUR help to play test my system. I had 2 PCs for an online setting development campaign but “real life” got in the way and caused one of them (you might know him as isdestroyer on the d20radio.com forums) to drop out. I think my system is pretty amazing/inspired and isdestroyer agrees; you can ask him yourself, it simply was a time factor. I don’t blame him because life got a little crazy for me too recently, but I’m going to have a lot of availability between Christmas and New Year’s for online gaming over roll20 (with a Google plus hangout for voice). So if you’re up for a one shot or a short series of adventures between Christmas 2013 and New Years please, please, please send a PM to EliasWindrider on the d20radio.com forums or leave a comment below.
And if me or Isdestroyer proclaiming my 3D RPG as a pretty slick system isn’t enough to motivate you to contact me, then maybe GM Dave’s endorsement will; he called it a “solid” system. You can listen for yourself in episode 14 and post show of episode 3 of the Order 66 podcast, Fantasy Flight Games reboot. In private communication, GM Dave also compared my system to Mouse Guard in terms of being simple and thus a system he’d like to GM.
In Part 8 of this series (next time) I’ll discuss defenses (using the term broadly), health, healing, recovery, death, and saving throws. Part 9 will cover organizations, mental combat, social combat, and mass combat (believe it or not those are all related to organizations in my 3D RPG). Part 10 will deal with balancing power levels and different scales, e.g. NPCs vs. PCs, characters vs. vehicles. In Part 11, I’ll cover equipment, magic, cybernetics, super powers, character modes, etc.; basically character schticks and options. In Part 12, I’ll discuss gaming aids; things like character sheets, GM screens, dice rolling apps, and character building software. About that, I found out three months ago that I’d need to get approval from my employer (a government research lab) to distribute self written software for free when it’s associated with any activity that could be perceived as outside employment (such as publishing a RPG); and getting approval to do any outside employment that involves “technical skills” (such as programming) is a lot more complicated than getting approval to do “non technical” work such as creating a pen and paper RPG.
Since GM Dave announced it in episode 14 of the Order 66 Podcast (Edge of the Empire edition), I can say that Gamer Nation Studios is the game design company that will be publishing my 3D RPG. The first book is going to feature a slightly stripped down version (e.g. no cybernetics/robots chapter) of my system for a traditional (D&D style) fantasy setting. If you have the time, inclination, and talent to be a world builder then send a PM to EliasWindrider on the d20radio.com forums. I’m looking for 1 to 3 fluff/setting writers who would get a share of the profits.
Here’s a brief outline of this setting. It’s a primitive prison planet; you can think of it being like real world Australia which started a penal colony for Great Britain. There are wormhole gates from various technologically or techno-magically advanced worlds that periodically open and deliver newly convicted felons: Humans, Elves, Dwarves, Halflings, Orcs, Trolls, Goblins, etc. Each race generally (but not always) arrives at a different gate. “Wardens” (NPCs with mega technology bionics) prevent the “inmates” from escaping through the gates when they open. While the world is a bit grim, gritty, seedy and rough, good people do live there, in fact many of the inhabitants were born on this planet and they tend to be a bit more “innocent” then the first generation felons.
Amid the wilderness and wild beasts, there are bastions of “civilization.” The major cultural centers sprang up in the vicinity of one gate or another under the protection of the Wardens. The level of technology among the “inmates” is at the late medieval or early renaissance level; for example, archaic firearms like muskets and flintlock pistols exist. Although Dwarves and Goblins will sell black powder to other races, how they make it is a closely guarded secret. Elves have the greatest mastery of magic. Humans are well known as traders and make the best ships; Goblin pirates and other swashbucklers are part of the setting. There is rivalry and bias between the different races, and even among the different human kingdoms, but the gate cities are neutral ground. The different races can intermarry and breed (they won’t get much if any page space in the first book but Atlanteans are the progenitors of all of the other “humanoid” races and are responsible for their common genetic makeup).
The “Karma effect” is well known and accepted as true (the evidence for it is overwhelming), but what causes it is the subject of a fierce debate among the inhabitants of the game universe. Some assert that conservation of Karma is a natural law like the conservation of mass, momentum, and energy. Others believe in an all powerful, primal, dualistic good and evil, semi-aware, mystical energy field created by all living things that binds the universe together. Still others place their faith in a supreme being who rewards the good and punishes the wicked, although different religions disagree on the exact nature of this supreme being. One of the more pragmatic religions, Karmaism, proclaims the existence of an omnipotent “unknown god” who is unknowable by mortals, but chooses to reveal his/her just nature through the Karma effect. If you want to create a religion in the game world, go for it, but it’s not going to be the only one, and won’t be established as “true.”
Two other things I’d like help with are:
- Coming up with spells, now I don’t mean rules, I mean effects, descriptions, inspiration. Fireballs, disintegrate, etc. are obvious. I’m looking for the less obvious stuff.
- A list of mythological creatures you’d like to see in the game.
If you’d prefer to publish a separate setting book using my system that’s OK too and I don’t need to be involved in it; my system is going to be released under the Open Gaming License.
This work, unless otherwise expressly stated, is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 3.0 Unported License.