D&D 5th Edition XI: The 15-Minute Designer Workday

[Saph]

Welcome to the 11th thread discussing D&D 5th Edition, aka D&D Next!

As is (by now) well known to every RPGer who hasn’t spent the past year hiding under a rock, a new edition of D&D is coming out. When? Well, they’re not telling us. What they are giving us is an open playtest, which you can sign up for right here. At the time of writing, the most recent playtest packet dates from June 7th, 2013.

Use this thread to discuss the playtest, the weekly mostly-weekly Legends and Lore update articles from Mike Mearls, and other news relating to D&D’s new edition.

Useful (and freshly updated!) links:

• Playtest Signup
• Legends and Lore Archive
• EN World D&D Forum

Penny Arcade / PvP 5e Podcasts:

• Part 1 of 4
• Part 2 of 4
• Part 3 of 4
• Part 4 of 4

Previous threads:

• First Edition
• Second Edition
• Third Edition
• 3.5th Edition
• Fourth Edition
• D&D 5th Edition: the fifth edition of the discussion thread
• D&D 5th Edition: 6th Thread and counting
• D&D 5th Edition: Thread #7
• D&D 5th Edition: 8th Revision and counting
• Pathfinder, Next, and the Future of D&D
• D&D 5th Edition IX: Still in the Idea Stage
• D&D 5th Edition X: Where's the Craft (RPG System) skill?

[DeltaEmil]

I like Flickerdart's task resolution system. I wonder if it could be improved even further, or what kind of pitfalls this system would encounter.

If you wanted to be really simple about it (with potential issues I'm sure) you could make skill checks cumulative; a door has a DC of, say, 30, and you just need to keep doing skill checks until you get enough to open it. Kind of like HP for tasks.

You could also impose an "damage reduction" of sorts - the door has a progress DC of 10, and a success target of 10, so subtract 10 from everything you get and add it to the total, and when it exceeds 10, poof.

So for instance, rolling with the lockpick example, Steve the Street Rat has a +10 to Open Lock, which he's applying with gusto to a merchant's money chest. The chest has a well-crafted lock - progress DC 18 and success target 10. His friend, Trevor the Thug, has a total modifier of -3 to Open Lock (8 Dexterity and no tools), so he can't possibly do anything to the lock. It won't take Steve more than a few rounds to open it - 5 rounds if he takes 10 - but if he decides to chance it, he has a 15% chance of cracking the lock open in just one roll (18+).

This also sort of solves the issue of unreasonable rolls somewhat - unless you're skilled enough to beat the DC in the first place, you don't roll because you can't possibly make it. If you CAN beat the DC, you get to try. This neutralizes the importance of the d20 somewhat, because if you have a 25% chance to hit the DC you'll be contributing a small number to the success target even if you make it, instead of just hitting it and winning forever.

And if you wanted to have traditional one-check DCs you just set the target to 0.

[Raineh Daze]

Well, there's the obvious one of 'get people to accept something that isn't just one 1d20 roll'.

It actually gives a non-binary system to skills, which is excellent for stuff like diplomacy and climbing.

[SiuiS]

Well, there's the obvious one of 'get people to accept something that isn't just one 1d20 roll'.

It actually gives a non-binary system to skills, which is excellent for stuff like diplomacy and climbing.

There's precedent. These rolls hosed up (but were ignored and disliked) throughout 3.5, and there was an entire system behind them in 4e. So really, it's not novel at all. It's just presented in a way that makes it seem more broadly applicable than the past iterations.

One new thread, made to order.

This is no longer the thread you're looking for. You can go about your business. Move along.

Thank you~ ^^

[noparlpf]

It looks like higher DC=harder task, and higher success target=longer task. So climbing a short sheer cliff would be a high DC but a low success target, but climbing up a hundred foot tall tree would have a much lower DC but an extremely high success target due to sheer size.

Obviously it depends a bit on the type of cliff and type of tree, but typically it's harder to climb a tree than it is to climb a cliff, as well as likely taking longer due to the distance (which you could compare by using two similar cliff faces where one is just twice as high). Cliff faces tend to have more natural handholds, whereas tall trees don't tend to have low branches or be narrow enough to climb up easily.

[Raineh Daze]

Obviously it depends a bit on the type of cliff and type of tree, but typically it's harder to climb a tree than it is to climb a cliff, as well as likely taking longer due to the distance (which you could compare by using two similar cliff faces where one is just twice as high). Cliff faces tend to have more natural handholds, whereas tall trees don't tend to have low branches or be narrow enough to climb up easily.

When I said 'sheer cliff', I was thinking of one that was rather... lacking in handholds. An abnormal cliff.

And for tree I was thinking some sort of really tall fir or something.

Either way, my choice of trees and cliffs doesn't really have much bearing on the system.

[noparlpf]

When I said 'sheer cliff', I was thinking of one that was rather... lacking in handholds. An abnormal cliff.

And for tree I was thinking some sort of really tall fir or something.

Either way, my choice of trees and cliffs doesn't really have much bearing on the system.

I used to climb a lot, and a cliff with no handholds is going to take wayyy longer to climb even twenty feet than it would to climb a hundred-foot fir.

[Raineh Daze]

I used to climb a lot, and a cliff with no handholds is going to take wayyy longer to climb even twenty feet than it would to climb a hundred-foot fir.

This is fiction, leave what happens in reality at the door and stop complaining about the inaccuracies of a single example. You're nitpicking something completely irrelevant.

[Person_Man]

I like Flickerdart's task resolution system. I wonder if it could be improved even further, or what kind of pitfalls this system would encounter.

I love it as well. I think the key is that the Skills just need to be well written. You need to be clear about when you should resolve a task through and when you need an opposed check. (For example, Tumble should be an opposed Dex check, not a set task). And you need to instruct DMs that the Difficulty and Success Target (or whatever they're called) can and should be modified according to various circumstances. And the Skills themselves shouldn't be so granular that you need to invest in multiple Skills to achieve a single basic action. (For example, you shouldn't have to Hide and Move Silently in order to accomplish the opposed check of not being noticed by an enemy).

[noparlpf]

This is fiction, leave what happens in reality at the door and stop complaining about the inaccuracies of a single example. You're nitpicking something completely irrelevant.

If you haven't noticed, a lot of people get upset when skills are completely unrealistic. Therefore, it is not completely irrelevant to request that skills be at least slightly realistic at the baseline.

[Flickerdart]

I posted some more thoughts on this idea in the previous thread, but someone pulled the thread out from under me while I was doing so. So, you know, it's there, but comment on it here.

Also, I think "10 feet of perfectly smooth stone" versus "100 feet of tree" is a pretty good example of high DC vs high success target. The stone might be something like DC 25 but have a success value of 5; if you're such a good climber that you can climb up smooth stone (like, you have suction cups or are a spider) then it's not super hard for you to do so, but anyone who isn't as skilled as you (say, a climber of moderate ability with a +5 to the check) would need five natural 20s to climb it, or an average of 100 attempts (assuming he doesn't ever fall off). Meanwhile, the tree can have a DC of 0 and a success target of 150; the average climber only needs 10 rounds to scale it, and the super awesome expert climber might do so in 6 rounds.

[Scow2]

I don't like Flickerdart's suggestion, especially because it doesn't really apply to anything that DOESN'T have visible progress. 3.5 already had two 'skills' that had a progress bar - Crafting, and Chopping Down A Door.

These kinds of checks CERTAINLY don't make sense for, say, Jump Checks - either you jump the distance you're trying to make, or you don't. You can't jump on air (Without a spell or supernatural ability at least), and successive jumps aren't usually higher than the previous.

Swingy mechanics keep the skills relevant without trivializing the challenge, while not requiring obscene investment in order to make it a possible task. D&D is designed to encourage taking risks.

...Flickerdart's later explanation in the now-dead thread swayed me over to his side, though. Yeah, it really does work well, especially given 5e's handling of accuracy.

[Ashdate]

This is fiction, leave what happens in reality at the door and stop complaining about the inaccuracies of a single example. You're nitpicking something completely irrelevant.

Well, I think it's interesting that Flickerdart's proposal can handle both extremes; one where success might literally hinge on rolling high once, or another where success might hinge on a careful pace.

As an example, you could set climbing a wall with a high progress DC, but a low success target to represent a very smooth wall that is relatively short (say, progress DC 15, success target 5), or alternatively, have a very easy to climb wall that is fairly large (say, progress DC 5, success target 15).

It's certainly more complicated than a single roll DC, but I think it opens up some interesting modelling possibilities. I especially like the possibilities for using it within contests:

Conan the 16 strength barbarian goes against Wimpy the 9 strength commoner. Each character spends a round rolling against the other; Conan has a progress DC of 9, and a success target of 9 against Wimpy. Wimpy has a progress DC of 16, and a success target of 16.

Can Wimpy win? Yes, but it's pretty slim odds.

Yet another possibility:

Conan the 16 strength barbarian and Wimpy the 9 strength commoner are going against an 22 strength Giant in a contest of strength; Conan and Wimpy are trying to push the giant over a cliff (and vice-versa).

Both Conan and Wimpy are therefore facing against a progress DC of 22, with a success target of 16. Both Conan and Wimpy's successes count towards the target of 22. This unfortunately means that Wimpy isn't going to be very helpful, but instead of rolling he can give Conan a +2 "aid another" bonus to his roll.

The Ogre meanwhile, uses the highest strength score between Conan and Wimpy (16) as his progress DC, but his success target is Conan and Wimpy's strength scores added together (9+16) therefore giving him a progress DC of 16, and a success target of 25.

Another possibility:

Conan the 16 strength barbarian and Wimpy the 9 strength commoner are going against a 22 strength Giant who is trying to bust through a reinforced door that Conan and Wimpy are bracing.

Conan and Wimpy aren't trying to bust through the door to get to the giant, so they don't roll. They're simply trying to keep the door shut.

The giant's progress DC is based off the strength of the door - it's reinforced, so the DM assigns it a DC 15 - but the progress DC is created by adding Conan and Wimpy's strength scores together (16 + 9 = 25) therefore making it progress DC 15, success target 25.

Certainly there's some math to consider in how this works (the low static modifiers to skills are potentially problematic here), but I think there's a skeleton of a pretty workable system there.

[Flickerdart]

These kinds of checks CERTAINLY don't make sense for, say, Jump Checks - either you jump the distance you're trying to make, or you don't. You can't jump on air (Without a spell or supernatural ability at least), and successive jumps aren't usually higher than the previous.

A jump check has a target success of 0. Not all skills must be low-DC, high-target success.

[Grod_The_Giant]

I don't like Flickerdart's suggestion, especially because it doesn't really apply to anything that DOESN'T have visible progress. 3.5 already had two 'skills' that had a progress bar - Crafting, and Chopping Down A Door.

These kinds of checks CERTAINLY don't make sense for, say, Jump Checks - either you jump the distance you're trying to make, or you don't. You can't jump on air (Without a spell or supernatural ability at least), and successive jumps aren't usually higher than the previous.

Swingy mechanics keep the skills relevant without trivializing the challenge, while not requiring obscene investment in order to make it a possible task. D&D is designed to encourage taking risks.

As I understand it, his system is for things that shouldn't be resolved with a single check.

Jumping across a pit trap would be a single Jump check. Jumping across a series of floating platforms Mario-style could use his system-- the DC being the difficult of jumping from one platform to another, and the success being based on the number of platforms.

[SiuiS]

I used to climb a lot, and a cliff with no handholds is going to take wayyy longer to climb even twenty feet than it would to climb a hundred-foot fir.

Not really? You're expected to have adventuring equipment and be the kind o person who can climb sheer stone with daggers and grit.

everyone has daggers and grit.

I posted some more thoughts on this idea in the previous thread, but someone pulled the thread out from under me while I was doing so. So, you know, it's there, but comment on it here.

Here ya go!

Spoiler

Show

((for reference))

You probably wouldn't raise the success target except in cases where objects are specifically crafted to do so. A lock, no matter how well crafted, only has so many tumblers, they're just harder to trip than on a shoddy lock. If you wanted to raise the success target, you could just add more locks; a rogue that got a result of 40 when unlocking a chest with three DC 10/success target 10 locks opens them all in one fell swoop, since it's really just a DC 10/success target 30 lock.

In the interest of exploring the viability of this, let's take a look at some more things one does as a skill monkey.

Trapfinding: Searching individual 5ft squares is horrible. A trap finder rolls a Search check; the DC represents the difficulty of finding a trap, the success target represents the number of traps. A trap with DC 25 to find in the old system might have a DC of 20 and target number of 5; thus, in a room with four such traps, a rogue that gets a result of 40 spots them all.

Disarming: Same deal as lockpicking.

Stealth: This one can be handled as simply an opposed check like always. Alternatively, you can have the rogue roll a target number (as above) and then the guard gets to make a Perception check against that. The target number is subtracted from the rogue's roll through some math; the rogue has a limited time frame in which to operate before the guard catches on. More experienced rogues have a lower subtracted number.

Actually, this gives me an idea. If Next hates scaling so much...this system wouldn't need to scale a lot, because the important math (how much of the DC is DC and how much is target number) is more or less a fixed total. A 20th level rogue might not have a much higher Stealth check than a 1st level rogue (let's say, a difference between +5 (+2 Dex and Skill Focus) and +11 (+5 Dex, skill focus, +2 masterwork boots, and a +1 from class features or something). If they both roll 10, that's a result of 15 for the level 1 rogue and 20 for the level 20 rogue. But the level 20 rogue's target number could be very small due to class features and such; so while a level 1 rogue's target number might be 5, and thus the DC to make progress on finding the rogue is 10, the 20th level rogue might have a target number of 0, so that the DC to spot him is a full 21. This kind of makes sense - a 1st level rogue will leave footprints and forget to close doors and stuff like that, so an average observer has a 50% chance of simply running into one of those signs and going to investigate. At this point, the rogue knows he should get out of there quickly! On the other hand, the 20th level rogue leaves no such traces. If anyone knows he was there, it's because they saw him move. The untrained person has no chance of spotting the level 20 rogue, even though his check is only 6 higher.

Target number modifiers are a pretty good example of something that might differentiate a PC from an NPC. While Joe Paladin with 0 Perception might be able to spot a skulking goblin after a few minutes of said goblin trying to sabotage the water supply, Gragnak the Warrior would fail to notice Joe's companion Joanna doing the same thing to the goblin water, even if the results of the skill checks were exactly the same.

I don't like Flickerdart's suggestion, especially because it doesn't really apply to anything that DOESN'T have visible progress. 3.5 already had two 'skills' that had a progress bar - Crafting, and Chopping Down A Door.

These kinds of checks CERTAINLY don't make sense for, say, Jump Checks - either you jump the distance you're trying to make, or you don't. You can't jump on air (Without a spell or supernatural ability at least), and successive jumps aren't usually higher than the previous.

Swingy mechanics keep the skills relevant without trivializing the challenge, while not requiring obscene investment in order to make it a possible task. D&D is designed to encourage taking risks.

...Flickerdart's later explanation in the now-dead thread swayed me over to his side, though. Yeah, it really does work well, especially given 5e's handling of accuracy.

Trouble! How do you handle a long jump that takes more than 6 seconds? Those situations where the player jumps, but runs out of movement before completion? Because 3.5 had two ways of handling it RAW. Either they purposefully declared the "begin full round action" as their jump, kept, and finished the movement on their next round, or they kept, ran out of movement despite having sufficient Jump to make the distance.... And suddenly stop dead and plummet hundreds of feet mid-round.

This thing here? Might fix it.

[Scow2]

Trouble! How do you handle a long jump that takes more than 6 seconds? Those situations where the player jumps, but runs out of movement before completion? Because 3.5 had two ways of handling it RAW. Either they purposefully declared the "begin full round action" as their jump, kept, and finished the movement on their next round, or they kept, ran out of movement despite having sufficient Jump to make the distance.... And suddenly stop dead and plummet hundreds of feet mid-round.

This thing here? Might fix it.

Not really. In 3.5. when making a jump, you make the skill check to bridge the gap. If you make it, nothing can interfere with that jump - you land the next round after moving the remaining distance. In 5e, you can make a jump check to go the distance, make a second jump check to perform your action mid-jump, and then move the rest of your speed. If you don't have the rest of your speed to finish, you stay "suspended" (But not falling - you're still jumping) and you can make a jump check to take your action next round, then move to land (Or move on either side of the action). But only one jump check's needed to see if you clear the distance. Of course, this leads to silliness where taking mid-jump stunts decreases your velocity.

[Raineh Daze]

If you haven't noticed, a lot of people get upset when skills are completely unrealistic. Therefore, it is not completely irrelevant to request that skills be at least slightly realistic at the baseline.

It is realistic.

If you have a realistic climbing skill, and a high DC, then it will take longer for you to accumulate the necessary rolls to reach the required target and climb the short sheer surface. A fir, meanwhile, you're unlikely to fail the rolls for, and you'll therefore make a lot more progress per roll--but if you roll poorly, it could still take longer.

If you have a really high climbing skill, such that you're unlikely to fail either, then you're going to get up the short surface quickly (low target) and the high tree faster than someone worse at climbing (high target, but low DC = more progress per roll).

[Icewraith]

With regards to sneaking, I think a reversal of the proposed procedure would work better. When the players are sneaking you want them to be actively rolling, and when you're trying to sneak something past them you don't want them to know it's there. Having people keep rolling perception or whatever until nothing happens is a dead giveaway, unless you're really into playing metagaming head games with your players.

It would be better for the awareness or perception or whatever to set the progress DC and the target be something like 20 + 1/5/10 per round the sneaker has been within long/medium/short range of the targets. Hitting the target lets you stay hidden in whatever square you occupy without having to make further stealth checks unless something unreasonable happens or you shoot someone, moving resets the counter and the time penalty starts accruing again. Flub the progress DC and you are noticed.

[Person_Man]

I posted some more thoughts on this idea in the previous thread, but someone pulled the thread out from under me while I was doing so. So, you know, it's there, but comment on it here.

Also, I think "10 feet of perfectly smooth stone" versus "100 feet of tree" is a pretty good example of high DC vs high success target. The stone might be something like DC 25 but have a success value of 5; if you're such a good climber that you can climb up smooth stone (like, you have suction cups or are a spider) then it's not super hard for you to do so, but anyone who isn't as skilled as you (say, a climber of moderate ability with a +5 to the check) would need five natural 20s to climb it, or an average of 100 attempts (assuming he doesn't ever fall off). Meanwhile, the tree can have a DC of 0 and a success target of 150; the average climber only needs 10 rounds to scale it, and the super awesome expert climber might do so in 6 rounds.

Very minor quibble. You might want to make the terminology different from previous editions, so that it doesn't get confused with similar but different rules from previous editions that had the same name. Maybe Difficulty Modifier (which gets subtracted from your roll) and Skill Target (the total result(s) you must accumulate in order to succeed).

Other then that, I'd like to reiterate that you should post your idea over on the WotC D&D Next forum, because it's a great idea.

[SiuiS]

With regards to sneaking, I think a reversal of the proposed procedure would work better. When the players are sneaking you want them to be actively rolling, and when you're trying to sneak something past them you don't want them to know it's there. Having people keep rolling perception or whatever until nothing happens is a dead giveaway, unless you're really into playing metagaming head games with your players.

It would be better for the awareness or perception or whatever to set the progress DC and the target be something like 20 + 1/5/10 per round the sneaker has been within long/medium/short range of the targets. Hitting the target lets you stay hidden in whatever square you occupy without having to make further stealth checks unless something unreasonable happens or you shoot someone, moving resets the counter and the time penalty starts accruing again. Flub the progress DC and you are noticed.

Maybe reverse the base, too, so that your stealth score gets slowly eroded by constant vigilance? You roll stealth, get your Stealth HP. Whenever something suspicious happens to make a guard actually check for you, or to dampen your stealth (light coming on, noise near you, new guard walking up, animal being disturbed) the enemy makes progress toward beating your stealth roll?

[Raineh Daze]

This would end up with an invisible master thief being caught simply because there's a bunch of normal town guards around. That's a bit... bad.

[Conundrum]

Also, I think "10 feet of perfectly smooth stone" versus "100 feet of tree" is a pretty good example of high DC vs high success target. The stone might be something like DC 25 but have a success value of 5; if you're such a good climber that you can climb up smooth stone (like, you have suction cups or are a spider) then it's not super hard for you to do so, but anyone who isn't as skilled as you (say, a climber of moderate ability with a +5 to the check) would need five natural 20s to climb it, or an average of 100 attempts (assuming he doesn't ever fall off). Meanwhile, the tree can have a DC of 0 and a success target of 150; the average climber only needs 10 rounds to scale it, and the super awesome expert climber might do so in 6 rounds.

Emphasis mine.

This raises an interesting point - how do you handle "failing" a task under such a system? It'd need to be on a skill-by-skill basis, almost... if you roll below the DC while climbing, do you fall off? If you roll below the DC while picking a lock, does the lock reset?

Having said that, I really do like the system and I'm sure you have an answer for this!

[Scow2]

This would end up with an invisible master thief being caught simply because there's a bunch of normal town guards around. That's a bit... bad.

No such thing as an "Invisible master thief" - 5th edition is deliberately avoiding that level of skill stratification.

The way to avoid having your cover blown by having too many guards on your case is to minimize the number of guards you're exposed to at any given time. A stealth check every round to replenish the pool might work in extended stealth missions. Heck... that could actually make stealth missions fun, and do a good job of simulating a "Detection meter"

[Raineh Daze]

No such thing as an "Invisible master thief" - 5th edition is deliberately avoiding that level of skill stratification.

The way to avoid having your cover blown by having too many guards on your case is to minimize the number of guards you're exposed to at any given time. A stealth check every round to replenish the pool might work in extended stealth missions. Heck... that could actually make stealth missions fun, and do a good job of simulating a "Detection meter"

Have they removed the concept of invisibility, then? I meant a literal invisible master thief (whether they're stealing masters or just a level-capped thiefy type is up to you).

[Kornaki]

Certainly there exists some density of town guards for which it's too risky to try to sneak past (one every 5 feet is an upper bound, but if there's like one every ten feet and they're all milling about there should be a chance that one of them just turns around suddenly and bumps into you)

[TuggyNE]

Maybe reverse the base, too, so that your stealth score gets slowly eroded by constant vigilance? You roll stealth, get your Stealth HP. Whenever something suspicious happens to make a guard actually check for you, or to dampen your stealth (light coming on, noise near you, new guard walking up, animal being disturbed) the enemy makes progress toward beating your stealth roll?

That would work well, assuming they have to beat a DC to make progress toward spotting you. (If it's just "auto-lose X Stealth points", that's lame.)

This would end up with an invisible master thief being caught simply because there's a bunch of normal town guards around. That's a bit... bad.

Nah, that actually makes sense, as long as they have some small chance of making progress, a large number of watchers is indeed quite dangerous to be around.

This raises an interesting point - how do you handle "failing" a task under such a system? It'd need to be on a skill-by-skill basis, almost... if you roll below the DC while climbing, do you fall off? If you roll below the DC while picking a lock, does the lock reset?

Generally, "critical"/substantial failures already have been customized to the skill, so it just needs a bit more. Failing a Climb check in 3.5 by 5 points or more means you fall, for example.

[Raineh Daze]

Nah, that actually makes sense, as long as they have some small chance of making progress, a large number of watchers is indeed quite dangerous to be around.

That was in response to the idea of an automatic loss of 'stealth HP' just because, say, a guard showed up.

[Zeful]

You probably wouldn't raise the success target except in cases where objects are specifically crafted to do so. A lock, no matter how well crafted, only has so many tumblers.

If it has tumblers at all. Not all locks need them. I've designed a couple of locks that require the key to open, mostly because of the underlying argument your making here.

[Grod_The_Giant]

You could have, say, a DC based on your Stealth skill rank, and skill HP based on your roll. So if you're supposed to be really stealthy, it's harder for guards to make any progress, regardless of your roll.