A Rather Odd Confluence of D&D and Apocalypse World, or, A Die Mechanic Late-Night Absurdium

edited May 2014 in Story Games
Recently, both D&D and Apocalypse World have been on my mind, for various reasons. (Warning: It's late at night, so please bear with me - I'm not at my most lucid. However, the math here is really entertaining, so if you're into that kind of thing, do read on.)

* I find the basic ability score check in D&D (roll a d20 below the ability score) to be a great basic mechanic, and very D&D-esque in flavour, but inelegant when modifiers come into play. Modifiers are linear, and sometimes hard to judge ("should this be a roll at -6 or -7?"). So, I thought: how about we treat a "difficult" roll as rolling two dice, then discarding the better result. Like rolling at a disadvantage in D&D Next - this is roughly equivalent to a -4 or -5 penalty, a reasonable step down (-25% chance on average), a good chunk of difference in difficulty, very tangible but not overwhelming. It feels about right, which is probably why they chose this mechanic for "advantage" and "disadvantage" in D&D Next.

Now note that rolling twice and keeping the worst result is equivalent to rolling twice and requiring both rolls to succeed... hold that thought.

* The Apocalypse World mechanic is fabulous and elegant, but has very discrete steps, which could be an issue in certain designs. It gives us three tiers of outcomes, and a nice bell curve.

* Using D&D stats in a game like Dungeon World is an interesting cross between these two games. We have things like "Strength 12, Charisma 9", but AW mechanics for moves, with a derived modifier translating in-between the two. Unfortunately, this makes it hard to, say, use standard 3d6 ability score arrays in Dungeon World (characters come out a little too wimpy, on average, and it means that there's no meaningful difference between Strength 9 and Strength 12).

This got me thinking: in what other ways can these two die mechanics cross-pollinate?

So, finally, I had a rather bizarre thought, and ran the numbers to check how it plays out. It sure does work out! There may be other uses for this little coincidental correspondence, but my first thought is that it allows effortless conversion of AW moves to D&D. (And could even be used without your D&D players catching on, if they're suspicious of such things for some odd reason.) You could play "stealth Dungeon World" (or whatever) with a group of D&D players, without having them realize that you're actually using a different mechanic, because you're still using D&D stats and your basic mechanic consists of rolling d20 ability score checks. You're using DW moves when you want to, but it still smells like normal D&D.

That's one possible application.

Here it is:

D&D and AW Dice Babies

The mechanic works as follows:

You have ability scores in the 3-18 range, with an average score around 10, and most scores falling within the 8-13 range. Unusual abilities might be rated as low as 3-6, or as high as 15-18. You know, D&D baseline, nothing new here.

To make an ability check, roll 2d20, and compare each die to the ability score (a roll of the score or lower is a success):

* If both fail, you have a miss, or a total failure. (6 or less)
* If one succeeds and the other fails, you have a partial success. (7-9)
* If both succeed, you have a full success. (10+)

The cool thing here is that the odds match up very well, within a few percentage points in most cases:

Ability Score | AW Stat Value
3 | -2
5 | -1
7-8 | 0
10 | +1
12-13 | +2
15 | +3
17-18 | +4

So a "typical" ability score is 8-13 (between 0 and +2), while extreme scores are like a -1/-2, or +3/+4. (-2 is the lowest stat AW can accomodate before a full success becomes impossible, and +4 is the highest stat AW can accommodate before failure becomes impossible. The range matches quite well here.)

The "center point" of an AW roll sits at +1 (smack dab in the middle of the three outcomes); here it sits at 10, in the middle of the ability score scale.

You can have fun rolling the dice one at a time, if you like, with some extra narration or drama in-between, or just roll both at once, AW-style.

Rolling 3d6 for stats produces results just a touch more powerful than standard DW characters, pretty close to standard AW characters. For example (actually rolling for this, hang on...):

INT 11
CON 14
WIS 12
CHA 11

Corresponds to roughly:

INT +1
CON +2 (and inching towards +3)
WIS +2
CHA +1

A very reasonable stat array for a character using AW moves, which tend to work best in the 0 to +2 range.

Here's another comparison (rolling randomly once more):

hard 9
cool 7
sharp 10
hot 10
weird 8

Corresponds roughly to:

hard +0.5 (or so)
cool 0
sharp +1
hot +1
weird 0

AW characters tend to have stat arrays which add up to +3, and this was a lower-than-average set of rolls, so this is exactly in the right ballpark.

(But remember that under these rules that last character does have different odds for cool and weird, even though they're both roughly like rolling+0 in AW: this character is as much weirder than she is cool as she is harder than she is weird.)
So, if you want to import, say, 3d6-in-order stat generation to an AW hack (World of Dungeons, perhaps?), now you can do it. Roll your six stats, then use ability score checks throughout the game, for a new form of "Primitive D&D".

Or if you're running a game D&D and you wish you could use that "Parley" or "Perilous Journeys" (or whatever) move from Dungeon World just about now, say something like:

"Ok, you're going to need to roll a Charisma check. If you succeed, you're almost there, a second roll will confirm it. If you fail, you're facing disaster, but a second roll can save your bacon."
Two successes? Give them a 10+ result. Only one? It's a partial success, like a 7-9. Both fail? Make a hard move.

It still looks familiar enough to a hardened D&D player, though: it feels like a D&D mechanic, familiar and in line aesthetically with other types of rolls.

Using the Parley move as a guide, you could handle a negotiation like this, for example:

"This negotiation will take two Charisma checks. Roll the first [...]"

"Okay, you succeeded. That means they're going to agree to your plan. Roll again to see if they want some serious concessions from you in return before they do what you ask, however."


"You failed. Ooh, that's not good: that means they're probably going to expect some concessions from you even if you win them over. Roll again to see if you can get them to agree to the plan despite their rather entrenched position."
Same results as a DW Parley, but couched in D&D vocabulary, in an old-school accent any Dungeon Master from 1974 would consider entirely familiar.


Want to make an AW hack with more fine gradations in ability? (Here you've gone from about six or seven tiers (-2 to +3 or +4) to fifteen (from 3 to 18). Have at it. What was a +1 before is roughly +2.5 now, so you can have a magic bow which gives you a +1 to Volley or a set of binoculars which give you a +2 to read a sitch or whatever.

(I think that AW doesn't need this, of course, but maybe some hypothetical hack you're working on does.)


If nothing else, we now have an easy to play World of Dungeons with standard 3d6 stats - it'll "look" just like D&D, if that's a nostalgia thing you're aiming for with those old blue character sheets. Not +1 Wisdom and 0 Strength, but Strength 9 and Wisdom 13, rolled on 3d6... but the resolution and moves still work the same way.

This is also an effortless to import tiered resolution or directly port AW/DW moves, should you need them, into any D&D game. Any AW or DW move, you now have an effortless conversion point.


  • edited May 2014
    Thinking on this a little further, you can look at other correspondences... I'm going to just brainstorm for a second.

    In WoD or DW, a head-to-head combat confrontation tends to be treated something like this:

    * On a full success, the attacker damages the defender.
    * On a partial success, they each deal damage to each other.
    * On a miss, the attacker suffers damage (or whatever calamity).

    This starts to look a lot like D&D if you treat it as two rolls, like this:

    "The Orc is in front of you! What do you do?"
    "I attack it! Rolling a d20... success!"
    "Ok, you hit him, but he's also trying to hit you! Make another check to defend yourself..."
    "Rolling a d20... oh, missed this time!"
    "Ok, so you hit the Orc, but he also got you in the meantime."

    Looks pretty much like standard D&D combat (except with the player rolling all the dice), but now we can see that the probabilities are the same in both cases. The player just effectively rolled a partial success (7-9) here.

    The advantage of the AW version is that you never get the "boring" version, where neither hits the other. The advantage of the D&D way... is that you can. You don't leave out one of the fictional possibilities here.

    This can be used to model the "hard choices" on an "acting under fire"-type move. The first roll gets you dragging the body back to safety, the second gives you a chance to avoid incidental fire. Roll them together and you get AW-style resolution; roll them separately and you get D&D-esque task resolution. Lots of options here.

    Breaking an AW move into two rolls has other applications which could be interesting to hackers, though:

    Splitting up one instance of resolution into two rolls like this allows you to mix and match abilities, should you be interested in that. Just like the D&D character above could have different odds for hitting and getting hit, you could now have a more "offensive" Gunglugger or a more "defensive" one, by splitting the hard stat into two smaller stats.

    Maybe instead of one Weird stat, which describes your affinity to the maelstrom, you have a one stat which shows how open you are to the maelstrom, and a second which describes your ability to force your will upon it.

    Let's call them Unhinged and Relentless. Opening your brain for just whatever insights might be just about being Unhinged: make two Unhinged rolls, inviting the maelstrom in, so to speak. Trying to reach out to find something specific might require an Unhinged roll first, to get vague impressions of what's out there and the psychic "terrain", and then a Relentless roll to go to the sources and find specific, sharp details of the thing in detail. Fail one roll but not the other and you might get only vague impressions, or only concrete details but completely without any context.

    In contrast, using direct-brain whisper projection could require two Relentless rolls - no need to let yourself invite the maelstrom in here, it's all about brute strength and forcing your will on the maelstrom itself and making it do something it doesn't normally do.

    Anyway, I'm not too sure where this is going or what it's for, but I'll put it out here and see if someone can do something fun with it.
  • edited May 2014
    In a less heroic variant of D&D, something more like basic D&D in an OSR style, this could be applied as follows:

    Easy rolls require a single ability score check, with the option of a second roll to turn a failure into a partial success if the first fails.

    Standard checks mean rolling a single check first:
    * On a success, you achieve the bare minimum, but you need a second roll to get the precise results you want or to avoid a complication.
    * On a failure, things are going sour pretty fast, but you can roll again to turn that failure into a borderline success.

    Challenging rolls require a successful check just to have a chance to do the thing, and a second successful check to actually pull it through.

    Characters with poor stats will tend to do well at Easy tasks, but only accomplish partial successes at Standard tasks and fail at harder tasks. (A character with a 6 Strength can succeed at an Easy task 50% of the time, and a character with a Strength of 8 will succeed 65% of the time.)

    Characters with average stats will ace Easy tasks, and usually do well at Standard tasks, but mostly fail Challenging tasks. (Someone with 11 Strength will succeed at an Easy task 80% of the time, but can only manage a Challenging task 30% of the time.)

    Characters with great stats will ace Easy tasks and hardly ever fail Standard tasks, but will find Challenging tasks to be an even gamble. (A character with a 14 Strength will succeed about 50% of the time with a Challenging task, for example.)

    Optional Variation:

    If you're in a difficult situation and you can't take any care in the action, maybe you don't get a second roll: Challenging tasks are impossible for now (you'll have to find a way to improve your circumstances first), and achieving a basic success is all you can manage when faced with a Standard task.

    Now we've brought back the basic D&D ability check as a special case of this system: it's just an Easy or Standard task being carried out under fire: perhaps time pressure, or while people are shooting at you, or some other inconvenience.

    (It's Easy if the negotiated outcome on a success is something really positive for your character, like when the GM says, "OK, roll Intelligence and on a success you know everything about this monster", and Standard when the negotiated outcome on a success is so-so for your character, like a saving throw for half damage.)

    (I kind of like the implications of a saving throw being a Standard roll being made by a character acting under fire.)

    An Example

    In yesterday's B/X D&D game (we were playing over IRC), our sturdy Fighter (played by @DWeird) wanted to pull a heavy stone door up from the opening it was covering. He'd wrapped a rope around the door and then around a nearby tree, and was putting all his weight into it.

    With him was a significantly weaker character - a halfling. I figured that the strong, heavy Fighter would have a decent shot at this, but the halfling was almost certain to fail.

    I decided that he would need one Strength check to get it open, and then a second to hold it open long enough for their plan to work (they wanted to be able to drop it on the monster they suspected was inside at just the right moment).

    Effectively, this was a Standard task rolled with Strength. (Or a Challenging task if you want to consider the crushing of the monster as the actual success here: this mechanic can be interpreted either way, and even both ways simultaneously in this case. It's quite flexible in that regard.)

    Now that I've run the numbers, I can see that the odds match what I was looking for in that situation pretty well: the Strength 15 Fighter had a 56% chance of complete success (but only a 6% chance of no results at all), while the halfling (Strength 8, I think) would have only had a 16% chance of pulling off this remarkable feat, and six times more likely not to budge the thing at all than the Fighter. Feels about right to me.
  • I don't know how anyone else feels about this, but I like it. It blurs the boundaries between what is done in Dungeon World and what can be done in OSR games.
  • edited May 2014
    Nice, nice work. Inspired me to do a 1d100 conversion table and start a discussion on hybridizing systems.
  • This is elegant. Well done!

    I've only played DW and not AW and found DW to be run like I always ran D+D (principle wise). The difference being the dice rolls and your idea takes that away. Neat.
  • I really like this. I've been wanting to get an OSR group going for a while now, and this is going in the stew I'm serving up if it happens.
  • I can safely say that you've stumbled across some awesome brilliance here. Amazing serendipity!
  • Well, aside from not seeming very OSR to me (not that it matters much, esp. here), it seems like a slightly convoluted way to get a mix of AW and Otherkind Dice (just without assigning dice). Not that it's bad, there's plenty of room in this area, it just doesn't scream D&D to me.
  • Seems like it could work. There's lots of "roll X, keep best" materials out there to pull from, from Alderac games of the 1990s to the modern Cortex+ stuff.
  • edited May 2014

    I hear you. I'm not sure yet what the application of this might be; I just noticed the mathematical coincidence and decided to write it down for posterity. I'm hoping someone else might find a clever application, in fact!

    (This whole thread was really a bit of "late-night drunken post", as far as I go, to the point where I was thinking of deleting it after I posted... so I'm just glad someone is finding it interesting for now. I'm not sure myself just how much sense I'm making here.)

    That said, I can imagine applications for stuff like your recovery move:

    "Ok, you're waking up from your drunken stupor just as the King's guards roll in. Roll Constitution twice: once to see if you can get up and keep moving before they get here ["recover quickly"], and once to see if you can sober up ["recover fully"]."

    So when the others get here, you might be still in bed and drunk, in bed but sober, up but still drunk, or up and sober.

    This isn't quite as OSR as, perhaps, "Ok, the effects will wear off in 1d6x10 minutes...", but it's perhaps more interesting and it factors in variations in character stats nicely. (A lot of OSR mechanics use different die sizes, which makes factoring in different character ability extremely difficult. I was just talking about this with Eero yesterday - for instance, if a Search roll gives you a chance of 1 in 6 to find something, and you're supposed to add your Wisdom modifier (like how Dexterity mods add to your initiative), does that mean that someone with a -1 Wisdom mod can never find anything at all? I find these little issues crop up a lot in most pre-3rd-edition D&D incarnations.)
  • If you roll them one at a time and give a chance to the player to correct errors for the second roll it gives lots of ways to use it. It could still lot like a task resolution roll just a half way roll. That could still be D+D.
  • edited May 2014
    I actually kind of like using the same bonus on different dice. It's neat how Strength 16 gives you +10% chance to hit, turns a 1-in-6 chance of opening a door into 50-50, and gives a variable bonus on 2d6. It's not always super-elegant, of course, and DMs can really make a mess of things if they screw around too much.

    As an aside, since I'm always looking for ways I can streamline play (though not always implementing them, since we're still trying to sort-of play "D&D" some of the time) myself, I noticed all the monsters in B/X have a second movement rate that's usually between 20' and 60', so I use the tens place as their initiative roll now, and have PCs roll d6+Dex mod, to see if they go before or after the monster.

    But anyway, I have the same problem here as I do with Otherkind Dice, I think. Looking at your recovery example, there's two dangers there: being caught by guards and being drunk (or perhaps a goal: escape and a danger: stay drunk), and both were made up on the spot. With Otherkind Dice, every roll you're inventing some possible dangers, but if the player rolls really well, all that imagination just gets swept off the table. With AW's act under fire, the MC doesn't have to come up with the specific consequences until after the roll misses or turns up 7-9.

    If it's a set of standardized moves, that's one way to get rid of the issue. Like your fighting example, where (I assume) you would always roll once for attack and once for defence. The DM doesn't have to decide what actual attack gets through the defence until the roll is failed, or whatever (also see next post).

    One advantage to using two rolls instead of one is right there in Otherkind Dice though: being able to switch the dice. Like if one die rolls high and the other rolls low, I can choose whether I sober up or escape the guards. Also, if those are separate stats, my bonuses could be such that actually one die only succeeds if I assign it to sobering up, because my Con bonus is higher than my Dex, but that might not be considered good by everyone.

    Another advantage to splitting one yes/sorta/no roll into two dice is that it is conceptually closer to the pass/fail rolls, which you might want to still use when you're playing D&D. There's lots of times when I just want a yes or no answer to something I'm wondering about. Like: does the PC notice a thing or not, does a PC know a factoid or not, can a PC build a boat or not, etc. and it's not about consequences or complications or how much it costs. So there would be times you'd roll one die for pass/fail. And then more dynamic situation like fighting, where you'd roll 2 or even 3 dice, each with a pass/fail, and the combination of pass/fail results would turn into a "yes, but" or "no, but" situation.
  • edited May 2014
    Or like if you have standardized moves that work along two axes, like attack/defense or evade/recover, say you have a social manipulation move, for charming and deceiving people and getting them to do things. So there's two questions: are you convincing, and are you motivating? So, you could be neither, in which case, nobody buys your shit and nobody will do what you want. Or you could be both, in which case people believe you are what you say you are (ie charming, good-natured, actually a retired army general, the mayor if you are disguised as such), and will also do what you want them to do (like make out with you or hand over their money).

    But you could also be convincing, in which case people don't see through your disguise, or they still think you're a nice guy, but can't or won't do what you want them to do, or they do it but they can't manage to do it right. Because you're not very motivating.

    Or, you could manage to get them to do what you want, but they see through you at the same time. Maybe this means you have to threaten them, and they don't like you anymore. Maybe it means they give you their wallet, but then realize you're just going to run off with it now. Time to fight!

    But again, the deal here would be to leave out the part where a lot of possible consequences are brainstormed before the roll, and then discarded. You do some brainstorming after the roll, and then pick something appropriate, so at least something gets used.

    Does that make sense?
    This may not be much of an issue for others, of course, just something I took away from playing Ghost/Echo that informed my own fiddling with AW+Otherkind Dice experiments.
  • edited May 2014
    It's Lasers & Feelings, in D&D. :)

    We've used that same multi-dice hack to add partial success to 3:16, Trollbabe, Basic D&D, and other "roll under" mechanics. I like it.

    Chris McDowall is doing fun stuff with multi-dice over/under in his new project, DIE BONEHEAD DIE.
  • edited May 2014

    All that makes TOTAL sense - I think we're on exactly the same page, because those are exactly the same concerns and questions I've been asking myself as I think about this.

    I was trying to showcase that a bit with the various examples: depending on whether you decide which factors are important a) before rolling, b) after the first roll, or c) after both rolls, you effectively get different mechanics.

    a) "So you want to catch up to the bandits while carrying all that gold? That's going to be tough... you'll have to make two successful rolls to pull it off." You roll twice, and only get one. "Well, you can catch up to the bandits, but only if you are willing to drop the gold. Which is it gonna be?"

    This is straight-up Otherkind Dice (like in Ghost/Echo).

    b) You roll once, and look at the result. Describe how things are going, then introduce a complication, and make a second roll to avoid it.

    This is... not like any system I'm familiar with, except maybe rerolls against a standing roll (like Sorcerer/IaWA). It's like my interpretation of the Parley move, above. It's got less flexibility, but also a nice back-and-forth to it.

    c) This is just like a typical AW move - it looks exactly like a), except for without the discussion up-front. You roll the two dice, then the GM makes up a complication/hard choice if necessary.

    I like the movement rate/initiative hack. I'll probably end up using that! Nice.

    (Incidentally, I love how you and I end up discussing Otherkind dice in one form or another on this forum every six months or so. I always really enjoy hearing your take on things; I've gotten a lot out of these conversations over the last few years. Thanks!)
  • @John_Harper: Awesome! How did I miss Lasers & Feelings? (Reading now!)

    Did you handle this in Basic D&D in the way I describe, or some other way?
  • b) You roll once, and look at the result. Describe how things are going, then introduce a complication, and make a second roll to avoid it.
    You get something similar in Whispering Vault though, right? You roll to attack, you don't roll well enough to defeat your enemy, and then they get to attack. But the player rolls to dodge instead of the GM rolling to attack.
    (Incidentally, I love how you and I end up discussing Otherkind dice in one form or another on this forum every six months or so. I always really enjoy hearing your take on things; I've gotten a lot out of these conversations over the last few years. Thanks!)
    I got some crazy shit to post like tomorrow or Tuesday related to those last couple posts, actually.
  • @Paul_T: Yeah, for Basic D&D, the same way. There was some other OSR game that used 2d20, too, which probably influenced this line of thinking. I can't recall it right now.
  • Whitehack uses 2d20 sometimes. In a roll two and take the most/least favorable under certain circumstances way.
  • The other thing related to this that I've been thinking about is how in B/X D&D a lot of things are a 1-in-6 roll.

    Stuff like finding secret doors, searching hexes while exploring, figuring out architectural details in a dungeon, initiative, and so on. It's very difficult to modify the odds of these rolls based on ability scores or modifiers. (For example, if your Wisdom score affects your ability to find secret doors, does that mean that someone with a negative modifier from Wisdom can NEVER find secret doors at all, but someone with just a slightly better Wisdom finds them twice as often?)

    Well, the odds for a full success in AW (10+) are precisely 1/6 for a character rolling two dice with no modifier.

    That means that a conversion for 2d6+modifiers roll, and success being 10+, is another common point between AW and B/X D&D which could to transfer rules back and forth. Once again, the range of modifiers matches the die range well. (A stat of 18 gives a modifier of +3, which is perfect.)

    It could be fruitful to use AW-style rolls (modified by the relevant ability score modifier) for a lot of these tasks. The 7-9 result could function very nicely as a toggle for easy or hard tasks:

    * A 10+ is always required for a full success on these tasks.
    * However, on an easy task, a 7-9 still gives some partial results. (A "No, but..." type of answer, perhaps meaning there is no time lost, or that you may try again, or that at least you know for sure that there wasn't anything to be found in the first place, or whatever.) A miss is just a failure, on the other hand.
    * On a hard task, a 7-9 means instead that you avoid disaster, while a 6 or less means some seriously bad consequences for you.

    This could be a useful transfer point, again, between AW and B/X D&D, allowing a group to borrow mechanics from either game. Perhaps your D&D module says that characters walking through the portal have a 1-in-6 chance of noticing the trap before it triggers. In your DW game, you can have them roll+WIS to see if they spot in on a 10+ (and make up some compromise or hard choice on a 7-9, I guess: this is basically a Defy Danger roll).

    Similarly, if you're playing D&D and the rules say that the odds of tracking the escaped prisoners through the woods are 1-in-6, have your players roll 2d6. On a 10+, they do it. On a 7-9, maybe you give them some information about the tracks, despite their ultimate failure in following them.

    This works for modifiers, too: If the written probability was 2-in-6, you could give them a +1 to the roll, if it's 3-in-6 it's a +2, and so on. The odds don't match exactly, but they are similar. 2-in-6 is equivalent to 12/36, whereas rolling a 10+ when your modifier is +1 gives you 10/36 odds. Same for 3-in-6: that's 18/36, whereas rolling 2d6+2 gives 15/36 odds. Not too far off.)

    (I suspect, but do not know for sure, that this might be quite similar to the approach taken in Stars Without Number.)
  • All of these coincidences can mean only one thing...

    Vincent Baker is a time-traveler.
  • Brilliant! This is going right into my current homebrew.
  • I was fooling around with some ability score-generating schemes for D&D NPCs, and I found that the following produces a very nice spread of "slightly lower" ability scores:

    Roll 3d6, keep the best two.

    This gives great values for a more "down-to-earth" Apocalypse World-type game; they're pretty close to 3d6, just a point or two lower on average, and with a range from 2-12, but with most results clustered between 8-11. That's a much better range for a *World game that's not supposed to give great odds for anything you try.

    If you want to match traditional D&D ability score ranges, then here's a modified version which is really fun:

    Roll 3d6, take the best two, and add one to the result.

    If the two highest dice are both 6's, add the value of the third die (instead of just adding one).

    This gives a 3-18 range, with most stats still in the 8-11 range, but occasional scores in the 13-18 range (maybe for every second or third character). I like that for, say, a crowd of country bumpkins in a D&D campaign.
  • This is interesting, but I think where it clicks is with multiple obstacles (sober up, escape guards) instead of two rolls for different levels of the same task. From my own experience in playing a zillion games and testing my own dice mechanics, two rolls that can both bite you in the ass are considerably less fun, and chances at the success you truly wanted happen less frequently.

    It's obviously just another way to play; but to me, the 2d20 for one action angle says loud and clear, "Just use DW!" Instead of new ground, it comes off as trying to retrofit D&D to have more diverse outcomes. An opinion I don't see passed around much, but in which I'm confident, is that granular success is usually nothing to scream about, while granular failure is frequently a powerful, fun storytelling tool.

    It ties into Burning Wheel's linked tests. Something that often flies under the radar is that if you set up a linked test, each test has its own outcome, BUT, the contract of a linked test is that none of the intermediate tests can prevent you from getting a chance at your ultimate intent. Let's say your intent is to sneak past guards and convince an idealistic prisoner awaiting his trial to escape instead of facing the justice system in which he puts so much faith. Stealthy-Persuasion. A stealthy failure CANNOT get you noticed! It can and should, however, make things more interesting (note: it makes failure, not success, more varied): the time it took you to sneak around the guards means that it's morning already and the guard shift has changed, so you'll need to find another way out (fight your way out, maybe get help from a prisoner who is very dangerous by freeing him). But you still have a chance at ultimate success at the linked test by persuading the prisoner, something that you're denied if you fail your first of the 2d20 system.

    Along those lines, the Otherkind-style system gets really interesting when you're rolling a larger dice pool with even more than two obstacles in your path. Especially because your entire roll can encompass an extended scene or timeframe. It can require a lot of forethought by the GM to know what kind of rolls to call for or allow, but is ultimately satisfying in my experience. For example, let's say you're rushing home from the market (obstacle 1) to prepare an exquisite meal (obstacle 2) and convince your fiancée's parents that you're worthy (obstacle 3). It's your call how to allocate successes when you roll. What's interesting is that you can, in theory, arrive home late and botch the meal, but still meet her parents' standards! That could say something interesting about them. But if you're going to marry her no matter what they say and you're an aspiring chef, you might honestly prefer to knock the meal out of the park; otherwise, rumors of your shortcomings in the kitchen might spread like wildfire from her parents' tongues.
    For the sake of completion, what's the fallout from arriving home late? They get mugged on your doorstep and convince their daughter to move back home. They may have loved the food AND you, but she won't live here, so you have to uproot and find a new restaurant that will help make you become a great chef.

    Pinning all of that (and more - that's just one way to interpret those failures) on one roll is powerful and not for every game type. I wouldn't use it for D&D or DW. But as long as you're reinventing the wheel, I say focus on making failure more interesting or feel better, not on making more gateways to achieve success.
  • Paul_T, very interesting! Wondering if you ever did any more with this?
  • Missed this the first time around!

    I think if I were to use it for 5E, I'd simply say that by default you roll 2d20 and look at both results. If you have Disad, you only roll 1d20 and even if you succeed there's a complication. If you have Adv, roll 3d20 and take the 2 best—a full failure is still possible, but very unlikely, and full success becomes dramatically more likely.
  • Nice!

    I'm not working on anything in particular, although I have a "House D&D" I'm working on, in an OSR-style, which uses the 2d6+adds model for a lot of rolls and skills.
  • Thinking about it more:

    In a way this change makes Disad less punishing, since you don't have a lowered chance of getting anything good out of it at all, but in a different way this idea makes Disad worse, since even if you roll well you don't get a full success. Maybe a natural 20 on the 1d20 counts as a full success when you have Disad?

    Also: I don't really think I'd use this system in 5E combat. Not needed, hard to implement. But it'd be great for skill rolls.
Sign In or Register to comment.