Quote from: Xanther;900430By complicated I don't necessarily mean intellectually so but in a more Methods Engineering sense.  That is to break down the steps and operations that take you from first determining a roll is needed to result.  More step and more complex the operation the more complicated the system.  

The terms complex and complicated are bit of a misnomers.   Adding three numbers ranging from 1-8 is not complicated but in the context of a game, where you want things to flow very fast and where the complication (time to resolution) increases non-linearly with the number of steps, even simple one, a lot of simple things become very complicated very quickly.

Yes I agree with you when it comes to the complication. Now on the other hand the smaller odds of the long tails allow for a single roll to represent what would otherwise require multiple rolls on simpler dice. For example if I'm rolling a d20 and do d6 damage I can represent the damage roll in a single die roll if my odds of hitting are 60% or 30%, in the former each two steps (10%) represent a hp of damage, and in the later each step (5%) represents a hp of damage. So if I need a 15 or better a roll of 15 also represents 1 hp of damage, 16 two hit points, 17 three, and so forth up to 20 being 6 hit points of damage. Yet if my odds of hitting are different or my damage is 1d8 then I'm out of luck and I can't simplify.

I use the term "creative bandwidth" for the point you're addressing. The amount of "perceived reality" for a given unit of effort (rolling, adding dice, adding modifiers, rolling again, etc.). I use the term perceived instead of create or determined because the mind can augment certain detail or filter out other. This is similar to watching a video on a slow connection, the frame rate is more important than the resolution. All the high detail of a low frame rate stream would get lost and on the other hand the low detail of a high frame rate stream would get augmented. I believe this holds true for tabletop rpgs as well in which detail (simulationism) is augmented by a high frame rate (narrative if you will?). This is at the essence of the argument with hit points and combat of attrition. I'm doing a lot of meta-data changes with the die roll, but in truth I'm not affecting the story much if the character or opponent remain as combat effective. What is really being changed by all those die rolls and all that math? It's also at the essence of the argument against Manzanaro's simulation vs narrative dilema. It's the effect and not the procedure that matters. If I can replace one operation with a "perceived equivalent output" then they're the same even if internally they're different or by some "theoretical definition" they're different. So we can discuss for days on end about the theoretical differences, but if the output is similar enough and one is more "efficient" than the other then it's a good candidate for a drop in replacement even if "theoretically" they're entirely different.

Now for a practical example. So I have this mechanism with 3d8 that can be rerolled and added. As you point out this can lead to complications and complexity escalation because the better my character the more dice I'll keep adding as they explode over and over again. This "reroll" bonus can turn into handicap. Now resolution is done by opposing die rolls so any value of 0 or better becomes a success for the one performing the action and more so the higher the difference. First point, I'm using degree of success instead of a binary success-failure that then requires a secondary roll to determine degree of success. The slopes are in such a way that if you're good is easy to get a +5 difference and if your not so good then a +5 is quite uncommon. Second point, since they're opposing die rolls I can resolve two actions at the same time. For example in hand to hand combat if I approach my opponent my opponent is also approaching me even if my opponent doesn't move, right? So I'm susceptible to my opponent's attack as much as my opponent is to mine. So I can say that if the difference is 0 or better I hit my opponent, if the difference is -1 or lower I miss, but if the difference is -10 or lower I'm the one who actually gets hit. So what would otherwise be two rounds and two set of rolls is now condensed into one. I can even define that if I get a +10 or better I've actually suppressed my opponent and get to attack again. So -10 is I get hit instead of just missing, 0 is ok and I hit and 10 is I hit and get to attack again immediately after. Now combat styles. If my opponent is way larger I can set myself in defensive mode. This lowers my odds of hitting (which are low already), and increases my defense (lowers odds of getting hit) while at the same time improving the odds of my opponent getting hit on a bad roll. So I can take a -4 to attack and a +4 on defense. So when I attack I get hit on a -10, hit on a 4 or better and suppress on a 14 or better (instead of the standard -10, 0, 10 bracket). When on defense (my opponent moves in for an attack) my opponent gets -6 to get hit, 4 to hit and 10 to suppress (instead of the standard -10, 0, 10 bracket). This means that it's convenient for me to wait for my opponents move (forfeit classic initiative) and by doing so actually keep initiative by leading the battle looking for a break in my opponent's defense. My odds of getting hit are reduced although if I do I'll most surely be suppressed, but at the same time my odds of finding a hole in my opponents defense increases dramatically from -10 to -6. So I'm getting more out of a more complex die roll so I'm trading in more complexity per action to less actions to resolve something. More so, since the die roll curve is infinite something outstanding can happen. On my opponents attack the outcome can be 5, 2, -4, -8, -15, -19, -21, -56, etc. So I can go from getting hit (5) to being missed (2, -4), to hitting back (-8, -15, -19, etc.) and the negative value means intensity so -8 can be slight effect, I evaded the attack went under the shield deflected the sword and struck my opponent's arm. On the other hand -19 can be an instant death as I swing low and shove the sword into the chest and directly to the heart. The -21 is even better and I'll leave the -56 to your imagination. I can't do this with a bounded and finite die roll range.

So yes, on one end it's more complex, but on the other I can get more form the same roll and more things I couldn't get from other types of rolls.

Back to the supposed thread topic, the discussion of large dice pools and extreme tail results, I think they are a good example of how to have mechanics that get players excited about what happens. That is, I've seen many players get excited about critical successes and failures, but if a low-grain system is being used, especially d20, then crits are actually quite common (10% chance of rolling a 1 or 20) and there's no difference between one crit and another, and it has nothing to do with the game situation. So if the system takes into account skills and circumstances, and measures the degree of critical effect, those issues can be addressed, and extra-juicy effects can be stuck at the rare extremes of the crit tables. It helps to do some statistics to figure out the extremes and the rates of certain results, so they aren't too common or inappropriate.

Quote from: Skarg;900757Back to the supposed thread topic, the discussion of large dice pools and extreme tail results, I think they are a good example of how to have mechanics that get players excited about what happens. That is, I've seen many players get excited about critical successes and failures, but if a low-grain system is being used, especially d20, then crits are actually quite common (10% chance of rolling a 1 or 20) and there's no difference between one crit and another, and it has nothing to do with the game situation. So if the system takes into account skills and circumstances, and measures the degree of critical effect, those issues can be addressed, and extra-juicy effects can be stuck at the rare extremes of the crit tables. It helps to do some statistics to figure out the extremes and the rates of certain results, so they aren't too common or inappropriate.

Meh. I HOPE this is not the supposed thread topic. Because it's really just a lot of mathurbation. That leg of the discussion exists isolated in a space of vague abstraction, safe from any semblance of story or actual play. My understanding of the topic is how to get better story in actual play. But it's worth my time to criticize each and every one of these points and assumptions because a) they are so widely accepted, and b) you can do the exact opposite, lose zero fun (if anything, you will on net game) but also save time in doing so.

1) Linear vs Curve systems. The differences are highly exaggerated. If you plot the probability of success curve in any curve system, you get something that fits the general shape of an italicized 'S' whereas presumably the linear system gives you something that looks like a slash '/'. I say "presumably" because there is a reality that you can't avoid in game design--namely that probabilities are bounded by 1 and 0. So whether you let chance of success/failure go straight to certainty, or whether you put in a ceiling/floor (be it 1%/99%, 5%/95% or anything else), the "curve" ends up looking more like a squared off lightning bolt. And if you do like many systems do, have exceptions to the main mechanic at the extreme highs and lows, the final "curve" ends up extremely close to the italicized 'S', just a little more squared off.
2) Pareto's 80/20 rule tells us you get the most mileage out of just 20% of your work. The other 80% is a lot of effort that applies to so few cases. Certainly tail results exaggerate this effect. Game designers get off on wasting so much time perfecting the tails when, by definition, they come up almost never.
3) That said, why is it a problem if crits are common? For one thing, as you increase in skill, you can go from failure being more common than success to success being more common than failure. No one bats an eye. So what's wrong with the idea if you continue to add skill (or sit mods or magical buffs, etc), eventually crits become more common then failure, and ultimately more common than regular successes? Fuck, when I play, there are some sit mods I apply directly to crits. E.g, got a 45% chance to hit, 5% chance to crit (50% total chance to at least hit), here, pick up this magic sword, then you'll have 45% chance to hit plus 15% chance to crit (60% overall chance to hit). This ADDS a dimension to the game, because if you have the choice between doing one thing that gives you a 20% regular sit mod vs doing something else that gives you a 10% crit mod, it's not automatically clear which is better. It's far more nuanced than the sterile mechanics game designers slave over.
4) As you sort of seem to get at, virtually every crit system is a generic pile of shit to begin with because it's baked into the mechanic and not specific to the situation. Not every situation has an "even better" option--especially if it's something that needed to turn out a really specific way to achieve the player's goal. Like if you're trying to KO an NPC but crit and kill him by accident. That is not a better result for what you were trying to do. KO by regular success is exactly what you wanted. No crit makes sense in that case. It's got to be treated as just another die roll. On the other hand, some situations present themselves with highly likely criticals. Stab the beholder in the large central eye? Doesn't seem like that small of a target.

So given all of that, let's NOT discuss goofy probability nonsense. Let's 80/20 this. Let's just say all that matters is we have a simple mechanic that works in 80% of cases, don't sweat the tails, and invest that time in actually creating interesting situations for the main game to play out. THAT'S how you "get a good narrative from rules of simulation." Now I know the odds of anyone taking that advice are slim. But that's EXACTLY why we keep getting topics like these, where people who act like playing a game of make-believe is a challenging problem that requires 100+ pages of discussion. Because, in the end, building a better core mechanic just doesn't work.

... from an adventure I'm currently writing.

The party has just crossed the river Acheron, two football field lengths away stand the gates to Hell, where the party is headed in search of a relic. The field in between is anything but pleasant. The ground shakes, flames spring up from the ground like geysers and jet through the air. This much is plain. What the party may not know, this is also the land where the wicked buried in shallow graves go. Treading on the ground is likely to wake the dead.

So here are my "rules of simulation"
-The standard rules of the RPG apply.
-Every round the party treads on this land, a zombie arm bursts through the ground near one member selected at random and attempts to grab that person.
-Attacking thusly by surprise in a way that is not predictable reduces the odds of a "miss" by 20%, that entire amount enhancing the odds of a "crit". A crit on the grab indicates the hold can't be broken, except by destroying the zombie (or at least its arm).
-The first time this happens, when it is completely unexpected, this modifier is double.
-It takes zombies 4 rounds to completely unearth--they seek to hold their victims in place while they dig themselves out, and only then do they use their normal attacks
-Every 4th round, one of the flames of this place jets out near the party. 1-3 party members may potentially be hit, but may avoid being hit with a successful avoidance check.
-Every 7th round, begging on round #2, the quaking reaches a crescendo. Each party member must make a precision check to keep balance. Those falling take 1-10 harm, fall prone, and may not act that round. Those who keep their balance may still take an action but the odds of success are half normal that round.
-The quaking also unearths 1-3 zombies, who do not attack by surprise but fully emerge from the earth in just one round.
-Whenever any character falls prone to the ground for any reason--losing balance during a quake, being tripped by a zombie, or diving to avoid an attack--there is a 50% chance one additional zombie will burst from the earth and attack by surprise that round
-Once reaching the gates, they'll have to figure a way to open them. Physical force works provided at least one pry bar is used and a combined Physique of 250+, or through any kind of magic that opens or unlocks doors or opens up passage ways, though additional energy needs to be invested such that the total energy cost is at least 13. Otherwise the magic just fizzles.
-Also, odds are the timing works out such that once the party reaches the gates, the first few zombies downed have fully regenerated and are getting back to their feet. And over the next few rounds, the next few zombies down will rejoin the fight. The time it takes them all to get where the PCs are acts as sort of a count-down for figuring out how to open the gates.

From the above, a pretty vivid story emerges. There's enough randomness that it's going to play a little different every time. There are also some broad choices players could make. Retreat? Stay and fight? Press on towards the objective? Also how much team work will they use? Do you take time to free any comrade grabbed by a zombie, risking the same fate for more party members? Or do you allow one to be a sacrifice for the benefit of the party as a whole who opportunistically moves along while one distracts the zombies? And of course this is just one encounter.

Lunamancer, it is you who is discussing from isolation in a space of vague abstraction. It is you who is presenting vague and poorly though out counterpoints like the addition of social security numbers and the airplane delay examples, examples which were simple erroneous or already contemplated. In the particular case of the airplane delay you talked about bell distributions and how delays could not be represented in such a way, to which I simply responded with a die roll mechanism that indeed behaves in such a way, and I'll respond with a few more in this post to address your "italicized 'S" description. To criticize you should first learn what others can, and are doing, and not reduce yourself to criticism just because you can't do what they're doing.

Quote from: Lunamancer;9007831) Linear vs Curve systems. The differences are highly exaggerated. If you plot the probability of success curve in any curve system, you get something that fits the general shape of an italicized 'S' whereas presumably the linear system gives you something that looks like a slash '/'. I say "presumably" because there is a reality that you can't avoid in game design--namely that probabilities are bounded by 1 and 0. So whether you let chance of success/failure go straight to certainty, or whether you put in a ceiling/floor (be it 1%/99%, 5%/95% or anything else), the "curve" ends up looking more like a squared off lightning bolt. And if you do like many systems do, have exceptions to the main mechanic at the extreme highs and lows, the final "curve" ends up extremely close to the italicized 'S', just a little more squared off.

It's more like you can't grasp the differences and thus believe everyone else is exaggerating. First of all, the famous S. Take a look at these curves:

[ATTACH=CONFIG]97[/ATTACH]

We see your famous "italicized 'S" in orange, and the "/" (slash) in black, the former being a 5d4 roll and the later a d20. Yet there are others that look nothing like an S or a slash, and if they do seem to you then your handwriting is considerably worse than mine. So no, your claim that "If you plot the probability of success curve in any curve system, you get something that fits the general shape of an italicized 'S' " is simply false.

The key difference between one and the other (curve vs linear) is that the probability steps are not the same for the same amount of point increments. That you fail to see this and address this when in the same post you address (and champion) the Pareto's 80/20 rule is beyond my comprehension. As you can see on the yellow line the first four steps account for 80% of the whole probability curve and the rest (up to 18) account for the rest. So adding a +4 bonus provides a huge boost within the lower values, but hardly so afterwards. Not the S shape in any of its variants: italics, bold, underline or strike-through. It seems you want everyone to do 80% of the job with 20% of the stuff, except your characters. They can't seem to be allowed to get 80% of the bonus out of 20% of their modifiers, they need to keep adding and adding because you love linear mechanics in game, but champion non-linear behavior out of game.

Quote from: Lunamancer;9007834) As you sort of seem to get at, virtually every crit system is a generic pile of shit to begin with because it's baked into the mechanic and not specific to the situation. Not every situation has an "even better" option--especially if it's something that needed to turn out a really specific way to achieve the player's goal. Like if you're trying to KO an NPC but crit and kill him by accident. That is not a better result for what you were trying to do. KO by regular success is exactly what you wanted. No crit makes sense in that case. It's got to be treated as just another die roll. On the other hand, some situations present themselves with highly likely criticals. Stab the beholder in the large central eye? Doesn't seem like that small of a target.

This particular paragraph has to be looked at from the point of view of "every crit system" you know of. Since for you all curve systems are like italicized 'S, and I've show that there are more than just italicized 'S curves, it's quite clear that "every" is a very limited set when it comes to you. So "every" actually refers to the very few you do happen to know.

Let me convert your d20 + damage to a single graph of 0 damage (miss) or odds of certain damage and compare this with some single roll non italicized 'S curves.

The following graph shows the d20 rolls  + damage (I need a 16 or better to hit) vs a simplified roll.

[ATTACH=CONFIG]98[/ATTACH]

The green line is d20 over 16 to hit plus 2d8 damage. So clearly you see a 75% chance of 0 damage as a green dot at the base of the blue line. The odds of delivering 2 hp or more is thus 25%, of 9 hp or more is 14% and so forth. This is similar to the orange line that produces the same amount of damage with about the same odds. It can do up to 16 points of damage as well, but can also deliver 17 at a 0.06% chance, which I can interpret as a kill instead of a KO. Although the roll to obtain the orange curve is too complex for my taste it comes to prove a point. I can convert a discontinuous mechanism like the d20 + damage into a continuous curve with a long tail that looks nothing like an italicized S. I can then obfuscate the whole odds thing by just requiring a target value large enough and the mechanism becomes continuous instead of plagued with exceptions.

That's one of the key features of the d6 system I presented and also a key limitation when compared to the unbounded exploding 3d8. I can set success starting at 0 and going upwards. It is possible to kill someone with your bare hands, it's just uncommon, but I can ask for a 15 or 20 instead of a 5, 8 or 10. As skill and power increases and since the mechanism is of opposing die rolls the better the attacker the higher the odds of a killing punch. It flows naturally and has a slow increase at first and then in rises quickly to the point that an inexperienced fighter might not kill on a punch, but a trained assassin that knows how to hit to kill can.

The unbounded 3d8 exploding dice is event better. Can my character jump over the Moon? Sure, roll 128. It's possible although highly, very, very, highly unlikely, and as you say probably useless, but shift the curves right far enough for a god or demi-god and it becomes all too possible. It's the same curve, no exceptions.

The issue you have is that your bounded by a limiting lower value be that 5% for d20 or 1% for percentile and that can be too much in some situations so you come up with the whole KO example which affects and concerns you, but not me. Because the statement:

Quote from: Lunamancer;900783building a better core mechanic just doesn't work.

is simply false. Building a better core mechanic does work and it works real well. It's just that you can't build it and you thus want everyone else to abstain from building it because you don't know curves beyond italicized 'Ss

There are systems and players, and game designers publishing games, where the core combat mechanic is roll 1d20, and it's a critical success on a 20, and critical failure on a 1, in systems where there are many attack rolls per combat, and there are no modifiers to crit chance. The problem with this, is that 10% of all attacks are critical something, and nothing in the situation affects that.

Meanwhile, the most common attitude I've seen from players, and one which I can relate to, is that they want there to be a chance of extraordinary events, they want them to be fun and exciting, and the idea of them (or "narrative" in the highly-contested wording of this thread) tends to involve some sort of reason for what the critical is about. When someone crit fails, they tend to deride the character or make up things about how that character messed up or is a klutz, while when there's a crit success, they tend to want to think of it in terms of someone doing something particularly skillful or badass or mighty or something. But all that imagination and desire and potential fun is undermined when someone notices that the mechanic is entirely about blind luck, and has nothing to do with the characters' skills or the circumstances. So that occurs as lame, to some players.

And I've seen other variations for other game systems. I find it important enough to try to make some good rules for it. Not that the core combat system shouldn't also be good.

I'm all for systems that allow for unusual results at extremes of the dice curve, and I don't find the idea entirely unrelated to the core topic, as the POSSIBILITY of unusual events/outcomes lends interest to a narrative for me (as opposed to "unusual" outcomes being directly authored into the game as you see in games like Fate where the supposedly unusual quickly becomes predictable and passe).

But how many games have that built into the system? I mean, in combat 'critical hits' do more damage. But what exactly does a critical Farming, or Scribe or whatever sort of roll do?

I guess typically, a really good farming roll might indicate s greater yield from the fields... but that isn't terribly narratively compelling or even all that interesting on the level of simulation. I want results like, you raise a crop of blueberries that heals certain forms of illness, or one of your pumpkins tops out at 200 lbs. Or whatever.

But there are basically only going to be two ways to get this.

1. You have some kind of table of critical results for lots of different skills.

2. If there is an extreme roll, the GM makes up an appropriate feeling result on the spot.

While the first method is closer to the premise of the thread, I am not sure how feasible of a method it really is.

Any thoughts or suggestions?

Quote from: Skarg;900927The problem with this, is that 10% of all attacks are critical something, and nothing in the situation affects that.

Yup

[ATTACH=CONFIG]100[/ATTACH]

Quote from: Manzanaro;900965But there are basically only going to be two ways to get this.

1. You have some kind of table of critical results for lots of different skills.

2. If there is an extreme roll, the GM makes up an appropriate feeling result on the spot.

While the first method is closer to the premise of the thread, I am not sure how feasible of a method it really is.

Why is the first method closer to the premise of the thread?

Quote from: Saurondor;900969Why is the first method closer to the premise of the thread?

Because what the second rule is doing is just giving the GM narrative authority to say what happens.

Quote from: Manzanaro;900978Because what the second rule is doing is just giving the GM narrative authority to say what happens.

But once again it can fall into authoring by proxy whereby I pull a table out of my ass and publish it, and then everyone else simulates my authoring. If the GM is an expert in the subject matter then GM's on the spot ruling may be more "simulationist" than the table itself. Not to say that the table or a set of information in similar format is bad. It is a great resource to transfer information from the setting creator (designer) to the user (player), particularly if the case of fantasy settings in which it easy to create something unknown to everyone else, but to claim that one is better than the other and generalize it is not correct. For example I can create a particular creature on the spot and I can make up an "appropriate feeling result on the spot". This is simply an "undocumented" version of 1).

No. The distinction is that if you roll on a table you are not CHOOSING the outcome. Why is this so hard for you to understand after all this time?

When you roll to hit is that 'authoring by proxy'? Why not just have the GM choose whether the attack hits instead of leaving it in the hands of whatever jackass wrote the rules?

Seriously... You seem to just totally miss the fundamental distinction between simulation and narrative, which makes conversation on the subjects with you feel very pointless and frustrating.

And again you start talking about me claiming one is 'better' than the other. Will you fucking PLEASE stop putting words in my mouth. How many times do we have to go through this?

Quote from: Skarg;900927There are systems and players, and game designers publishing games, where the core combat mechanic is roll 1d20, and it's a critical success on a 20, and critical failure on a 1, in systems where there are many attack rolls per combat, and there are no modifiers to crit chance. The problem with this, is that 10% of all attacks are critical something, and nothing in the situation affects that.

Two things. First, I think Manzanaro nails it when he talks about how these things often have built-in meaning in combat but not other skill areas. This is why I don't like degrees of success built-in to the mechanic. The age hold hit roll + damage roll is brilliant in that capacity. The damage roll is your de facto degree of success roll, but the system only has you roll it as a standard thing when, as a standard thing, it has meaning.

Second, one thing I am very particular about is to distinguish between what we call things in game terms, what the actual effect is. Say we're playing old school D&D with the house rule that a natural 20 is a crit, and crits basically do double damage. You've got your fighter with a long sword that does d8 damage. On a normal hit, you will do 1, 2, 3, 4, 5, 6, 7, or 8 damage. On a crit you will do 2, 4, 6, 8, 10, 12, 14, 16. So notice right off, half of your crits are producing numbers that fall within the range of a normal hit. We only call them crits because you rolled a natural 20. As far as the math is concerned, though, only half of those natural 20's are really doing something beyond ordinary. So while you think the game is having you crit 5% of the time, in terms of numerical effect, you only really crit 2.5% of the time.

So I'd like to suggest that the probability of producing an effect above and beyond a normal hit is different from what the RPG says it is on the face. I'd also like to suggest that there IS a degree of situation-dependence to this. If you're fighting a goblin with 7 hit points, I would argue a damage roll of 7 or 8 on what the game terms a normal hit is pretty damn critical, as it slays the goblin. So against weaker foes, your "crit" chance as determined by effect (rather than system naming convention) can become far more probable. On the other hand, against stronger creatures, say a 200 hit point demi-god, even a max damage crit doesn't even amount to 10% of the guy's hit points. There is essentially zero chance for a real crit. A natural 20 is a CHINO--critical hit in name only.

QuoteMeanwhile, the most common attitude I've seen from players, and one which I can relate to, is that they want there to be a chance of extraordinary events, they want them to be fun and exciting, and the idea of them (or "narrative" in the highly-contested wording of this thread) tends to involve some sort of reason for what the critical is about. When someone crit fails, they tend to deride the character or make up things about how that character messed up or is a klutz, while when there's a crit success, they tend to want to think of it in terms of someone doing something particularly skillful or badass or mighty or something. But all that imagination and desire and potential fun is undermined when someone notices that the mechanic is entirely about blind luck, and has nothing to do with the characters' skills or the circumstances. So that occurs as lame, to some players.

This is probably my biggest beef against crits and fumbles (which, to be clear, I DO use crits and fumbles, and find they work well the way I use them). You know what you describe feels like to me? Playing a video game where for the most part you're controlling the character in real time, but as soon as you do some magical combination, the character goes on auto-pilot for a second or two to do his cool move, only after it's done do you resume control. So for the most part in RPG combat, we know how it works and feel a great deal of control through our decisions. But, fuck, as soon as the "fumble" comes up, I'm suddenly spinning 180 degrees and plugging the party mage.

To me this is not fun for a few reasons. One, as I was alluding to, at that point I'm not actually controlling my guy. He's just off doing goofy shit 'cause the dice said so.

Two, it even undermines agency in the bigger picture. I alluded to in an example a couple pages ago, if you're at the firing range, let's say, certain precautions are taken to eliminate the chance of accidentally shooting someone. Of course this is at the expense of being able to do cool shit like shit the chandelier to make it drop. You have to stick to the boring old paper target. But that's the trade off. You give an inch of what you'd like to do to eliminate even the small probability of a mile's worth of damage. I mean, in the shooting-the-mage fumble, my guy's an archer, he'd be well suited in behind the party so enemies can't get at me. So why am I shooting from in front of the mage, exposing myself to the dangers of my enemies? To avoid a tragic friendly-fire incident. Want to say my bow string breaks? Fine. That can be coded into the system. Want to just throw out this abstract thing called a "fumble" and require even the totally illogical to happen just because the dice told us something bad needs to happen is, quite frankly, stupid. Which segues into the third point. It hurts suspension of disbelief. Sadly, however, that's how probably the majority of RPGs are designed.

Quote from: Lunamancer;901031Two things. First, I think Manzanaro nails it when he talks about how these things often have built-in meaning in combat but not other skill areas. This is why I don't like degrees of success built-in to the mechanic. The age hold hit roll + damage roll is brilliant in that capacity. The damage roll is your de facto degree of success roll, but the system only has you roll it as a standard thing when, as a standard thing, it has meaning.

There are a couple of issues with this point. First of all, and as I've shown previously, a hit roll+damage can be emulated by a single die roll with degree of success. If I hide the actual die rolling from you and just output the values generated you'll feel like it's d20 + 1d6 or d20 + 2d8 and like it as much as the real thing while being unaware that internally it's a degree of success mechanic all along. Which means you are actually liking a degree of success mechanic.

Damage, as the destruction of the Death Star in A New Hope shows us, is a function of the hit location and not so much the weapon. A well placed dagger is more effective than an ill placed long sword. In many games damage is a fixed range sometimes modified by strength, but hardly by skill. Sure, I might have some feat that improves it, but overall once I hit the actual hit location is not that relevant to the damage done. This is mainly due to the limited range of the d20 and the lack of long tails you're so quick to disregard. If my peasant can only hit a knight on a 19 or 20 then there is little room to express damage there. It's not that "The age hold hit roll + damage roll is brilliant in that capacity.", it's that for the age old d20 that's the only capacity it has, brilliant or not. On the other hand if the 10% is spread out over 15 values ranging from 19 to 34, each with smaller and smaller odds (<1.0%), then it is possible for such values to actually represent damage and the have the attack be resolved with a single roll.

Now for a little graphing. The following image shows the opposing Nd6 mechanics I've shown before and a d20 + damage roll graphed as a single roll. The actual damage roll is a 2d6.

Notice how for skilled vs skilled and experienced vs experienced (black and orange lines respectively) the damage (rolls above 0) behaves quite similar to the d20 + damage (green line) for a required roll of 16 or better. Now the experienced vs skilled roll automatically shifts the curve right due to the higher skill of the experienced PC over the skilled NPC. This cyan curve behaves similar to a d20 + damage when the required roll to hit is 13 or better (lower than before because the PC is more experienced than the target NPC) and the damage is taken to 2d6+2, enjoying a bonus to damage for higher skill.


[ATTACH=CONFIG]101[/ATTACH]

Notice how all the d6 curves go on and on. So on the cyan line I can easily simulate critical hits for 15, 16, 17, 18 and more points of damage. There are very small, but yet possible, odds of delivering truly devastating amounts of damage in the range of 41 hp in a single strike (about triple damage). The key here is that the curve allows for more than "one crit" outcome. I don't have to decide what 20 means, I don't have to try to cramp many outcomes into a single value, sort them out or pick the best. I can have more than one crit outcome available for me at any given time.

Last, but not least, there's the usage of mechanics outside combat. If degree of success is set by a die roll for hp then other skills must have hp too? Farming points, and litigation points, and negotiation points, and fun points, and entertainment points, and what not. Question for you, since you seem to like damage rolls so much, why do other skill usages lack the equivalent of hit points and just reduce the outcome to a success or failure roll?

Quote from: Manzanaro;901025No. The distinction is that if you roll on a table you are not CHOOSING the outcome.

I get this and I agree. It's a tough pickle, though, as you illustrated before. I can't imagine the amount of effort it would take to come up with tables upon tables of criticals for all kinds of skills and situations. There would by necessity be a lot of holes left that would need to be filled by the GM's imagination, even in the most simulationist of simulationist systems.

As a thought experiment (and sorry if this has come up already), imagine that instead of a human, you had a very advanced AI inference system that did not work off of crit tables, but was connected to the internet at large. Imagine that it could somehow synthesize creative responses appropriate to the situation from speedily trawling the internet, would you consider that simulation or not simulation? I'm just trying to see where you might draw the line.