How to write no rules

[PhoneLobster]

OK so I played some guy's home brew system the other day. Or maybe it was an "indie" system he merely supported, but I think it was home brew.

It is what you would call a lite system. Very light.

He ran a good game and handled the system, as much as there was one, rather well.

But it reminded me that I wanted to make this thread a while ago when mechanics dealing with arbitrary crap like skill challenges and other informal/semi-formal rules resolution was cropping up a lot in discussion around here.

So I'm going to start with a description of this guy's system. A critique of what I see as the failings, and later I will present an explanation of how I think these sort of semi-formal rules lite arbitrary resolution mechanics should actually work.

The system
You had various attributes, gear and skills. Each represented by a d4, d6 or d8.

To do, anything, you needed to roll a 4 on at least one dice. You got to roll any dice the GM felt were associated with that, thing, you were doing. Any 4 made it an over all success, any 3 or less was a negative "complication" of some kind. A complication which may or may not have negated the success...

And you could add a limited resource of bonus d6s to your rolls. A resource which refreshed when the GM felt like it.

Your gear dice ran out if the GM felt like it.

Despite abilities being broken up strictly into attributes, gear and skills you could easily roll two or more dice from the same group on one roll. So the intuitive max 1 attribute, 1 gear, 1 skill that might have made sense, wasn't happening.

You could take some portion of the dice you were about to roll and instead remove them from the roll. They became positive complications IF you succeeded. And mostly/sometimes negated negative complications of rolling under 3 on a dice you left in the pool.

Also occasionally the GM required multiple successes rolled from the pool for the action to succeed.

Though the game was apparently meant to be heroic action pulp you needed to roll 4's to succeed and the majority of the dice were d4s.

You had a bunch of negative status conditions that the GM could declare you acquired from action failures or negative complications, or from enemy actions which as far as I could tell were purely arbitrary in nature. They refreshed basically when the GM felt like it. I scored the only refresh in the game session when my oriental man of mystery took a nap at the back of a train, and I only scored it by pulling a Confuscius once said style "Even the end of the train will reach it's destination on time" statement.

There were no further rules. No positioning system, no initiative or turn taking system, no differentiation between combat or other types of action, no clear guidelines on what a success would achieve, one minute it was knocking one guy unconscious, the next it was taking out a room full of 20 or so enemies. There were common attributes everyone had an entry in (for some reason) but you were just as likely to kill a guy with "Stealth" and "Stunts" as you were with "Fight" or "Survival".

The Good Stuff
OK so this all basically boils down to strategies you use playing games like Paranoia. With enough creative thinking any one of your skills might be useful in any situation. All you need to do is convince the GM that one of your biggest skills is usable in a relevant action. And that is the case even if your only good skill is "Hygiene".

Though in this case it was about stacking up the biggest pile of dice you could. Removing every d4 in sight to boost a potential success (and stop being potential negative results) and rolling one or two d6s or maybe a d8.

And with a group of creative proactive players and a GM that is happy to encourage creative skill uses and negotiate results it works out OK.

The Bad Stuff
Having the majority (I think it was 9 out of 16) of every character's abilities represented with d4s, and only one d8 out of the lot with a "success" result of 4+ required on each and every dice roll...

Was bad. Failures were too easy to come by. With the negotiated and optional addition of dice to the pool then the optional removal of dice from the pool those d4s and the 4+ result meant some of the players were struggling with not just understanding the (few) rules of the game but worse, making really bad tactical choices with them.

I impetuously split from the main party in game, and I succeeded massively as a result. The rest found themselves engaging in "cooperative" scenarios where if ANY of them had the sense to put aside their d4s and succeed some other joker in the group would use a full pool of 4 or so of them to contribute 4 or more failures or complications to the group for them to have to compensate for.

Which not only punished people for not being able to (instantly) comprehend and manipulate the raw mechanics but also punished everyone anywhere near them. So it ended up being a game of "Lone oriental mystic hero" and a parallel adventure of "5 clowns crammed on a plane and crashing into shit repeatedly". Fortunately the GM managed to handle it well enough, but still...

So it punished lack of game mastery. It punished lack of ability to negotiate your best skills into the relevant scenario. It punished co-operation. It punished you for even being NEAR someone who lacked game mastery, or who just plain got unlucky. Being near someone actively trying to sabotage events by rolling every d4 they could into shooting "accidentally" in your generally direction with an aircraft machine gun again and again and again was especially punishing. Though I'll certainly forgive the system of that last one.

In addition the lack of transparency sometimes came around to bite you on the ass. The whole thing were sometimes a 3 or lower wasn't just a complication but a failure that had to be negated for the overall action to succeed, and sometimes it wasn't. And then that other thing where sometimes you needed some variable amount of (rolled not removed) successes greater than one in order to get a passing grade on your action. That was troublesome.

And the thing where you were actively encouraged to not just negotiate lots of dice into your pool but then selectively weed all the d4s back out again... That was OK, sort of. But the problem was not only did people not really catch on to how to use that properly but also it made every action an extreme. Let me explain...

Mostly you could negotiate or burn bonus dice for a pool resembling d6, d6, d4, d4. You would generally need one result of a 4+ and a 3 or less on any dice wouldn't make the action fail (most of the time) it would just cause a "complication".

The sensible thing to do was to remove both your d4's to reduce negative complications and to use them as potential "boosts" of positive complications if the over all roll succeeded. Then you rolled 2d6, if you got a 4+ on either you succeeded and had Two additional upgrades to your success Even if (likely) you rolled one of the d6s as a 3 or less you could cancel the negative complication and STILL had a d4 to upgrade the success.

Basically you either failed outright. Or you succeeded MORE than your basic intended action by several steps. Bare and basic success+nothing was rare as all heck.

And finally with all those d4 abilities on my character sheet I had myself asking... Why even have those entries? Why wasn't the default "try and add something or other" dice a d4 and then my character sheet could have been 9 entries lighter with just the 6 d6 and 1 d8 entry appearing on it. And why was it that everything NOT on my character sheet COULDN'T even contribute a d4, let alone anything better than that? If I swing of the chandelier I want a "Furniture" d4 out of it, maybe a d6+.

So if you are that guy
(Who's name I can't remember) I hope you aren't too annoyed at me critiquing your system (the name of which I can't remember). It was a good game but an excellent example of a number of commonly made mistakes in rules lite resolution systems.

Everyone else
Later I'm going to post a piece on how to actually write this sort of thing and get it right.

In the meantime I suggest you consider the failings of the example system.

Consider the way the multiple rolls, the complexity and lack of transparency largely contributed only negatively to the experience of asking the GM permission to roll some vague skill against a purely arbitrary DC for entirely arbitrary results.

Then later when you've thought about it, or good god, discussed it a bit, I'll come back with my "how to do it right" post.

[zeruslord]

That sounds like a hybrid of Savage Worlds and Houses of the Blooded done very poorly. Most of the basic ideas are present in interesting or well-done games, but the implementation of the failure mechanics is pretty bad. The other problem is that enemy actions are completely arbitrary, and I don't know how to fix that without making rules more complicated than they are now. The d4 problem should have been obvious in any previous games with this system. The big issue is that it makes sabotage easy and non-obvious.

Some of the system issues may have been the result of the guy running the game trying to get the outcomes he wants and not the fault of the actual system. It sounds like this game system is sitting in a zone where it requires more analysis than people looking for a rules-light game want and isn't complicated enough for people who want to really sink their teeth into a system.

[PhoneLobster]

I think the issues were basically all system related with just a little bit of player sabotage. The designer did pretty well in the role of GM.

But other than that yeah, it straddles a strange position between "No Rules" and "Enough Rules To Have Bad Side Effects".

Which is why it was an interesting example of a rules lite system that had conflicting ideas about what it is or where it is going.

But anyway, as much as this is apparently not interesting to people, I'm off to work on the How To post and liven this up a bit...

[PhoneLobster]

How To Write No Rules

Sooner or later you are going to have some portion of your RPG that lacks coverage from formalized and abstracted rules systems like those we are familiar with from combat resolution mechanics in D&D and similar games.

And basically you want some kind of rules system to cover the "And Everything Else" portion of your game.

For some "rules lite" systems that "And Everything Else" portion might be ALL of your game system.

You need and want the massive versatility and power of letting the GM just make arbitrary rulings about whatever the hell it is that the story now demands to have some sort of resolution mechanic for.

And you don't know WHAT that thing that will need resolving might be in advance. It might be jumping a pit, building a boat, deciphering alien text or something as obscure and unpredictable as attempting to skateboard along the length of an angry sea serpent during a thunder storm.

Which is why you need to leave the mechanic for resolving these issues wide open to GM arbitration, because you are NEVER going to have full and comprehensive list of formal rules prepared in advance. No matter how much effort you put into preparing rules for every scenario "And Everything Else" will always be out there, hanging over your head.

How Big Should "And Everything Else" Be?
The first question you need to ask is how big should your "rules lite" system be?

Is it intended to resolve everything ever? Everything that doesn't fit in your formalized physical combat system? Everything that doesn't fit in your multiple formalized mini games of combat+social+whatever? Everything that doesn't fit in your formalized system and ALSO has not been accounted for by the designer providing an arbitrary formalized ruling for in advance (effectively all the gaps in an attempt at a comprehensive "general" skill system)?

Well. There is no real wrong answer to this. You can run an entire RPG with "rules lite"/"GM arbitration heavy" rules (if you do it right).

But it is important to note that some rules. Like an actual attempt to comprehensively cover every action you can possibly imagine with specific rules in advance during game design. Are largely a wasted effort when some vast swathe of those rules could very well be much better covered with a "rules lite" mechanic, which will need to exist to cover the gaps anyway.

The meaning of "Arbitrary"
So to handle "everything at once" we harness the mighty power of the GM making completely Arbitrary rulings.

Player "Can I do X?"
GM "Yes/No/On a roll of Y+ you succeed"

The important thing here is that the GM uses his feel for what he wants the outcome to be and what he thinks the players will like to just arbitrarily decide that action X, whatever the hell it is, does well, whatever the hell he thinks it should.

And that is what arbitrary means "Whatever the hell you feel like", and that is good because that broad definition happens to entirely cover the subject material of "And Everything Else" without leaving any gaps at all.

The GM and the players can use their full personal creativity to just deal with whatever comes up.

But it also has it's down side.

God is not great
"The GM is God" or so say the fanatics, many of whom are strong supporters of "rules lite" systems and arbitrary GM decision making as a resolution system.

But what they fail to recognize is that simply ceding authority and power to a GM as if they were a god does not actually give the GM the wisdom and benevolence of a god.

GMs make mistakes. Worse a GM who regards themselves as the "GM as God" type GM is often divinely petty and vengeful in nature.

Giving a section of the game over to the GM to simply arbitrate entirely as they please is a big gamble that you can lose on BIG TIME just by having a shitty GM. Or even a mildly inattentive one.

That's a major reason why most RPGs have heavily formalized segments of game play. To limit and control this problem.

But we can't actually do that here. Heavily formalized rules balanced at design time are pretty much impossible in the field of "And Everything Else".

But we need SOMETHING so we use two other tools.

Tool 1) Good Advice
Instead of wasting rules space with large tables of example DCs for tasks based on the arbitrary decisions of a guy not even at the game table you should spend some time advising GMs on how to do it right and not be a dick.

So you tell them to be open to player input, to be reactive where possible. If a player presents a suggestion that some task should be easier or harder, or relevant to their "Profession (Yo Momma)" skill the GM should at least consider that, if not openly adopt player suggestions as eagerly as possible. If the GM allows arbitrary power to go to their head and refuses to listen to player input then the players are no longer contributing ANYTHING to the game, and that is bad.

You should advise them not to be afraid to simply say yes or no rather than having to roll for every single thing ever. Some actions/events are simply unacceptable to allow success, or to allow failure. Having a PC's very life hang in the balance of a single completely arbitrary roll like jumping a pit is largely a bad thing, probably a very bad thing. Having the continuation of a story REQUIRE a successful arbitrary spot check, or worse, REQUIRE a failed one, is NOT acceptable.

But advice is only so good you need something a lot harder to keep this in check, and that's where you get...

Tool 2) Transparency
If the GM is going to just arbitrarily say "Yes", "No" or "Maybe, X+ roll to find out..." it is vital that both the GM and the players can clearly tell exactly what the GM is saying, and what the implications of that are.

So you want your "Rules Lite" systems to be as amazingly transparent as possible.

And I'm sorry to all you dice fanatics out there, but that means that a lot of the time the answer SHOULD just be "Yes" or "No" without ANY roll, just an arbitrary consideration by the GM of the situation and maybe player input or character "attributes" with a simple decisive result.

And even MORE insulting for the dice fanatics, even if the GM decides to let it go to a roll, it needs to be the simplest possible roll. If he is screwing you with near impossible odds needs to be absolutely clear to you, and even more importantly to him.

And I'm sorry but for me that means dice pools are right fucking out. Certainly ones as complex and quirky as the example system.

One dice, rolled once, vs a single DC number, with MAYBE a single net bonus to the dice roll result.

THAT is an acceptable transparent mechanic. The GM picks the DC and maybe the bonus you get on your roll and pretty much anyone should be able to see the implications of that on your likely hood of success.

If the GM is being overly generous or overly dicky, you know. And HE knows, which is also handy. You never get a situation where the GM has been screwing you forever and you can't tell, or worse where he has been screwing you forever and not only not doing it deliberately but not even realizing he was doing it.

Formal+Arbitrary=Arbitrary
Notice in the transparency bit I said DC vs Dice Roll +MAYBE a bonus. Well. It's because of this.

If the DC is just some arbitrary number the GM pulled out of his ass to account for his perception of the situation, the "skills" involved, the "circumstances" involved and what he would like to be the probable results...

Then the bonus is largely rendered irrelevant. Because the DC could just be an entirely different value for no provable or observable reason.

Giving a bonus, even if it is ALSO arbitrary has a potential place as a way of telling a player "In this arbitrary circumstance you are arbitrarily this much taller than someone else who doesn't get the same, arbitrary, special consideration that you do".

But if EITHER the bonus, or the DC, EVER have a number which is "whatever the GM thinks it should be", then the entire exercise is purely a matter of GM whim. Pretending otherwise by giving out large blocks of supposedly static DCs or bonuses to slip into an arbitrary equation and pretending it is a complex formal system is only going to damage transparency and thus break things all over the place.

Degrees Of Success
Hey, everyone LOVES degrees of success hey?

The example system certainly was all about degrees of success with its negative and positive complications.

And you can find lots of "Rules Lite" mechanics out there where your skill checks or whatever have varying degrees of success.

As far as I am concerned degrees of success can very much have a place in your "rules lite" system.

But they MUST meet the strict transparency requirements of rules lite systems. It must be very obvious and clear to players how wide or likely a range of degrees of success they might achieve.

In addition degrees of success should never allow an action to achieve MORE (or less) than the limits to success or failure that the GM has arbitrarily determined. Remember back in "Good Advice" where I mentioned that some results to actions are simply unacceptable? Well those results are STILL unacceptable as part of a "degrees of success" system.

If a PC is jumping a chasm and they fail by whatever margin triggers really bad degrees of failure, that doesn't mean that instant PC death is an appropriate result for your game (it might be, but if it isn't then just saying "But degrees of success?" doesn't cut it as an excuse). This should be reflected by the game mechanics or the advice to the GM in some way.

The example system above handled degrees of success and failure rather badly. It wasn't transparent, and results were not in line with genre conventions the game was supposed to adhere to.

Influencing results with resources
A lot of "rules lite" mechanics like to give players some sort of limited resource they can use to boost success chances or degrees of success or whatever.

The example one certainly did.

By doing this a system based largely off GM arbitration is giving a tool to the other players that says "You can spend this resource to change the GMs arbitrary decision by X amount".

Again the whole "Arbitrary+Formal=Arbitrary" moderately devalues this, but whatever.

Depending on your attitude this is either good or bad in nature. I'm inclined to not particularly mind it. As long as the GM has the balls to have said an outright "Yes" or "No" to those actions/results that require it the boosts will only be a productive element that lets players have a positive input on the story.

Of course then you have to be concerned about resource refresh rates and such. The example resource refresh rates were basically entirely arbitrary or existed as some reward for "Entertaining role play". Which I'm actually largely OK with, it's a resource of arbitrary value used to influence arbitrary rolls, so why not have it refresh at an arbitrary rate?

Taking Turns
If your "rules lite" system is covering ENOUGH ground sooner or later someone somewhere is going to start asking questions about competitve actions.

If character 1 wants to shoot character 2 and character 2 wants to also shoot character 1 at the same time...

Well. With sufficient generosity to the attitudes of parties involved in the game simply leaving resolution down to who calls it first MIGHT be acceptable.

But really the second the system covers competitive ground it would be helpful if there were SOME form of turn structure or interaction between competitive actions that is clear and transparent.

A simple "the party who succeeds by the largest margin shoots first and negates the other's action, A tie means both shots go off" is a good way to go. But there are potentially other options.

The example system had no mechanics to cover this at all as far as I could tell. It seemed to be down to the GM simply having some vague prerogative to TRY and share screen time among players.

NPCs taking a turn
A rules lite system should also answer the question "If a tree falls in the forest and it isn't a PC does it make a save vs lumberjack roll?"

As far as I could tell the example system didn't make a clear distinction.

Certainly the NPCs weren't rolling... but sometimes they were taking a turn (in the non-existent turn sequence) which just automatically did something (being thus somewhat superior to half the players at the table who couldn't figure when to roll d4s or not), and sometimes they only "did something" in the form of causing negative consequences to a players failed roll or "negative complications" result.

The system needs to tell us. Do NPCs function just like PCs, including the same potential roll mechanics? Do NPCs just get to always ignore rolls even if the GM would have required a roll from a PC? Or are NPC actions purely a special effect that is influenced by PCs who are the only active parties in all roll resolved actions?

Co-operation and contagious failure
This was a serious element of the example system. More people acting together had larger failure rates. And PCs failures often led to negative consequences for nearby PCs.

The thing where one PC failing leads to "complications" where another PC is shot in the back is largely an issue of advice, guidelines and genre conventions. There is not much you can do other than try to make those things clear when designing your game so that GMs know whether or not they are supposed to be pulling that sort of "critical fumble" shenanigans. I am unsure if the example system really had a clear idea of what it was doing in that regard.

But the higher failure rate/negative complication rate of co-operative actions was very much a direct mechanical implication of the example system.

And it can easily be a mechanical implication of many other potential systems.

If jumping a chasm requires a 5+ roll on a d20 for one character that is clear enough. But if I'm tied to another guy in a three legged race and jumping the chasm requires BOTH players to roll a 5+ on a d20 the chances of success have dropped. But in addition they have dropped in a way that is LESS readily clear than if it had been changed to a single d20 roll of an 8+.

Ideally co-operative actions should probably become single roll events that have modified DCs or bonuses. The players can more immediately tell if tying another PC to my PC is a wise move for jumping the chasm or not.

They should almost certainly NOT be resolved by adding size, complexity and unpredictability to an already complex and unpredictable dice pool.

And as a side note, 4e style "skill challenges", to the limited degree that that term means anything, are BAD. They force more rolls than transparency can permit, and forcing all PCs to roll is also bad for co-operative actions obvious ways described in detail here abouts already.

So actions and their consequences should be as separate as possible. And when combined you need to have a mechanic that does that in a neat, simple, transparent way. Not a big damn cluster fuck.

Status Effects
The example system had status effects. But it was utterly unclear what the hell they did, or even what some of them were.

Generally having status effects just be another arbitrary factor in an arbitrary system is probably entirely fine. But it would be nice if that were clear.

And at that point you probably don't need a common list of 10 different named status effects that each and every character tracks. If I might just arbitrarily acquire the status effect "Angry" (as in the example system) and it might just arbitrarily do... something, I really don't NEED it to be an entry on that list of 10 status effects and to know that the other 4 players haven't got that box ticked.

What does an arbitrary attribute look like?
Which brings up another thing. The example system character sheets looked like this.

Attributes:
(8 attributes you shared the names of with all other PCs, different dice entries in each, 5 were d4s, 2 were d6s one was a d8 )

Items:
(4 attributes probably unique to yourself, with 1 d6 and 3 d4s)

Skills:
(4 attributes unique to yourself, with 1 d6 and 3 d4s)

Negative Status Effects:
(About 8 or so bad things, all shared with other character sheets, each being a tick box that did... something).

A box for tracking your limited bonus dice.

Names, titles, background text, a picture, some vague pieces of pseudo rules text.


Now I regard large parts of this sheet to be irrelevant.

Status effects should have been an empty box that you just write words and other notes in. Having that odd list of 8 or whatever named effects that did "?" and were obtained by "?" both meant I was tracking things that I had no idea why I cared about (like being "Angry" or "Confused"). And also meant that I wasn't tracking other status effects I might suddenly arbitrarily care about, like "Manic" or "On Fire".

But what about the rest?

Why were my various arbitrary attributes in 3 different categories? Those categories were entirely meaningless. They could have just been one block.

Why was I tracking the abilities that I was merely average at? In an arbitrary system like this I only really give a damn about stuff I am better at or worse at. More than half the entries on the sheet could be replaced by "Default of d4" as a simple rule.

Why did the abilities on the sheet share common names with those on other sheets (some of the time)? The game is entirely arbitrary, I have to negotiate the situational relevance of ANY ability with the GM anyway, so why does it matter that me and that other guy BOTH have an ability called "Brains" but only I have an ability called "Kung Fu"? Wouldn't it have been better if the guy with average brains had no entry and the guy with big brains and an oriental background had "Inscrutable Intellect" as a uniquely named exciting (and equally vague) ability?

Ideally the abilities should have been specific areas in which characters were better or worse than some vague average. There was no reason for them to be anything other than unique abilities, and no particular reason to categorize them in the strange manner listed, especially since inventory was pretty much functionally static.

But in so much as you should have SOME ability entries what should they look like?

I'll give you two options for now.

1) "Good At", "Bad At".
You have an entry on your sheet that makes an unspecified claim about your abilities in some vague area. This is a tool you use to negotiate better DCs or other special consideration from the GMs arbitrary rulings.

This is also handy because your character background notes essentially become a functional part of the game mechanics in a similar manner. And for this sort of "Rules Lite" scenario that is a good thing.

2) "Something + X", "Something Else - X"
You get bonuses or penalties to your something-or-others. This means if the GM makes you roll something that is determined to be relevant to "something or other" you get the +X bonus against the arbitrary DC.

If no roll is called for you can still use your bonus as a lever to negotiate for special consideration from the GM much as you could in option 1 or with your character background or "circumstantial" factors.

The bonus of course isn't incredibly meaningful due to the factors mentioned in the "arbitrary+formal=arbitrary" section. But it's a way of representing your "Rules Lite" abilities on a character sheet in a productive manner.

So what should this guy's game have done?
Later I might give it some example reformed rules that match my requirements for a good "no rules" rules system. In another post.

[Absentminded_Wizard]

...but the implementation of the failure mechanics is pretty bad.

I agree. Your odds of getting at least one 4 on 3d4 is about 58%. Not bad as a floor, but if you almost never get more than a d4, or if the DM requires extra successes.... And having a 75% chance of a negative complication on most of your dice? Ouch!

And PL, I will read your long, detailed treatise later when I have more time/am less tired.

[violence in the media]

I don't really have anything to add to this, but I wanted you to know that this is an entertaining read. Please continue.

[Lich-Loved]

One example of a "lite system" is the "Just Fudge It" rules in FUDGE. There was a great essay on how this approach worked, but I couldn't find it in a brief online search. Anyway, it went something like this:

Suppose the player wants to jump out of 3rd story window to avoid some baddies breaking down the door (or maybe you are Rambo jumping off a cliff to avoid a sniper). How should the "degrees of success" system be used to resolve this task when there are no falling rules and no rules to handle jumping into the nearby tree which *might* be reachable or at least break the fall?

"Just Fudge It" recommends that the GM assign the boundary conditions of what *could* happen (or what is acceptable to the story to have happen) to the dice and let the player roll. Using dice that generate a bell curve (no flat d20's), assign the Worst Possible Result (whatever that is - from death to a broken bone to a Serious Wound or whatever) to the low end of the curve and the Best Possible Result (complete success, scratched up, getting away with a slight limp, whatever) to the high end of the curve. The GM should also assign a Likely Result and assign this to the arithmetic mean of the bell curve. This determination is made on the fly and in front of the player and the dice are rolled to determine to outcome. In the example above and using the familiar 3d6 as the generator (or whatever your Dice of Choice), the GM rules that the character will be unconscious with a broken leg (and thus likely captured) on a roll of 6 or lower, will suffer a Serious wound on a 10 - 12 and will make the jump into the tree on a 16+ and suffer only moderate bruising. The dice are rolled and the GM interprets the outcome, interpolating between points as necessary to get the degree of success (so if the player rolled a 14, the GM might rule that the character made the tree but failed to hold on initially and fell crashing through the branches and is now Moderately wounded). This approach can be used whenever the character attempts something that is not an Opposed Roll and with a little practice from the GM can be handled very quickly in game and cover just about every scenario that could conceivably occur, all without having specific rules for each situation.

There is nothing mystical about this approach, but it does bound the result and informs the player ahead of time about the possible outcomes and degrees of success available. I like this approach better than let's say setting a Difficulty in Shadowrun of 4 (very difficult) or a DC of 30 in D&D and then having the player roll all applicable dice since it gives the GM a very clear picture of a range of outcomes (not just pass/fail). Trying this in Shadowrun would require that the GM know the binomial distribution associated with the dice pool the character was going to throw, something that most people are not going to have a their fingertips. Doing this in d20 really isn't possible with a fixed target number (rather than a series of DC's) since the d20 is a flat distribution.

[PhoneLobster]

I immediately dislike the just fudge it technique because it dares to disagree with me on a single point.

Transparency and how many dice to roll.

I dislike the effects of "bell curve" rolls in ANY RPG mechanic. I think they are unpleasant, opaque and largely don't do what most of their proponents think they do. I regard the whole bell curve thing more as a sort of wanky zombie fad rather than a productive or sensible design decision.

But I REALLY don't like it for this particular sort of scenario for the reasons I outlined under the transparency section.

"No rules" rules need to be extra super transparent. 2d6 is pushing that transparency limit well beyond lowest common denominator audiences as it is. Anything more "curvey" like 3d6+ is making life uncomfortable for ME, and in a former life I had a university education in advanced mathematics.

I like me the one roll one dice thing, it's simple, it's straight forward, even the slower gamers get it in a snap, and I barely need to think about it to know the implications of setting a target number.

And I'm not giving it up for some fruity la-de-da "bell curve" that isn't really very curvy and just makes the extremes of target "to roll" numbers less useful. I mean on 3d6 a god damn 18 virtually might as well be the GM saying "impossible, I won't let you do it".

[Starmaker]

I dislike the effects of "bell curve" rolls in ANY RPG mechanic. I think they are unpleasant, opaque and largely don't do what most of their proponents think they do.

I'm not going to argue about your personal preferences, but our group has long since switched from d20 to Bernoulli (Shadowrun dice pools) for rules-lite games. Opacity is not necessarily bad. As in computer games where formulae are commonly hidden, players start thinking not in terms of exact probabilities but in terms of dice pool vs. difficulty. On the other hand, when I'm optimizing the shit out of my character in WoW or D&D, I do want to know exactly how much DPS each +1 buys me.

[MartinHarper]

Really, I don't think that 3d6 is fundamentally harder to grasp than 1d20. The 1d20 has the advantage that it is one dice. The 3d6 has the advantage that six-sided dice are everywhere. I think it's better to have a unified mechanic, so players can get used to how likely they are to roll a 16+ on whatever dice system they are using.

[Username17]

I think that bell curves do exactly what their proponents think they do. Here's an example of some dude explaining why he thinks GURPS uses 3d6:

I don't have any specific numbers to show you, but I can give you some basic theory. When you add up multiple dice, your totals tend to all land in a fairly tight grouping. 'Most' rolls on 3d6 tend to be 8-12. Therefore a character can be reasonably confident of succeeding any skill roll of 13 or higher. With a single d20, all totals are equally possible, as with only one die any number can come up. This means you get more high and low swings; i.e. you have a 5% chance of rolling any number on the die, including 1 or 20, the extremes. In order to get the extremes on 3d6 all three dice have to hit '1' or '6' at the same time, which is much less likely. I seem to recall the odds of getting a total of '3' are something on the order of 1% or less? So basically, with 3d6 you tend to perform at a moderate level most frequently, whereas with a single d20 your rolls are all over the place.

The guy doesn't have the exact odds or anything, but he is exactly right as far as what the 3d6 bell curve does.

-Username17

[Lago PARANOIA]

PhoneLobster, I do hope that you realize that the reason why D&D uses the d20 instead of 3d6 is because that game is meant to be swingy. Critical hits are supposed to happen fairly often (every edition seems to think that at least someone in the party should get at least one critical hit per combat) and lower-level critters are supposed to have a fighting chance against higher-level ones.

It's not because the d20 roll is superior or because they want people to grok the mechanics more easily. For example, for Mutants and Masterminds, you can almost tell they the game designers wanted to use 3d6 but they couldn't.

[PhoneLobster]

I've seen the "bell curve" people claim that their bell curves do everything from solve the generalist vs specialist problem, solve off the RNG problems, to curing cancer. Mostly they just say it is "more realistic" or "feels better", proving their mastery of game design right there....

As far as I see it 3d6 falling "mostly" on 8-12 means your RNG is now basically limited to 8-12 and all your bonuses and game mechanics are restricted to that 8-12 range. Making any mechanic assuming the RNG will fall in the extremes is an incredibly shoddy idea and you get shit where most of the time you may as well not be rolling and then suddenly...

BAM way off the standard RNG. Infrequently enough to be risky but not frequently enough to be remotely reliable.

It's a lot like the stupidity that is infinite exploding dice.

And Lago, what you are describing isn't what I would call "Swingy" it is what I would call predictable 3d6 is swingy because it's almost always "the middle" then has rare and unpredictable wild mood swings to low or high.

But anyway. I've never seen a bell curve in practice that I liked. They were always systems that suffered more than ever from specialization because all you had to do was make it just over that hump and BAM win city for 85%+ rolls, and investing in any skill or ability that DOESN'T make the hump is basically a complete waste. And the designers never seem to grasp that shit, giving out bonuses and setting target numbers more as if the game were running off a nice "flat" dice like a d20.

But anyway you want to argue "curvy" dice rolls I suggest another thread because it will only swamp the whole "no rules" topic if it stays here.

Suffice it to say now. 3d6 is too "curvy" and opaque for these sorts of mechanics. A 1/72 chance of 17, or a 1/216 chance of an 18 is not something any GM should be expected to calculate on the fly while pulling numbers out his ass, it is not something players should be expected to follow on the fly, and 1/216 is not really an acceptable "chance of success" to hand to someone on an arbitrary action anyway. And I genuinely expect no disagreement on those specific claims. 3d6 is just NOT conducive to pulling target numbers out of your ass for.

[Crissa]

So, PL doesn't want the edge conditions to happen less often.

Right.

-Crissa

PS, it's totally easy to find the probabilities in this new fangled internet thing.

[Kaelik]

So, PL doesn't want the edge conditions to happen less often.

Right.

-Crissa

PS, it's totally easy to find the probabilities in this new fangled internet thing.

1) You could also not be near the internet, and it still makes making up numbers take longer, defeating one of the main advantage of making up numbers.

2) Don't be an idiot. Even a five year old can tell that's not what PL is saying. He's clearly explaining that bell curves are so curvy that what you actually have is a RNG of 5 numbers that claims to be larger. And so when you design the game as if your RNG is 15 numbers, you get a bunch of people who should be able to succeed, but can't, and a bunch of people who should risk failure but don't.

[Lago PARANOIA]

And Lago, what you are describing isn't what I would call "Swingy" it is what I would call predictable 3d6 is swingy because it's almost always "the middle" then has rare and unpredictable wild mood swings to low or high.

On what planet could a random result that produced an extreme situation happening more often could be considered less "swingy"?

Or to put it another way, in d20, your chances of getting your shit wrecked by an orc with a greataxe as a fourth level character is higher than in 3d6, because d20 orcs will roll a critical hit more often--but unless they roll a critical hit, you'll chew through them because you're fourth level characters and they're pathetic mooks that barely even have double-digit hit points.

So unless there's a critical hit, the outcome of a level 4 party versus five level 1 orcs is a foregone conclusion (the party walks away barely even hurt). A bell curve makes the 'PC goes down against otherwise easy enemies' result less likely to happen. So how is the d20 system not 'more swingy'?

[Kaelik]

And Lago, what you are describing isn't what I would call "Swingy" it is what I would call predictable 3d6 is swingy because it's almost always "the middle" then has rare and unpredictable wild mood swings to low or high.

On what planet could a random result that produced an extreme situation happening less often could be considered less "swingy"?

Or to put it another way, in d20, your chances of getting your shit wrecked by an orc with a greataxe as a fourth level character is higher than in 3d6, because d20 orcs will roll a critical hit more often--but unless they roll a critical hit, you'll chew through them because you're fourth level characters and they're pathetic mooks that barely even have double-digit hit points.

So unless there's a critical hit, the outcome of a level 4 party versus five level 1 orcs is a foregone conclusion (the party walks away barely even hurt). A bell curve makes the 'victory with no losses' less likely to happen. So how is the d20 system not 'more swingy'?

It's more swingy to him because the RNG is 5 numbers and you are more likely to be off the RNG in one direction or another.

Yes, this is totally the opposite of what we mean when we say swingy.

PL is saying different situations have drastically different results, which is totally perpendicular to swingy, IE a single situation could go either way.

You have to read PL in context, because he never ever uses words to mean what they actually mean.

[PhoneLobster]

You have to read PL in context, because he never ever uses words to mean what they actually mean.

Oh I guess I'll just have to double check my definition of swingy with google there... and... oh it's a made up word

All you get is (edit you can get more than that with other searches, but none of it matches Lago's idea of swingy)

Definitions of swingy on the Web:

* lilting: characterized by a buoyant rhythm; "an easy lilting stride"; "the flute broke into a light lilting air"; "a swinging pace"; "a graceful ...
wordnetweb.princeton.edu/perl/webwn

* (of ice) Allowing stones to curl more than usual

Predictability and human gambling is an interesting and strange world.

By making the middle ground a much more reliably predictable result but still having extremes a negligible possibility the sensible gamble becomes an increasing reliance on the middle ground. Accounting for a less than half of one percent chance of an 18 on 3d6 is just not sensible, and yet... the result it possible.

It's like the definition of a "Black Swan" event. Just because you almost never see an 18 on 3d6 doesn't mean it can't happen, and when it does it will totally break the expectations of the players and the RNG.

Now the orc example is an interesting one. Because if you just swap 3d6 for a d20 in 3.x D&D game (which is often openly advised by bell curve enthusiasts) you create situations where an advantage of a mere 4 points of RNG in the right part of the range functionally puts you right off the actual practical RNG.

Sure if you define critical hit as an 18 on 3d6 you make it a lot less common and that big splat event might not happen so much. But that isn't a definition of "swingyness" that's a definition reliant on critical hit mechanics.

Meanwhile the 3d6 orc encounter suffers from significant tipping point issues.

Lets imagine we aren't using D&D as a skeleton for this example. Lets imagine we are using in fact a "rules lite" system. The ENTIRE ENCOUNTER with the orcs will now be resolved with a single roll

In this case either a d20 roll, or a 3d6 roll.

Now the orcs are represented by a target number on that single roll.

In the d20 case changing that target number by one point has a linear easily predictable outcome. A change by 4 points is always a nice even 20% change to chance of success. Sticking in an extra orc, a circumstantial bonus, a player skill it's no big deal.

In the 3d6 example an extra point of target ranges from a 0.4% to a 12% change in odds. A change by a mere 4 (so you've got say up to four single points of difficulty factors being accounted for or not) can see anywhere from a 10% to a 50% change in odds.

That makes the effect of some small bonus unpredictable. That makes the functional RNG shorter, and that is very much a potential tipping point scenario which frankly by the nature of tipping points sounds pretty "swingy" or "sway-ey" or "wobbleriffic" to me.

But really must we argue about broken bell curves on a thread that really is only very tangentially about them?

[Ice9]

When we're talking about a single roll with results assigned by the GM, then all the the type of dice changes is the granularity at different parts of the scale - 3d6 gives you more at the ends and less in the middle. For instance, let's covert that 3d6 example to d20:

Code: Select all

3d6      Chance		 d20
6-        9.3%   ->	1-2
10-12    36.5%	->	8-14
16+       4.6%	->	20

And actually, in a case like this, I think a flat roll is more transparent. It becomes immediately obvious that the DM is saying "you've only got a 1 in 20 chance for the best result", whereas "16+" might seem more likely.

[Username17]

Interestingly, 3d6 does have a deeper RNG than does a d20 roll. Seriously.

Let's say that you succeed on a 10+ in ether system to begin with. The bonus that pushes you to having a 90% chance of success is +3 on the 3d6, and +7 on a d20. There are more numbers on the d20's bonus to get to that point, but that doesn't mean the RNG is deeper - it means that each bonus is smaller. What makes the RNG deeper on 3d6 is that you can add that bonus again on the 3d6 roll and you'll still be on the RNG (although only barely), while on the d20 you'll have gone off the RNG entirely without even getting a third of the way through a repeat of the bonus.

Yeah, on 3d6 a bonus is really fucking big. People who use 1d6 systems have to deal with that fact as well. But large bonuses and penalties can be added or not added dynamically within a game without pushing the RNG completely out the window when you use a bell curve. That's pretty sweet if you want to have bonuses and penalties appear contingently in the middle of games.

-Username17

[Starmaker]

Probabilities! and "How long is the RNG?" discussion.

[PhoneLobster]

Thats a load of circular logic bullshit Frank.

"The granularity is finer as long as you only count the part of the RNG that only even happens 10% of the time!"

WTF?

On 3d6 you have a total range of whole numbers of bonus that is smaller. You ALSO have a portion of that range of whole number bonuses that we don't care about. +0.4%? That last point of bonus may as well not exist. +1.3% the second last point is laughable. +2.7% Hey you might almost maybe one day notice the third last... and so on!

You quite simply have a very small variation in bonuses that the RNG gives a shit about. It is a low granularity system. The fact that part of the RNG has finer granularity is totally occluded by the fact that the part with the super chunky granularity is on the same damn scale and will completely eclipse the smaller valued increments all added together the vast majority of the time.

That is only "deeper" by defining the term "deeper" as "complete bullshit".

[Username17]

Thats a load of circular logic bullshit Frank.

"The granularity is finer as long as you only count the part of the RNG that only even happens 10% of the time!"

No. The granularity is simply a matter of how big a die you roll. If you want more granularity, roll percentile dice.

The depth is how much bonus you can get before you hit an RNG wall after the bonus you have is already large compared to the RNG as a whole.

A d20, or any other uncurved die roll has very little depth, because the difference between the end of the die and a large bonus are by definition a very small bonus. The depth on a curved roll can be quite large because a very large bonus can be added more than once before getting to the end of the RNG.

I understand that you don't care about that. But while it's certainly not for every game, the fact is that you're being a jackass by insisting that it's always bad. It's a great and justifiably well liked and used mechanic. A lot of people get upset when the game tells them that an action cant get any more difficult once they have their eyes closed without becoming impossible. They want to be able to have their eyes closed and have a wound penalty and still have a chance. They are not wrong to do so.

On a 3d6 roll, being blinded should be maybe a -3 or -4. Being blinded on a d20 should be more like a -8. On a 3d6 roll, you can have two such penalties (or equivalent bonuses) and still be on the RNG (albeit only barely). On a d20 you can't. And if you can't see why this would be valuable to people when they are making a game, you have a problem. And the people who like 3d6 don't. They are making sense and you're not.

-Username17

[Lago PARANOIA]

That said, the d20 does have advantages over the 3d6 aside from the fact that you roll less dice and it's easier to calculate iterative probability.

Going back to the 'swingy' issue (or vacillate if you don't like that word), while D&D does expect your 4th level party to curbstomp orcs without thinking about it too hard, the d20 roll does allow said orcs to become a threat again just by attacking at dusk and standing on some higher ground. As Frank said, while it's harder to push people off of the RNG on a bell curve, on a d20 it's easier to put people back on it if they're already off. This is valuable for games like D&D where they do expect your mid-level adventurers to panic if the king sends his Red Guard after you.

I can see why they wanted to use a d20 for the soak roll in Mutants and Masterminds d20, since they wanted combat to be chaotic--but since combat is also a low-risk affair in that system it has the benefits of getting combat over more quickly and also allowing a modestly outmatched side to make a comeback... without making it so that players get jilted after their enemies have a hot streak in a routine bank robbery.

[PhoneLobster]

On a 3d6 roll, being blinded should be maybe a -3 or -4.

What does that even MEAN Frank? That bonus has an inconsistent effect Dependant on other bonuses and the target number.

You say "it should mean -X or -Y on a d20" because you feel it should have a certain effect on the possible outcome, but on a 3d6 we don't even KNOW what that effect on outcome is without a full context of the net bonus situation before the additional "-3 or -4". So your arbitrary ruling on how strong an effect blindness is... isn't even consistent anymore on a 3d6.

On a 3d6 roll, you can have two such penalties (or equivalent bonuses) and still be on the RNG (albeit only barely). On a d20 you can't.

Wrong. You can have 2 modifiers of 4 to the target number in each system. Your comparison where you just make the 3d6 number smaller and the d20 one bigger is a deceptive farce.

You justify it with the large granularity of a particular part of the 3d6 range and translate that to a bonus of the same over all size in the d20 granularity. But you just fucking arbitrarily cherry picked the granularity at which to read in the value of your +4 modifier on a 3d6!

Using your SAME spurious reasoning I can declare the d20 range to be "deeper" because +4 at another portion of the 3d6 is in fact only worth a 10% change in odds. Which is only a +2 bonus in d20, so you can add that modifier MORE in d20 than you can in 3d6! Tadah! "Deeper"

Your particular defense is completely based on really bad inconsistent reasoning. If you want to defend these stupid dice roll methods you are doing it wrong

And the people who like 3d6 don't. They are making sense and you're not.

So how much is that +4 on 3d6 worth then Frank? +8 on a d20 or +2? It's your line of reasoning which doesn't make sense and is utterly internally inconsistent.