The purpose of the rules

[Lord Raziere]

On Freeform: My early role-playing experience was actually entirely in free-form. And it can work, you just have to make up the missing system knowledge with setting knowledge. From what technology and/or magic is available in the setting, to an idea of the tone and the power of the protagonists.

My experience with freeform is that it works much better for PbP than system-based roleplaying because you don't have to go through the arduous process of rolling and can think out your actions, thus every post can have a reasonable response as long as you think it through.

the worst freeform I've ever seen was one where everyone tried to make themselves as invincible as possible, with only a single obscure weakness for each person so battles became long drawn out affairs full of nothing but counters and counters to the counters, and long OOC arguments about the battles. Didn't have a GM, so players being powergamers was the problem

As long as you don't powergame in freeform, your good.

[Thrudd]

In D&D we had a problem: when critting against normal humans we described the hit as a lethal hit, but then the enemy still had HP. The rules forced the fiction to go in a direction we tought was wrong.
...
The problem is that many rules are important in some situations, while in others they are just a drag. So if we ignore them we are playing the wrong game, but if we switch to a simpler ruleset what happen when we are in the first type of situation and we don't have the rule ? The idea that "stuff don't happen because the rules don't support it" is something that I can't accept. The rules should follow the fiction, not the other way around.

That problem means you either need to change the way you narrate, or change the rules. Add rules for specific injuries and hit locations. Change the way critical hits work (auto kill instead of just extra damage, if that's what you think it should be). Add a rule that says the GM can fiat injuries, long-term penalties or restrictions to actions at their discretion (not a great rule for fairness, but at least the players would have a heads up).

The rules and the narrative should never work against each other. If they aren't in-sync, you should examine that and fix it. They should help you narrate what and how you want to narrate. It isn't that "stuff doesn't happen if the rules don't cover it." If the rules don't cover it, then it's a thing that shouldn't affect game play - it isn't a strategic consideration for players in creating characters or taking actions. If there are no rules for injured hands, it doesn't mean you can't describe an injured hand- it means injured hands don't affect the things characters can do in the game. You can describe a hurt hand, but maybe like an action hero the character can power through it and use it anyway. That's the type of game/story that rule-set would promote.

[Jormengand]

The point which I think is being obnoxiously and deliberately ignored is that the rules are supposed, whether 3.5 is good at it or not (it is; do the work: extrapolate) to provide players with consistent results for consistent actions. You shouldn't have to ask the DM "Can my character climb buildings like the Brotherhood and the Memory Hunters can?" Your character should have a set of rules which tells them whether they can climb walls like an Assassin or an Errorist can, and how likely they are to succeed if they attempt that, and what happens if they fail. Not just "The DM says it happens so it happens."

But fine, whatever, create a game where the only rule is "What I say goes". See how far that gets you.

[Max_Killjoy]

The point which I think is being obnoxiously and deliberately ignored is that the rules are supposed, whether 3.5 is good at it or not (it is; do the work: extrapolate) to provide players with consistent results for consistent actions. You shouldn't have to ask the DM "Can my character climb buildings like the Brotherhood and the Memory Hunters can?" Your character should have a set of rules which tells them whether they can climb walls like an Assassin or an Errorist can, and how likely they are to succeed if they attempt that, and what happens if they fail. Not just "The DM says it happens so it happens."

I have to generally agree with that, especially the part I bolded.

[Knaight]

The point which I think is being obnoxiously and deliberately ignored is that the rules are supposed, whether 3.5 is good at it or not (it is; do the work: extrapolate) to provide players with consistent results for consistent actions.

The quality is relevant here. If the rules aren't good at it (and they aren't), then a lot of the alleged advantages are lost.

You shouldn't have to ask the DM "Can my character climb buildings like the Brotherhood and the Memory Hunters can?" Your character should have a set of rules which tells them whether they can climb walls like an Assassin or an Errorist can, and how likely they are to succeed if they attempt that, and what happens if they fail. Not just "The DM says it happens so it happens."

This is the primary point of contention. I don't need these rules. I don't wan these rules. I have consistently had more fun, as a player and GM, in systems where we didn't deal with these rules. A general climbing skill rank and one general difficulty table is enough for me, and I've found that it works better.

[PhoenixPhyre]

I like having rules, but I differ with a few things here.

Your character should have a set of rules which tells them [1]whether they can climb walls like an Assassin or an Errorist can, and [2]how likely they are to succeed if they attempt that, and [3]what happens if they fail.

[1] This should be a tunable parameter, as long as the system isn't setting/tone-locked. For a multi-setting system, the information should be available at session 0. "For this game, we're running in action-hero mode. Things like Assassin's Creed-style parkour occur regularly" vs "For this game, we're playing dark-and-gritty mode. Expect to be limited to 'regular people' stunts." and then have some mechanical backing.

[2] As long as exact precision isn't required, I'm fine with this. Knowing "I have a 25% chance to succeed" is, to me, very meta. Real people don't know that. They know "can succeed without a challenge (usually because I've done this exact thing before repeatedly)", "easy but a chance of screwing up", "challenging but doable", "difficult but possible with luck", or "I will only succeed if I get really lucky." And this is at best. Humans are bad with assigning probabilities to things. DMs should telegraph the rough difficulty, but knowing the exact numbers ahead of time is unnecessary, as is knowing all the factors that go into the DC calculation.

[3] Consequences of failure depend strongly on the situation, so encoding this in the actual mechanics is very difficult and failure-prone, IMO. The consequence of failing a climb check in one case may be that you never start climbing; in another it means you fall to your certain doom. And this depends on the fictional situation which can't be given on a table somewhere.

As a general rule, rules that require multiple table lookups (especially conditional table lookups) and chains of logic (If X, then Y, but if Y then Z, but only if not Y') make me not want to use a system. Things that require strong consistency, have high consequences, and are outside the normal experience of prospective players (e.g. combat) should have the most well-defined and nailed-down rules. Things that have weaker consequences or are not consistent (not all "rough walls" are the same!) or are within the normal experience of the prospective players can have much looser rules with only a framework to guide DMs.

[Segev]

I'll just third (or fourth, or whatever) the point that the purpose of rules in an RPG are to allow players to have shared expectations of what the results of their actions will be. They may be uncertain, but the uncertainty will be bounded by parameters they can judge based on the rules. "Can Billy the Bat get across the chasm?" is a question that should be answerable by examining his character sheet and knowing a few details about the width of the chasm and any extraneous terrain or environmental features (such as up/downdrafts, if important).

If Billy the Bat can fly, his sheet should say so, and the answer of "yes, of course he can" would be obvious. If he instead is a gadget-based superhero, then rules for what his gadgets can (and can't) do should inform whether he definitely can, has to make some sort of roll, or definitely can't. If Billy the Bat is just a really awesome athlete, estimates of minimum and maximum distances he can leap based on a roll should, when compared with the chasm's width, tell us whether he has a chance to succeed or fail, or if the results are certain one way or the other. (A 2-foot-wide 'chasm' is a guaranteed success; a mild-wide chasm is a guaranteed failure for even a comic-book "normal human.")

The reason the rules are there is to let the players make this kind of determination without having to ask the GM whether they can or not. To take guessing out of it.

To make it so that the game of cops & robbers with a judge doesn't just amount to asking the judge, every time, if he decides the shooter hit the target or not.

[BlacKnight]

Well, first I'd examine some of those choices and ask 'is this rule really doing what I wanted it to do?'. Like, when you describe all crits as lethal, why did you make that choice? What was the intended purpose of making that (narrative) rule, as opposed to just describing a critical hit as, say, a strike to a weak point or just any particularly effective strike?

We tought a normal human should die to the best hit possible with a weapon. The rules didn't work with the fiction we had in mind.

Similarly, lets take the example of the wounded enemy. If we go by the breakdown I described before, a disabling shot is defined as 'injuring in a way such that they are out of the fight'. So, ergo, even if you didn't shoot them in the hand they're succumbing to shock or they've passed out or they're writhing in pain or something like that, at least for the next 30 seconds or however long fights last. If they aren't, then what happened wasn't a disabling shot but rather a graze that the enemy is playing up to make it look like they were taken out. Now lets say the fight finishes, but there's some issue that later on requires you to know where the enemy was shot - for example, maybe you take them captive and you need to know if they can march or if they need to be carried or something like that. In that case, you can simply leave it up to the person who shot them to decide post-facto where the hit was - even if its a metagame advantage to be able to decide that after the fact, its not an advantage which will come up often enough to really cause more problems than the ability to abstract hit location (which could come up all the time) will solve.

As to 'the group who wants detailed hit locations' again I ask - what is this level of detail supposed to accomplish? It shouldn't be there just because it could be there, it should be there because there is some way that the gameplay engages with it. That means that it needs to be something which is sufficiently non-random that it can be part of strategy and planning. If this isn't going to be the case, its probably a mistake to have rules about hit locations at all, because those rules aren't actually doing anything or accomplishing any design goals.

Why I want hit location ? Let's see: why would I want a system where the combat is divided in actions, instead of having a single roll deciding the entire fight ? I would say because I think that level of detail is interesting. So on the same page I would say that having to deal with specific wounds and penalties is interesting. Even if it's random. I mean the rolls to hit are random too.

I think I have a problem with the level of detail. Basically I like detail. So for any situation there are going to be situational modifiers. One time could be "the range of your weapon is not enough to hit that target" the other "your weapon can penetrate that door" and the other again something about hit location. These things are interesting in that specific situation, but not in the majority of other situations.
So I should make rules only for the common situations, which could be (for example) rules for combat rounds, initiative and rolls to hit. But then a lot of situations would have GM calls that modify the basic rules.
But following your line of tought I shouldn't do that: the rules are for players to understand stuff and plan around it.
So I have to ignore that level of detail. But I don't want to do that. We like those details. They are important to us, even any one of them comes out rarely.

If something like that is in the rules, it lets players know what it's going to be like to play a time traveler in this game and to what extent they will be able to do classic time travel shenanigans.

I have to say your time power description is quite good and detailed. I wish RPG designers made more stuff like that instead of "this power makes XD6. Go figure what actually happens".

That problem means you either need to change the way you narrate, or change the rules. Add rules for specific injuries and hit locations. Change the way critical hits work (auto kill instead of just extra damage, if that's what you think it should be). Add a rule that says the GM can fiat injuries, long-term penalties or restrictions to actions at their discretion (not a great rule for fairness, but at least the players would have a heads up).

The rules and the narrative should never work against each other. If they aren't in-sync, you should examine that and fix it. They should help you narrate what and how you want to narrate. It isn't that "stuff doesn't happen if the rules don't cover it." If the rules don't cover it, then it's a thing that shouldn't affect game play - it isn't a strategic consideration for players in creating characters or taking actions. If there are no rules for injured hands, it doesn't mean you can't describe an injured hand- it means injured hands don't affect the things characters can do in the game. You can describe a hurt hand, but maybe like an action hero the character can power through it and use it anyway. That's the type of game/story that rule-set would promote.

If a thing can't affect the characters is just like it doesn't exist.
If I can describe an injured hand, that means there is an injured hand in the fictional world. If it's a world of heroes that can be a neglectable thing, but if it's not ? I need that injured hand to affect the fiction. So I need rules for it ?
It would seems so by your answer. But you haven't really answered about the problem of rules weight. I mean if I want to promote realistic combat I need a ruleset that simulates realistic combat ? But that would be basically unplayable.


EDIT: I tought of a better example to explain my crux. The PCs want to kill a bad guy that is behind a heavy wood door. Al shoots with his 9mm handgun against the door. Then Bob shoots with his .50 machine gun. There are no rules about material penetration. What happens ?

[NichG]

We tought a normal human should die to the best hit possible with a weapon. The rules didn't work with the fiction we had in mind.

Yeah, you need a different system to make a rule like that work. If you have a system where any hit could randomly kill you, it also has to be a system where the primary tactical element is preventing enemies from even getting a chance to hit you. So it fits more in a Counterstrike-like stealth game than a game like D&D where the system is built on the expectation of standing there and trading blows.

You could also do it in a system where armor or cover can just outright prevent crits, which would get you something that plays kind of like the recent X-Com games.

Why I want hit location ? Let's see: why would I want a system where the combat is divided in actions, instead of having a single roll deciding the entire fight ? I would say because I think that level of detail is interesting. So on the same page I would say that having to deal with specific wounds and penalties is interesting. Even if it's random. I mean the rolls to hit are random too.

Here we get into things that really need to be analyzed from a somewhat 'gamist' point of view though. Having rolls to hit rather than just a straight out deterministic comparison primarily has gamist consequences - in terms of the design purpose, it's twofold: one is that, psychologically, chance is exciting; adding uncertainty to something can make fairly boring strategic elements more interesting because of the way that risk feels to a player. The other is that the alternative would be very binary - always hit or always miss - and so adding a probabilistic element smooths the outcomes and therefore makes it so that improvements or situational modifiers to character abilities can have a finer level of granularity. You could also solve this by folding to-hit into damage and taking average, which actually plays almost the same in terms of outcomes.

The bigger issue of, why divide combat into actions, has to do with making tactical decisions relevant. If combat is just a single roll, there are no decisions that can be made within combat which change things. It would be like playing chess and just looking at the Elo rankings of players rather than having them actually play the game. By breaking combat down into smaller elements, you make it so that there are decisions and evaluations to be made which influence the outcome from within the activity, which ultimately makes those things more relevant to the game and to the story. That's the different between detailing something which has player input (where do you stand, how do you move, what action do you choose?) and something which can just be immediately re-abstracted (roll to hit, then roll for damage; versus take-average or a combined roll method). The second is almost entirely about illusion and perception - you might feel like there's more going on, but actually there isn't.

I think I have a problem with the level of detail. Basically I like detail. So for any situation there are going to be situational modifiers. One time could be "the range of your weapon is not enough to hit that target" the other "your weapon can penetrate that door" and the other again something about hit location. These things are interesting in that specific situation, but not in the majority of other situations.
So I should make rules only for the common situations, which could be (for example) rules for combat rounds, initiative and rolls to hit. But then a lot of situations would have GM calls that modify the basic rules.
But following your line of tought I shouldn't do that: the rules are for players to understand stuff and plan around it.
So I have to ignore that level of detail. But I don't want to do that. We like those details. They are important to us, even any one of them comes out rarely.

The alternative is, you can have that level of detail but it's your responsibility to make sure that it's somehow tactically meaningful - by which I mean, there must be relevant choices tied to controlling those details. In which case the players then can understand it and plan around it. Otherwise you get that 'harmful complexity' effect you were noticing in your first post, where you end up with all of these rules you 'have to have' but they're not actually doing much of anything other than taking time and attention.

If we take your above example of 'crits kill', we can build such a system around that idea.

Spoiler: Strategic Hit Locations and Deadly Crits

Show

Lets start with the idea that whether something a crit is not just a function of the roll, but is also a function of the cover or armor the target has. We'll resolve an attack in the following way: the attacker has a rating which is derived from other stuff, but is generally from 1 to 10. Attackers always make called shots at a bonus or penalty to their choice of: legs (+0), arms (-1), head (-1), torso (+1). The attacker rolls 10 d10's and each d10 which rolls equal to or less than the attacker's rating is considered a 'success' (optionally, on rolling a 1 add another d10 and roll that too, as long as 1s keep coming up). Subtract the defender's location-specific armor and cover from the number of successes, where someone shooting blind from cover has for example arms exposed, someone looking at shooting has head and arms exposed, etc.

On 1-4 successes, the attacker causes damage. On 5-7 successes, the attacker causes damage and compromises the armor at the hit location. On 8-9 successes, in addition the attacker also causes an injury to the hit location which has a corresponding debuff. On 10 successes, the target loses the limb if struck on an extremity, or is killed outright if struck on torso or head.

Armor on arms interferes with aim, armor on legs interferes with reflexive movement and swimming, torso armor is heavy and slows the character, head armor penalizes perception, etc. Armor piercing weapons reduce the effective level of armor by a fixed amount. Rapidfire weapons allow the attacker to subdivide their successes into a certain number of dice-groups based on the weapon type (so if they roll 6 successes, they could deal two 3-success chunks of damage with a Rate 2 weapon) but must apply armor separately to each chunk, etc. Lots of different ways you can play with this kind of setup.

So now there's a couple of potentially interesting choices tied to the detail. A character who really wants to be the best possible scout probably has to go lightly armored in certain ways at least, which leaves them vulnerable to specific attacks that an attacker can choose intentionally. Because of the armor-degrading aspect of attacks, followup attacks to a very successful hit might tend to focus on that location with the tradeoff that the attacker loses some flexibility. Furthermore, you can have automatically lethal crits (10 successes) but characters can ensure that the conditions that would permit them are avoided by e.g. riding around in a tank or making sure to secure cover.

This also answers your door question - ad hoc rule the number of armor points of cover something is worth, but it integrates into the armor system with everything else.

To really make this shine, I feel like it'd need some other thing that is both location-connected and diverse between targets to tie it all together. Location-specific cyberware perhaps.

Anyhow, the point is if you're going to take up time and space in the rules with something, it should be doing some work towards a specific purpose.

[Zombimode]

I tought of a better example to explain my crux. The PCs want to kill a bad guy that is behind a heavy wood door. Al shoots with his 9mm handgun against the door. Then Bob shoots with his .50 machine gun. There are no rules about material penetration. What happens ?

I guess by "there are no rules about material Penetration" you mean that there are no rules specifically called out to deal with material Penetration, right?

But there are likely other rules that are relevant to the context of the Situation.
Thus, how to map the effect of the door against the weapons as well as the differences between the weapons depends on the specific rules language and rules constructs of the System in question.

So lets view your Situation through the lense of some Systems I'm familiar with:

Fate:
The most simple solution is to increase the difficulty of the check of attacking (through the door).
Fate is not big on granularity so there might no difference in the weapons in this situation.
Or the GM could declare an Aspect of the .50 machine gun "overpenetration" or so that could be invoked by the player to ignore the increase in difficulty of the attack.
Or the machine gun could just ignore the difficulty increase per default.
Really depends on the mood of the game.

Savage Worlds:
Since Toughness governs if a hit actually affects (and how much) the target, having a closed door increasing the Toughness against an attack trough the door seems reasonable to me.
The weapons are likely to differ in their damage (and rate of fire), both of which are naturally linked to attacks and Toughness.
The GM has to make up the actual increase in Toughness, of course.

2d20 (Conan, Star Trek Adventures):
Reduce the damage dice or apply a flat penalty on the damage.

d20:
Many Options. The GM could use the provides stats for Hardness and Hit Points of Wood and calculate the difference of the attack's damage to that.
Or provide the target with a Bonus to Armor Class.
Or apply a flat damage reduction on the attack's damage.


The gist of all that is: once you have decided what the door will be, conceptionally, in our case: a hindrance to the attempt of Shooting a target on the other side of the door, most if not all Systems are able to map this into rules.

[Knaight]

We tought a normal human should die to the best hit possible with a weapon. The rules didn't work with the fiction we had in mind.

Different rules are made for different fiction, and the rules aren't necessarily bad because they don't support what you're trying to do with them. They're just the wrong tool for the job. A chisel is an excellent tool, but you wouldn't pound in a nail with it (even if you technically can).

Why I want hit location ? Let's see: why would I want a system where the combat is divided in actions, instead of having a single roll deciding the entire fight ? I would say because I think that level of detail is interesting. So on the same page I would say that having to deal with specific wounds and penalties is interesting. Even if it's random. I mean the rolls to hit are random too.

Cool. So play a system which has hit locations.

I have to say your time power description is quite good and detailed. I wish RPG designers made more stuff like that instead of "this power makes XD6. Go figure what actually happens".

A lot of RPGs do this. There's a concept I like to call qualitative mechanics, where there are game mechanics that do game mechanical things, but they don't work numerically. A lot of RPGs are full of them, D&D mostly avoids them (and the player base tends to hate them when they show up).

If a thing can't affect the characters is just like it doesn't exist.
If I can describe an injured hand, that means there is an injured hand in the fictional world. If it's a world of heroes that can be a neglectable thing, but if it's not ? I need that injured hand to affect the fiction. So I need rules for it ?
It would seems so by your answer. But you haven't really answered about the problem of rules weight. I mean if I want to promote realistic combat I need a ruleset that simulates realistic combat ? But that would be basically unplayable.

If you want to simulate that detail there are some pretty lightweight options to do so. They often rely on those qualitative mechanics (e.g. wound Aspects in Fate), but not always.

EDIT: I tought of a better example to explain my crux. The PCs want to kill a bad guy that is behind a heavy wood door. Al shoots with his 9mm handgun against the door. Then Bob shoots with his .50 machine gun. There are no rules about material penetration. What happens ?

The GM makes a judgement call.

With all that said, you should really check out Nemesis. It's not a super heavy system (although it's heavier than the length of less than 60 pages would suggest, the rules explanation leans towards density), but it covers everything you want. It's got a location specific wounding system with wounds that influence later mechanics. An unarmored headshot is usually a one shot kill, and the best shot is shooting someone in the head, so that meets your normal human dying criteria. It has an explicit weapon penetration stat and weapon penetration table, and in addition to it being generally useful it has details on it that help evoke the implied action horror setting (human shields show up on the penetration difficulty table, as just one example). It's got powers with qualitative mechanics to them.

Also it's free, and it's a quick read. The mechanics are pretty easy once you get your head around them, although getting the basics firmly might take two readthroughs.

[Jormengand]

The quality is relevant here. If the rules aren't good at it (and they aren't), then a lot of the alleged advantages are lost.

True. Rules should be of high quality of course. Obviously.

This is the primary point of contention. I don't need these rules. I don't wan these rules. I have consistently had more fun, as a player and GM, in systems where we didn't deal with these rules. A general climbing skill rank and one general difficulty table is enough for me, and I've found that it works better.

Fine: play roll to dodge or something. I prefer systems where I actually know that climbing a tree won't arbitrarily change difficulty because the DM has a different idea of how climbing works than most people.

[1] This should be a tunable parameter, as long as the system isn't setting/tone-locked. For a multi-setting system, the information should be available at session 0. "For this game, we're running in action-hero mode. Things like Assassin's Creed-style parkour occur regularly" vs "For this game, we're playing dark-and-gritty mode. Expect to be limited to 'regular people' stunts." and then have some mechanical backing.

Right, and this is definitely a thing you can do, but you need a method of doing that responsibly, not just making the DM make stuff up like in 5e.

[2] As long as exact precision isn't required, I'm fine with this. Knowing "I have a 25% chance to succeed" is, to me, very meta. Real people don't know that. They know "can succeed without a challenge (usually because I've done this exact thing before repeatedly)", "easy but a chance of screwing up", "challenging but doable", "difficult but possible with luck", or "I will only succeed if I get really lucky." And this is at best. Humans are bad with assigning probabilities to things. DMs should telegraph the rough difficulty, but knowing the exact numbers ahead of time is unnecessary, as is knowing all the factors that go into the DC calculation.

The point here is that the DM shouldn't have to make up the DC of climbing a rough wall with a few handholds, because it should be in the book. The rules should say what the DC is so that you can have shared expectations. And honestly I don't really mind people knowing the percentage chance because "55%" isn't really more useful than "A little over half".

[3] Consequences of failure depend strongly on the situation, so encoding this in the actual mechanics is very difficult and failure-prone, IMO. The consequence of failing a climb check in one case may be that you never start climbing; in another it means you fall to your certain doom. And this depends on the fictional situation which can't be given on a table somewhere.

Again: no. It shouldn't depend on the DM's decision whether you get partway up, or get all the way up, or don't get anywhere at all, before you fall. The DM shouldn't have to choose how many falling dice they want to throw at your character. The DM doesn't need a rule to tell them to do what they want. They can already do what they want.

[Knaight]

Fine: play roll to dodge or something. I prefer systems where I actually know that climbing a tree won't arbitrarily change difficulty because the DM has a different idea of how climbing works than most people.

The thing is "roll to dodge or something", as you put it, is a category that encompasses a great many RPGs. The idea that it's fine for the GM to determine difficulties isn't an unusual one, and it's well supported. I prefer systems where I can be confident that stupidity like not being able to see the moon because the designers screwed up an explicit penalty on a spot skill doesn't come up. I prefer systems where the game moves along at a decent pace and isn't slowed by constantly having to check tables all the time.

As far as your preferences go, cool. There's games for that too, and there should be. I'm all for having a wide variety of RPG systems. What I'm not for is taking personal preferences and claiming that they're a mandate by which all RPGs should be designed, with RPGs that don't follow them being objectively worse.

[Pelle]

To the OP: Cool thread! In theory I agree if one can accept the premise, that if all players are on the same page with regards to the setting, character abilities, challenges presented and so on, we don't need many rules, since everyone agrees on what should happen. We could even agree on the chance of success if some action is uncertain as well, and roll for it. In practice however, getting to this level of shared understanding is hard, especially when playing games that are involving superhumans. I think it's worth striving for though.

I'll just third (or fourth, or whatever) the point that the purpose of rules in an RPG are to allow players to have shared expectations of what the results of their actions will be. They may be uncertain, but the uncertainty will be bounded by parameters they can judge based on the rules. "Can Billy the Bat get across the chasm?" is a question that should be answerable by examining his character sheet and knowing a few details about the width of the chasm and any extraneous terrain or environmental features (such as up/downdrafts, if important).

[...]

The reason the rules are there is to let the players make this kind of determination without having to ask the GM whether they can or not. To take guessing out of it.

In theory I agree with this too. But what if the rules get too extensive, so that no player is able to memorize everything?

When no one remembers the rules, the players are not able to expect what the results of their actions will be anyways. Then looking up the rules to get the true answer achieves the same as the GM making something up on the spot. Sure, sometimes the players can be proactive and look up the rules before the situation arises, but you can't do that with everything.

[Max_Killjoy]

To the OP: Cool thread! In theory I agree if one can accept the premise, that if all players are on the same page with regards to the setting, character abilities, challenges presented and so on, we don't need many rules, since everyone agrees on what should happen. We could even agree on the chance of success if some action is uncertain as well, and roll for it. In practice however, getting to this level of shared understanding is hard, especially when playing games that are involving superhumans. I think it's worth striving for though.

This need for a shared set of fictional-facts is part of what lead me to my deep interest in extensive, iceberg-model worldbuilding.

However, it is my experience that even when we can get everyone to agree on the facts, what those facts mean can still be contentious, especially when it's a fictional world being considered with all the personal subjective interpretation and gap-filling that people do.

In theory I agree with this too. But what if the rules get too extensive, so that no player is able to memorize everything?

When no one remembers the rules, the players are not able to expect what the results of their actions will be anyways. Then looking up the rules to get the true answer achieves the same as the GM making something up on the spot. Sure, sometimes the players can be proactive and look up the rules before the situation arises, but you can't do that with everything.

Good point, there's a complexity and scope limit beyond which the rules would become effectively as arbitrary-feeling as whim.

We see this in real law too... people end up breaking the law not via negligence or malice or unethical intent, but simply because the body of law becomes so vast, self-referential, and arcane, that no one can know all the laws in detail.

[BlacKnight]

The bigger issue of, why divide combat into actions, has to do with making tactical decisions relevant. If combat is just a single roll, there are no decisions that can be made within combat which change things. It would be like playing chess and just looking at the Elo rankings of players rather than having them actually play the game. By breaking combat down into smaller elements, you make it so that there are decisions and evaluations to be made which influence the outcome from within the activity, which ultimately makes those things more relevant to the game and to the story. That's the different between detailing something which has player input (where do you stand, how do you move, what action do you choose?) and something which can just be immediately re-abstracted (roll to hit, then roll for damage; versus take-average or a combined roll method). The second is almost entirely about illusion and perception - you might feel like there's more going on, but actually there isn't.

The alternative is, you can have that level of detail but it's your responsibility to make sure that it's somehow tactically meaningful - by which I mean, there must be relevant choices tied to controlling those details. In which case the players then can understand it and plan around it. Otherwise you get that 'harmful complexity' effect you were noticing in your first post, where you end up with all of these rules you 'have to have' but they're not actually doing much of anything other than taking time and attention.

Your argument makes me think "We should play FATE". Aspects enter the game only when they are relevant, so you can avoid checking the rules just to add tedious +1 or -1 that have little effect.
In my experience players plan around meaningful details. For example if they can't shoot trough the door they will think of a way to go beyond it or to crush it. If the door just gives a little penalty to damage they will just shoot trough it and the outcome is that we just need more time to resolve the action.
But aspects are tied to GM call. Players can't know in advance that the door can stop 9mm cartridge but not .50 (if not being on the same page with the GM about the setting).

If we take your above example of 'crits kill', we can build such a system around that idea.

Spoiler: Strategic Hit Locations and Deadly Crits

Show

Lets start with the idea that whether something a crit is not just a function of the roll, but is also a function of the cover or armor the target has. We'll resolve an attack in the following way: the attacker has a rating which is derived from other stuff, but is generally from 1 to 10. Attackers always make called shots at a bonus or penalty to their choice of: legs (+0), arms (-1), head (-1), torso (+1). The attacker rolls 10 d10's and each d10 which rolls equal to or less than the attacker's rating is considered a 'success' (optionally, on rolling a 1 add another d10 and roll that too, as long as 1s keep coming up). Subtract the defender's location-specific armor and cover from the number of successes, where someone shooting blind from cover has for example arms exposed, someone looking at shooting has head and arms exposed, etc.

On 1-4 successes, the attacker causes damage. On 5-7 successes, the attacker causes damage and compromises the armor at the hit location. On 8-9 successes, in addition the attacker also causes an injury to the hit location which has a corresponding debuff. On 10 successes, the target loses the limb if struck on an extremity, or is killed outright if struck on torso or head.

Armor on arms interferes with aim, armor on legs interferes with reflexive movement and swimming, torso armor is heavy and slows the character, head armor penalizes perception, etc. Armor piercing weapons reduce the effective level of armor by a fixed amount. Rapidfire weapons allow the attacker to subdivide their successes into a certain number of dice-groups based on the weapon type (so if they roll 6 successes, they could deal two 3-success chunks of damage with a Rate 2 weapon) but must apply armor separately to each chunk, etc. Lots of different ways you can play with this kind of setup.

So now there's a couple of potentially interesting choices tied to the detail. A character who really wants to be the best possible scout probably has to go lightly armored in certain ways at least, which leaves them vulnerable to specific attacks that an attacker can choose intentionally. Because of the armor-degrading aspect of attacks, followup attacks to a very successful hit might tend to focus on that location with the tradeoff that the attacker loses some flexibility. Furthermore, you can have automatically lethal crits (10 successes) but characters can ensure that the conditions that would permit them are avoided by e.g. riding around in a tank or making sure to secure cover.

This also answers your door question - ad hoc rule the number of armor points of cover something is worth, but it integrates into the armor system with everything else.

To really make this shine, I feel like it'd need some other thing that is both location-connected and diverse between targets to tie it all together. Location-specific cyberware perhaps.

Anyhow, the point is if you're going to take up time and space in the rules with something, it should be doing some work towards a specific purpose.

Nice system, altough it rolls too many dices for my taste. Also there is a way to add successes (other than rolling 1) ? Bacause otherwise having a 10 successes kill would be basically impossible.

I guess by "there are no rules about material Penetration" you mean that there are no rules specifically called out to deal with material Penetration, right?

But there are likely other rules that are relevant to the context of the Situation.
Thus, how to map the effect of the door against the weapons as well as the differences between the weapons depends on the specific rules language and rules constructs of the System in question.

The gist of all that is: once you have decided what the door will be, conceptionally, in our case: a hindrance to the attempt of Shooting a target on the other side of the door, most if not all Systems are able to map this into rules.

But the door could also totally prevent the PCs from shooting trough. Either way the GM is making up how the door interacts with the rules. Yes, often following the ruleset style, but even so the difference between -1 to damage and -10 can be massive.

A lot of RPGs do this. There's a concept I like to call qualitative mechanics, where there are game mechanics that do game mechanical things, but they don't work numerically. A lot of RPGs are full of them, D&D mostly avoids them (and the player base tends to hate them when they show up).

If you want to simulate that detail there are some pretty lightweight options to do so. They often rely on those qualitative mechanics (e.g. wound Aspects in Fate), but not always.

Qualitative mechanics is a good definition. I can't think of many RPGs with similar stuff. I suppose Fate, Dungeon World... and nothing else. Do you have some advice ?

With all that said, you should really check out Nemesis.

I will do, thanks. Altough it's maybe a little too heavy.

In theory I agree with this too. But what if the rules get too extensive, so that no player is able to memorize everything?

When no one remembers the rules, the players are not able to expect what the results of their actions will be anyways. Then looking up the rules to get the true answer achieves the same as the GM making something up on the spot. Sure, sometimes the players can be proactive and look up the rules before the situation arises, but you can't do that with everything.

I noticed the same issue too. Usually the GM knows the rules better than the players, which are not interested in checking the handbook in the middle of the session. Thus if the GM doesn't remember a rule he can invent something reasonable and 99% of the cases nobody notices anything. This leads to not studying the rules, which leads to making too many unnoticed mistakes, which leads to the Dark Side to me wondering what's the point in checking the handbook to be sure I'm following the rules.

[Zombimode]

But the door could also totally prevent the PCs from shooting trough. Either way the GM is making up how the door interacts with the rules. Yes, often following the ruleset style, but even so the difference between -1 to damage and -10 can be massive.

Sure.
Your point being?

[Segev]

In theory I agree with this too. But what if the rules get too extensive, so that no player is able to memorize everything?

When no one remembers the rules, the players are not able to expect what the results of their actions will be anyways. Then looking up the rules to get the true answer achieves the same as the GM making something up on the spot. Sure, sometimes the players can be proactive and look up the rules before the situation arises, but you can't do that with everything.

As Max_Killjoy said, that is a problematic level of rules complexity. However, it still doesn't change that there exist rules for those situations, so if the player is trying to prepare a character who will engage in certain activities as part of his concept, he can hunt down the rules and develop his character to be good at it. The rules empower him that way.

Not advocating for all the rules at maximal complexity everywhere. But the purpose of rules remains to tell people what they can do, and how to make it possible to do it.

[NichG]

Your argument makes me think "We should play FATE". Aspects enter the game only when they are relevant, so you can avoid checking the rules just to add tedious +1 or -1 that have little effect.
In my experience players plan around meaningful details. For example if they can't shoot trough the door they will think of a way to go beyond it or to crush it. If the door just gives a little penalty to damage they will just shoot trough it and the outcome is that we just need more time to resolve the action.
But aspects are tied to GM call. Players can't know in advance that the door can stop 9mm cartridge but not .50 (if not being on the same page with the GM about the setting).

I don't like how aspects tend to collapse down into interchangeable roll modifiers in FATE, but something conceptually aspect-like but with more structure might be good.

Nice system, altough it rolls too many dices for my taste. Also there is a way to add successes (other than rolling 1) ? Bacause otherwise having a 10 successes kill would be basically impossible.

It's potentially fewer dice than a high level Fighter's full attack in D&D 3.5, and its probably faster to resolve (since you only have a single roll all at once), but you could tweak the numbers a bit. As I wrote it, you don't need a 1 to get a success, you just need to roll <= your threshold (based on skill or other factors). If you have a threshold of 8, for example, then 10% of your shots are instant kills. For both your concerns, dropping the dice pool to 5d10 might be better. That way a threshold of 5 means that 3% of attacks are instant kills on unarmored enemies. I tend to consider a 1% chance of instant kill per action to be on the high side, since a character might receive something like 2 attacks a session on average, 100 sessions a campaign, meaning at 1% each player is looking to lose two characters to lucky crits regardless of anything they do at that level, but the armor system helps a lot with that.

Anyhow, its just an example.

[PhoenixPhyre]

As Max_Killjoy said, that is a problematic level of rules complexity. However, it still doesn't change that there exist rules for those situations, so if the player is trying to prepare a character who will engage in certain activities as part of his concept, he can hunt down the rules and develop his character to be good at it. The rules empower him that way.

Not advocating for all the rules at maximal complexity everywhere. But the purpose of rules remains to tell people what they can do, and how to make it possible to do it.

But more complex rules come at significant cost even well before you reach unplayable levels. It's more moving parts that have to be balanced against each other, it's more room for strange interactions, it's harder to get into, it takes more effort to DM, etc.

Having defined DC values also strongly limits the scenarios that can be described to those that match the prescribed values. Even within a given setting, not all "iron doors" are the same. If you allow circumstance modifiers that are anywhere near the size of the actual value, you've just got "DM decides" but hidden. There's also strong pressure to make all such entities the same. To me, it feels like playing a video game where everything's stamped out from a uniform backing entity (game asset) rather than an actual world where the variation between "similar" (at the level of the DC tables) things is on the same magnitude as the variation between entities that supposedly vary. I've seen wooden doors that were much stronger than metal ones. And for more complicated things this gets progressively worse. Better to have an expected range (the vast majority of DCs are between 10 and 20, with most in the 10-15 range) than to have these purported exact numbers that either actually vary strongly at DM fiat (circumstance modifiers) or that force all "similar" things to be identical.

But that's just my preference--this isn't an objective right/wrong. Only a "this is better for style X, that's better for style Y, but YMMV" situation.

[CharonsHelper]

But more complex rules come at significant cost even well before you reach unplayable levels. It's more moving parts that have to be balanced against each other, it's more room for strange interactions, it's harder to get into, it takes more effort to DM, etc.

Very true.

This gets into one of my guiding design principles based upon the below three points.

1. Depth is good.

2. Complexity is bad.

3. Depth is purchased with complexity.

Therefore -

Only purchase significant depth for the game's focus/strengths and try to get the best depth/complexity bargain that you can. Streamline everything else as much as you can get away with.

Of course: how much to streamline & how much complexity is okay are both extremely subjective.

[Max_Killjoy]

But more complex rules come at significant cost even well before you reach unplayable levels. It's more moving parts that have to be balanced against each other, it's more room for strange interactions, it's harder to get into, it takes more effort to DM, etc.

Having defined DC values also strongly limits the scenarios that can be described to those that match the prescribed values. Even within a given setting, not all "iron doors" are the same. If you allow circumstance modifiers that are anywhere near the size of the actual value, you've just got "DM decides" but hidden. There's also strong pressure to make all such entities the same. To me, it feels like playing a video game where everything's stamped out from a uniform backing entity (game asset) rather than an actual world where the variation between "similar" (at the level of the DC tables) things is on the same magnitude as the variation between entities that supposedly vary. I've seen wooden doors that were much stronger than metal ones. And for more complicated things this gets progressively worse. Better to have an expected range (the vast majority of DCs are between 10 and 20, with most in the 10-15 range) than to have these purported exact numbers that either actually vary strongly at DM fiat (circumstance modifiers) or that force all "similar" things to be identical.

But that's just my preference--this isn't an objective right/wrong. Only a "this is better for style X, that's better for style Y, but YMMV" situation.

This is where I'd say that DCs (or TNs, or whatever they're called in a system) for things like "how hard is a door of this type to break" should be clearly presented as "typical" or "average" or "suggested" instead of "this is how hard every iron door is to break down".

If the GM needs a "typical" iron door on the fly, they grab that number. If the GM has a specific iron door in mind, they adjust for great construction or for decades of rust or whatever.

[Pelle]

This is where I'd say that DCs (or TNs, or whatever they're called in a system) for things like "how hard is a door of this type to break" should be clearly presented as "typical" or "average" or "suggested" instead of "this is how hard every iron door is to break down".

If the GM needs a "typical" iron door on the fly, they grab that number. If the GM has a specific iron door in mind, they adjust for great construction or for decades of rust or whatever.

But isn't it better if what door is considered hard to break (iron or wood) is dependent on the setting, genre, and a shared understanding by the players, instead of hardcoded by the system?

Again, I guess it depends on how difficult it is to have a consensus, but if possible I greatly prefer that.

[PhoenixPhyre]

This is where I'd say that DCs (or TNs, or whatever they're called in a system) for things like "how hard is a door of this type to break" should be clearly presented as "typical" or "average" or "suggested" instead of "this is how hard every iron door is to break down".

If the GM needs a "typical" iron door on the fly, they grab that number. If the GM has a specific iron door in mind, they adjust for great construction or for decades of rust or whatever.

A large part of this depends on the expected range of difficulties over the entire game is. If I (as DM) am supposed to set an integer DC over a range of 40 or so (ie 3e D&D), I'm gonna need a table of default values. Because each individual point may matter strongly (especially given the range of skill points available).

If, on the other hand, DCs are 90% of the time in the range 10-20, with suggestions of using 10, 15, or 20, it's a lot easier. Instead of thinking "this is an iron door with modifications XYZ, so DC = base + mod_1 + mod_2 + ..." I think "this is a really hard door to break. DC 20". And then players know that the range of expected values is 10-20, with most being 10 or 15.

Edit: And having it vary by setting (or even micro-setting) is also a thing. The dwarves of Fuar-Uulan make much better metal doors than the people of Byssia (who barely use any metal). On the flip side, a well-made Byssian wood door is going to be much better than a well-made Uulanian wood door (because the Byssians have master woodcrafters and tree-growers).

[Max_Killjoy]

But isn't it better if what door is considered hard to break (iron or wood) is dependent on the setting, genre, and a shared understanding by the players, instead of hardcoded by the system?

Again, I guess it depends on how difficult it is to have a consensus, but if possible I greatly prefer that.

My personal preference is that the "average door" breaking DC be set as example in a master list of example DCs, based against how hard it would be for the average person to break the average door. If the person is stronger than average, it will be easier for them to break the door because their great strength is represented in their character build. If the door is better or worse than average for doors, then its DC is set higher or lower.

Personally, I would much rather have a frame of reference (that doesn't cause "WTH?" moments) than operate in a vacuum.

[Tinkerer]

My personal preference is that the "average door" breaking DC be set as example in a master list of example DCs, based against how hard it would be for the average person to break the average door. If the person is stronger than average, it will be easier for them to break the door because their great strength is represented in their character build. If the door is better or worse than average for doors, then its DC is set higher or lower.

Personally, I would much rather have a frame of reference (that doesn't cause "WTH?" moments) than operate in a vacuum.

Yeah, there are a couple of problems which can easily cause this. I call the first one the "5-15" problem. That's where (in a d20 system) a GM will routinely rule that a rolled 5 will fail and a rolled 15 will succeed regardless of other circumstances. Now that is admittedly an exaggerated example (although one I've seen played straight on a number of occasions) but the more common example that I've seen is that if you have a character with a +19 to a task who rolls a 3 and a character with a +2 to a task who rolls a 19 several GMs will rule that the character who rolled a 3 failed and the character who rolled a 19 succeeded.

The other one is "Oblivion scaling" where the GM will scale things which really don't make sense such as a stuck rotted non-magical door in an epic level dungeon having a higher bash DC than a well constructed locked door in a 1st level dungeon.

I've seen both of these examples come up more often than I would care to admit.

[Pelle]

My personal preference is that the "average door" breaking DC be set as example in a master list of example DCs, based against how hard it would be for the average person to break the average door. If the person is stronger than average, it will be easier for them to break the door because their great strength is represented in their character build. If the door is better or worse than average for doors, then its DC is set higher or lower.

Makes sense. I guess I kind of do the same in practise. I prefer 'realistic' games (not a fan of superheros or god-wizards), so I just calibrate the DCs on what is possible for normal people, world records, real physics and so on.

I realize it comes down to for me; when I estimate a DC, it will either roughly match what the system rules say or it will not. If it matches, good, but what is the point of having an extensive ruleset then? If it doesn't match, the system is bad, since it produces nonsensical results

I know, this isn't helpful for making the players understand what their characters can do. Though, I'm assuming that if I judge a task to be 'difficult', it is more likely that other people will also do that than them considering it 'easy'...

[Max_Killjoy]

Makes sense. I guess I kind of do the same in practise. I prefer 'realistic' games (not a fan of superheros or god-wizards), so I just calibrate the DCs on what is possible for normal people, world records, real physics and so on.

I realize it comes down to for me; when I estimate a DC, it will either roughly match what the system rules say or it will not. If it matches, good, but what is the point of having an extensive ruleset then? If it doesn't match, the system is bad, since it produces nonsensical results

I know, this isn't helpful for making the players understand what their characters can do. Though, I'm assuming that if I judge a task to be 'difficult', it is more likely that other people will also do that than them considering it 'easy'...

To expound on the part in your first paragraph -- if a game using this system does end up with a character who is a superhero or demigod (and thus possessing or able to access fantastic strength), then their attempt (roll, whatever) will be able to easily overcome the DC of a normal door, maybe even to the point that they automatically succeed.

[Quertus]

Yeah, there are a couple of problems which can easily cause this. I call the first one the "5-15" problem. That's where (in a d20 system) a GM will routinely rule that a rolled 5 will fail and a rolled 15 will succeed regardless of other circumstances. Now that is admittedly an exaggerated example (although one I've seen played straight on a number of occasions) but the more common example that I've seen is that if you have a character with a +19 to a task who rolls a 3 and a character with a +2 to a task who rolls a 19 several GMs will rule that the character who rolled a 3 failed and the character who rolled a 19 succeeded.

The other one is "Oblivion scaling" where the GM will scale things which really don't make sense such as a stuck rotted non-magical door in an epic level dungeon having a higher bash DC than a well constructed locked door in a 1st level dungeon.

I've seen both of these examples come up more often than I would care to admit.

Ugh. I'd find both rather jarring. I've occasionally called for a roll, only to have a player tell me that they have X in Y, and ask if they really need to roll, to which I'll generally just respond "no".

[PhoenixPhyre]

To expound on the part in your first paragraph -- if a game using this system does end up with a character who is a superhero or demigod (and thus possessing or able to access fantastic strength), then their attempt (roll, whatever) will be able to easily overcome the DC of a normal door, maybe even to the point that they automatically succeed.

If person X should automatically succeed, they shouldn't have to roll at all. I'm not fond of critical fails for skills (or critical success, for that matter).

5e's DC setting process basically asks this same question (ie calibrate with "normal" people):

1) Is the outcome in doubt? If not, don't roll. Either auto-success or auto-failure. Err on the side of success (presume the PCs are competent).

This includes if any reasonable DC would be less than the bonuses of the character (so they'd succeed on a 1). For example, I have a rogue in one game who (because of a class feature) can't roll below a 21 (minimum roll of 10, +11 modifier) on an acrobatics check. That means that many acrobatics tasks can't fail for him, since normal DCs are in the range 10-20. So he just succeeds without rolling unless it's super special.

2) Decide how hard it should be for a beginning adventurer who's trained in the skill (+4 modifier). Easy (DC 10) = 25% chance of failure. Medium (DC 15) = 50% chance of failure. Hard (DC 20) = 75% chance of failure. Very hard (DC 25) = 100% chance of failure, although someone with very good stats (+5 modifier) can succeed by sheer chance (5%). An impossible task (that could be overcome by someone very good) has DC 30. DC 20+ should be pretty rare (especially at lower levels)--those things usually just fail. DCs don't change as you level--a certain door that's DC 10 is still DC 10 when you're level 20. But by then you're probably auto-succeeding.

You could use an untrained commoner (+0), but that just shifts things by 20% (so "medium" is actually hard, etc).

3) Roll the dice. Partial success, partial failure are allowed (usually failing by 5 or less or succeeding by 5 or more). What a success or a failure looks like depends too much on the exact situation to really encode (at least for me), although the guidance is that partial success (or success at a cost) should change the environment and make further progress require a new strategy.

4) If they failed (but not really bad), but that failure doesn't really have bad consequences other than lost time, they can succeed if they take 10x as long. Often this is used by default for non-hurried "search the whole room exhaustively" type things.