Mathematical Musings: On Skills, Scaling, and Advantage

After playing with the skill die for a number of sessions I have come to the conclusion that I do not like it. I like the intention behind it, I like the math of it, but I do not like how awkward the rolling of it works out. The skill die gets especially confusing when you throw in rogues skill mastery and advantage. There are easier ways to represent competency. I think scaling overall could be better presented, and I think that another mechanic that already exists in 5e might better represent skill training.

Part 1: Scaling

Now many people will probably disagree with this, but I think minor level based scaling should return. I know that "accuracy is bounded" in 5e, and I don't suggest changing that. I would like the see PCs grow in capability as they level in a numerical sense. We can see some automatic scaling already with fighter attack bonuses and the skill die improving from d4 to d12. The scaling I propose is +1 to all ability checks at 4th level and every 4 levels thereafter. This would apply to ability checks used for skill rolls, attack rolls, and saving throws. This bonus represents a high level PCs increased capability through skill and experience. At 20th level the bonus is still relatively minor at only +5. This bonus would also only apply to PCs. Monsters would not have recieve this level based scaling.

This bonus would replace the ability score increases at those levels and the increasing attack/save DC bonuses. A class who is better at combat (like the fighter) would simply recieve a minor bonus (+1)  to weapon attacks from level 1 but have no need of a further scaling bonus to attack. Instead their superiority with weapons comes from increased weapon damage much like we have now. *This scaling might also need to be applied to AC.

Part 2: Skills

As I stated earlier, I do not like the skill die, but I think 5e already has a mechanic that can represent skill training quite easily - Advantage. Rolling 2d20 and choosing the best result has a few key features that makes it ideal for skill training. It doesn't raise the maximum value but does increase the tendency for higher values. This means a trained PC is more likely to succeed at a typical task and is less reliant on the luck of the wildly variable d20. This also means that variance between trained individuals and untrained individuals is lessened so that everyone still has the opportunity to participate.

For those who want to specialize in skills even further, I suggest a skill mastery feat which grants a 3rd d20 to the mix. This will even further push up the average and drive down the likelyhood of failure. Rolling more d20s helps to shift the focus from luck to a PCs competency. 

So combined with part 1's level based scaling we have a difficulty chart that looks like this:

        +10 bonus (+5 ability mod and level 20)       +5 bonus (+5 ability mod and level 1 or +0 and 20)

Difficulty  Number of Dice - Chance of Success                Number of Dice - Chance of Success
DC 30    1d20 - 5%     2d20 - 10%   3d20 - 14%          1d20 - 0%     2d20 - 0%     3d20 - 0%
DC 26    1d20 - 25%   2d20 - 44%   3d20 - 58%          1d20 - 0%     2d20 - 0%     3d20 - 0%
DC 22    1d20 - 45%   2d20 - 70%   3d20 - 83%          1d20 - 20%   2d20 - 36%   3d20 - 49%
DC 18    1d20 - 65%   2d20 - 88%   3d20 - 96%          1d20 - 40%   2d20 -64%    3d20 - 78%
DC 14    1d20 - 85%   2d20 - 98%   3d20 - 100%        1d20 - 60%   2d20 - 84%   3d20 - 94%
DC 10    1d20 - 100% 2d20 - 100% 3d20 - 100%        1d20 - 80%   2d20 - 96%   3d20 - 99%

So the most difficult tasks (DC30) can only be accomplished by those with a 20 ability score at level 20 with only a small chance of success. As PCs advance in level they become more likely to accomplish the easier DC 10-18 tasks, but the very difficult tasks (DC22+) remain out of reach to to most low level adventurers and are still difficult to high level trained individuals. Also, as ability scores do not increase with level, having an 18-20 score on a PC will be much harder to find so DC 30 tasks can truly be godlike in difficulty. Note that while similar to the current 5e design, I have reduced the top end DCs by 5 (reducing the DC of most tasks overall).

Edit: Multiple Dice also allows for tasks with varying complexity requiring multiple successes. A reinforced door might require two DC 15 checks to break down. A untrained 10 strength wizard will not be able to break down such a door with a single lucky roll, but a trained 16 strength warrior could.

Advanced Options:
Show
Lastly advantage could be easily added with this system by having advantage grant an additional d20 to roll (4d20 for a master with advantage). If the number of d20s PCs rolled for attack rolls also increased (perhaps to represent multiple attacks), all minor (+1, +2, +5) bonuses could be removed from the game entirely in favor of a universal advantage mechanic. When a PC is already rolling 2d20, a 3rd d20 is only worth about a +2 bonus. This means conditions that grant these small bonuses could be consolidated within the advantage mechanic. Flanking, attacking a prone enemy, etc could all simply grant an extra d20. This has the benefit of removing conditional modifiers which tend to be a big hangup/slowdown for many groups. Remembering to roll an extra d20 is easier than remembering a +1 here and a +2 there.

My 5e Homebrew Material

The Warblade: A Mythic Fighter

The Hero: A Modular Class

1) I'd rather see the stats increase directly (+1 to all stats).

2) Disagree.  For the same reason i don't like the current skill dice.  Mainly that it's "all or nothing".  You can't get "minor advantage".  Though the current skill dice can be modified to work with that.

guides
List of no-action attacks.
Dynamic vs Static Bonuses
Phalanx tactics and builds
Crivens! A Pictsies Guide Good
Power
s to intentionally miss with
Mr. Cellophane: How to be unnoticed
Way's to fire around corners
Crits: what their really worth
Retroactive bonus vs Static bonus.
Runepriest handbook & discussion thread
Holy Symbols to hang around your neck
Ways to Gain or Downgrade Actions
List of bonuses to saving throws
The Ghost with the Most (revenant handbook)
my builds
F-111 Interdictor Long (200+ squares) distance ally teleporter. With some warlord stuff. Broken in a plot way, not a power way.

Thought Switch Higher level build that grants upto 14 attacks on turn 1. If your allies play along, it's broken.

Elven Critters Crit op with crit generation. 5 of these will end anything. Broken.

King Fisher Optimized net user.  Moderate.

Boominator Fun catch-22 booming blade build with either strong or completely broken damage depending on your reading.

Very Distracting Warlock Lot's of dazing and major penalties to hit. Overpowered.

Pocket Protector Pixie Stealth Knight. Maximizing the defender's aura by being in an ally's/enemy's square.

Yakuza NinjIntimiAdin: Perma-stealth Striker that offers a little protection for ally's, and can intimidate bloodied enemies. Very Strong.

Chargeburgler with cheese Ranged attacks at the end of a charge along with perma-stealth. Solid, could be overpowered if tweaked.

Void Defender Defends giving a penalty to hit anyone but him, then removing himself from play. Can get somewhat broken in epic.

Scry and Die Attacking from around corners, while staying hidden. Moderate to broken, depending on the situation.

Skimisher Fly in, attack, and fly away. Also prevents enemies from coming close. Moderate to Broken depending on the enemy, but shouldn't make the game un-fun, as the rest of your team is at risk, and you have enough weaknesses.

Indestructible Simply won't die, even if you sleep though combat.  One of THE most abusive character in 4e.

Sir Robin (Bravely Charge Away) He automatically slows and pushes an enemy (5 squares), while charging away. Hard to rate it's power level, since it's terrain dependent.

Death's Gatekeeper A fun twist on a healic, making your party "unkillable". Overpowered to Broken, but shouldn't actually make the game un-fun, just TPK proof.

Death's Gatekeeper mk2, (Stealth Edition) Make your party "unkillable", and you hidden, while doing solid damage. Stronger then the above, but also easier for a DM to shut down. Broken, until your DM get's enough of it.

Domination and Death Dominate everything then kill them quickly. Only works @ 30, but is broken multiple ways.

Battlemind Mc Prone-Daze Protecting your allies by keeping enemies away. Quite powerful.

The Retaliator Getting hit deals more damage to the enemy then you receive yourself, and you can take plenty of hits. Heavy item dependency, Broken.

Dead Kobold Transit Teleports 98 squares a turn, and can bring someone along for the ride. Not fully built, so i can't judge the power.

Psilent Guardian Protect your allies, while being invisible. Overpowered, possibly broken.

Rune of Vengance Do lot's of damage while boosting your teams. Strong to slightly overpowered.

Charedent BarrageA charging ardent. Fine in a normal team, overpowered if there are 2 together, and easily broken in teams of 5.

Super Knight A tough, sticky, high damage knight. Strong.

Super Duper Knight Basically the same as super knight with items, making it far more broken.

Mora, the unkillable avenger Solid damage, while being neigh indestuctable. Overpowered, but not broken.

Swordburst Maximus At-Will Close Burst 3 that slide and prones. Protects allies with off actions. Strong, possibly over powered with the right party.

The scaling I propose is +1 to all ability checks at 4th level and every 4 levels thereafter. This would apply to ability checks used for skill rolls, attack rolls, and saving throws. This bonus represents a high level PCs increased capability through skill and experience. At 20th level the bonus is still relatively minor at only +5. This bonus would also only apply to PCs. Monsters would not have recieve this level based scaling.

Agreed, although the bonus would need to apply to humanoid-NPCs who possessed anything like class levels. That's a relatively small subset of "monsters", but it would serve to re-inforce the meaning of levels.

This bonus would replace the ability score increases at those levels and the increasing attack/save DC bonuses. A class who is better at combat (like the fighter) would simply recieve a minor bonus (+1)  to weapon attacks from level 1 but have no need of a further scaling bonus to attack. Instead their superiority with weapons comes from increased weapon damage much like we have now.

Probably my favorite part of 4E was how fighters were just flat better at fighting, rather than increasingly better at fighting. The game should never assume that only the specialist should have a chance to succeed during ordinary circumstances, and increasing gaps only pigeonholed characters further.

As I stated earlier, I do not like the skill die, but I think 5e already has a mechanic that can represent skill training quite easily - Advantage. Rolling 2d20 and choosing the best result has a few key features that makes it ideal for skill training. It doesn't raise the maximum value but does increase the tendency for higher values. This means a trained PC is more likely to succeed at a typical task and is less reliant on the luck of the wildly variable d20. This also means that variance between trained individuals and untrained individuals is lessened so that everyone still has the opportunity to participate.

I really like the idea of attaching 2d20 to enduring traits of the character, rather than leaving it up to contingent environmental circumstances. Skill checks come up rarely enough that anyone with proper training really should shine and have minimal failure chance.

The metagame is not the game.


2) Problem with your system is you can't make a character who is "very good at climbing", for example, if skills don't raise the maximum DC of tasks you can reach.

Yeah you can make a character with high Str but that's saying "my character is Arnold Schwarzenegger" and not "he's really good at climbing."

No matter how many Advantage d20s you roll for being trained or for having skill mastery... you still can't beat a task that has a higher DC, you'll only score more high numbers in that same DC-margin you already had with 1 die.

With the system you propose, the only real way for a character to stand out in a particular task (say climbing for example) is to have a high base ability, which doesn't really represent being an expert in an area.
Even the +1 bonus every 4 levels would apply to all characters in all checks, so it wouldn't serve to differentiate your "expert climber" in that particular area from everyone else.

Specializing in a skill or task should allow your character to accomplish more difficult tasks (higher DCs) than those who don't train in that area.



That said, I do find the whole roll d20s plus other dice together somewhat odd and off.
Maybe I just need to get used to it, or maybe trade that for static bonuses.
With the system you propose, the only real way for a character to stand out in a particular task (say climbing for example) is to have a high base ability, which doesn't really represent being an expert in an area.

That part really makes sense to me, especially when it comes to physical skills.  If someone doesn't have the physical ability to ... say, lift themselves up by their fingertips... then no amount of practice will allow them to do so - unless it's just exercise that builds your strength the normal way. I can easily imagine a surface which is only climbable to someone with that sort of ability, and anyone incapable of that would have no chance - regardless of how much practice they've had climbing fences or trees or other surfaces where you can get a foothold.

It makes sense to me that the person with practice and training will be super consistent about their results, because they know how to make the most of what they have and not make dumb mistakes - which corresponds to consistently rolling high on the die roll.

Now, it doesn't do so great of a job with mental skills, so that's definitely a tradeoff worth considering.

The metagame is not the game.

Now many people will probably disagree with this, but I think minor level based scaling should return. I know that "accuracy is bounded" in 5e, and I don't suggest changing that. I would like the see PCs grow in capability as they level in a numerical sense. We can see some automatic scaling already with fighter attack bonuses and the skill die improving from d4 to d12.


I think, to a large extent, that means precisely that you are suggesting changing that.  There is a big difference between particular classes getting level-based bonuses to attack or skill dice increasing (which only applies to certain rolls based upon character-defined qualities) and everyone receiving a level-based bonus on most/all d20 rolls all the time. 

The scaling I propose is +1 to all ability checks at 4th level and every 4 levels thereafter. This would apply to ability checks used for skill rolls, attack rolls, and saving throws. This bonus represents a high level PCs increased capability through skill and experience. At 20th level the bonus is still relatively minor at only +5. This bonus would also only apply to PCs. Monsters would not have recieve this level based scaling.


Define "relatively minor."  The difference between a character that receives a +5 bonus to attack rolls, ability rolls and saving throws (and perhaps AC) and a character that does not is enormous, all other things being equal.  Such a change would have ripple effects throughout the game, and you would find for example that monsters absolutely must be designed with this in mind.

I'm not saying that can't work, however it doesn't seem to be the direction the devs are going, and I for one am happy with that.  I don't want to see the degree of increasing specialization of characters that 3E had, where characters could reasonably participate in similar activities at low levels but became worlds apart as time went on.  I also don't think I want to see automatic bonuses to the extent of 4E.  I like the goal of allowing characters to develop in specific areas without moving non-specialized characters or lower-difficulty challenges into obselence.  In this sense, it seems we mostly agree on the overall goal, but with an important distinction.

Part 2: Skills

As I stated earlier, I do not like the skill die, but I think 5e already has a mechanic that can represent skill training quite easily - Advantage. Rolling 2d20 and choosing the best result has a few key features that makes it ideal for skill training. It doesn't raise the maximum value but does increase the tendency for higher values. This means a trained PC is more likely to succeed at a typical task and is less reliant on the luck of the wildly variable d20. This also means that variance between trained individuals and untrained individuals is lessened so that everyone still has the opportunity to participate.

For those who want to specialize in skills even further, I suggest a skill mastery feat which grants a 3rd d20 to the mix. This will even further push up the average and drive down the likelyhood of failure. Rolling more d20s helps to shift the focus from luck to a PCs competency.

So combined with part 1's level based scaling we have a difficulty chart that looks like this:

        +10 bonus (+5 ability mod and level 20)       +5 bonus (+5 ability mod and level 1 or +0 and 20)

Difficulty  Number of Dice - Chance of Success                Number of Dice - Chance of Success
DC 30    1d20 - 5%     2d20 - 10%   3d20 - 14%          1d20 - 0%     2d20 - 0%     3d20 - 0%
DC 26    1d20 - 25%   2d20 - 44%   3d20 - 58%          1d20 - 0%     2d20 - 0%     3d20 - 0%
DC 22    1d20 - 45%   2d20 - 70%   3d20 - 83%          1d20 - 20%   2d20 - 36%   3d20 - 49%
DC 18    1d20 - 65%   2d20 - 88%   3d20 - 96%          1d20 - 40%   2d20 -64%    3d20 - 78%
DC 14    1d20 - 85%   2d20 - 98%   3d20 - 100%        1d20 - 60%   2d20 - 84%   3d20 - 94%
DC 10    1d20 - 100% 2d20 - 100% 3d20 - 100%        1d20 - 80%   2d20 - 96%   3d20 - 99%

So the most difficult tasks (DC30) can only be accomplished by those with a 20 ability score at level 20 with only a small chance of success. As PCs advance in level they become more likely to accomplish the easier DC 10-18 tasks, but the very difficult tasks (DC22+) remain out of reach to to most low level adventurers and are still difficult to high level trained individuals. Also, as ability scores do not increase with level, having an 18-20 score on a PC will be much harder to find so DC 30 tasks can truly be godlike in difficulty.

Lastly advantage could be easily added with this system by having advantage grant an additional d20 to roll (4d20 for a master with advantage). If the number of d20s PCs rolled for attack rolls also increased (perhaps to represent multiple attacks), all minor (+1, +2, +5) bonuses could be removed from the game entirely in favor of a universal advantage mechanic. When a PC is already rolling 2d20, a 3rd d20 is only worth about a +2 bonus. This means conditions that grant these small bonuses could be consolidated within the advantage mechanic. Flanking, attacking a prone enemy, etc could all simply grant an extra d20. This has the benefit of removing conditional modifiers which tend to be a big hangup/slowdown for many groups. Remembering to roll an extra d20 is easier than remembering a +1 here and a +2 there.


I'll ask a similar question: what do you mean by "a 3rd d20 is only worth about a +2 bonus"? (emphasis added.)  The number of d20s rolled acts as an exponent on the probability of failure, and that's pretty significant a lot of the time. 

I'm not sure I understand your goals or complaints.  The skill dice mechanic may make it possible for specialized characters to accomplish tasks that nonspecialized characters cannot, but such tasks never really become routine, and in many ways skill dice allow for greater customization and gradation than a strict pool-of-d20s mechanic.  What is it precisely about skill dice you find awkward?

2) Problem with your system is you can't make a character who is "very good at climbing", for example, if skills don't raise the maximum DC of tasks you can reach.

Yeah you can make a character with high Str but that's saying "my character is Arnold Schwarzenegger" and not "he's really good at climbing."

No matter how many Advantage d20s you roll for being trained or for having skill mastery... you still can't beat a task that has a higher DC, you'll only score more high numbers in that same DC-margin you already had with 1 die.

With the system you propose, the only real way for a character to stand out in a particular task (say climbing for example) is to have a high base ability, which doesn't really represent being an expert in an area.
Even the +1 bonus every 4 levels would apply to all characters in all checks, so it wouldn't serve to differentiate your "expert climber" in that particular area from everyone else.

Specializing in a skill or task should allow your character to accomplish more difficult tasks (higher DCs) than those who don't train in that area.



I want to "debunk" this myth you present. You have to think of things in terms of bounded accuracy. You aren't going to find DC 30 walls for the most part. In fact most things you will climb will be DC 22 and below. Even a DC 22 wall might be something as challenging as climbing a sheer cliff in the rain.

Let us say we have two average +0 ability modifier humans both trying to climb some walls. One is trained in climbing, the other is not.

The DC 11 easy challenge wall: Trained individual climbs this wall 75% of the time, while untrained one only 50%. A master climber climbs this wall 88% of the time.

DC 14 moderate challenge wall: The untrained individual only succeds at climbing this wall 35% of the time. The trained individual 58% and the master 73%. That seems pretty significant right there. The master is twice as likely to limb such a wall. Remember these are inexperienced (level 1) PCs with only average ability scores.

The DC 18 difficult wall: Untrained climber only succeeds 15% of the time. Trained individual 28% and master 39%. This is a very tough wall, but the trained individual is almost twice as likely to succeed as the untrained one. The master 2.5 times as likely.

The DC 22 very difficult sheer cliff in the rain: It is too much for our inexperienced climbers to tackle. At level 20 though after gaining much climbing experience during the course of their dungeon delving. Untrained individual 20%. Trained 36%. Master 49%. Still a difficult task even though our PCs are challenging gods in combat. The master is clearly the one to choose for this task though as the untrained individual has a fairly good chance of falling to his death (failure by 5+).
Define "relatively minor."  The difference between a character that receives a +5 bonus to attack rolls, ability rolls and saving throws (and perhaps AC) and a character that does not is enormous, all other things being equal.  Such a change would have ripple effects throughout the game, and you would find for example that monsters absolutely must be designed with this in mind.

One side effect of this change is that, by reducing the accuracy/AC of low level monsters compared to high level ones, you no longer need to scale HP/damage so quickly in order to differentiate them.  If a goblin has a 35% chance to hit you, regardless of your level, then you need to have a lot more HP at level 16 to make the goblin as ineffective as it would otherwise be if you'd scaled its chance to hit down to 15 percent.

The metagame is not the game.

If someone doesn't have the physical ability to ... say, lift themselves up by their fingertips... then no amount of practice will allow them to do so






Well, see? I'm sure there's plenty of bodybuilders out there who are 10x stronger than him and can't even begin to hope accomplishing that.
Cause that guy's just really good at doing what you said... ;P
So he can do what for most is impossible.
Well, see? I'm sure there's plenty of bodybuilders out there who are 10x stronger than him and can't even begin to hope accomplishing that.
Cause that guy's just really good at doing what you said... ;P
So he can do what for most is impossible.

That's just physical strength, though. Granted, it's an entirely different sort of physical strength than bodybuilders have, but that's a system limitation where you only have three stats to represent the entire spectrum of physical capabilities. I really don't want to play a system where you can have different strength scores for your fingers relative to your wrists, legs, or back.

The metagame is not the game.

1) I'd rather see the stats increase directly (+1 to all stats).



I thought about +1 to all ability scores at every even level, but that didn't sit well with me. I already don't like the +1 to 2 ability scores every 4 levels. I mean why should an adventurer miraculously get stronger, smarter, more dexterous etc. Most adventurers seem to be in peak physical/mental shape, and the adventuring life doesn't lead itself to self improvement in a such a way that would meaningfully impact the base attributes.

What adventuring does give you is skill and experience. Repeatedly performing tasks helps you do them better. You might not be getting stronger, but you get better at knowing how to use your strength. Listening to the wizard babble about arcane lore for 20 levels and you pick up a thing or two. The "experience bonus" to rolls feals more natural in that your PC learns how to better use his abilities with level.

2) Disagree.  For the same reason i don't like the current skill dice.  Mainly that it's "all or nothing".  You can't get "minor advantage".  Though the current skill dice can be modified to work with that.



While there is not much granularity, my proposed skill mastery feat leads to three tiers of competency: untrained, trained, and master. It also allows for gaining an additional die through advantage or losing a die through disadvantage much more fluidly.

Hell a 4th tier could be added with Grandmaster Skill Training feat that grants the much sought after +3 bonus!
That's just physical strength, though. Granted, it's an entirely different sort of physical strength than bodybuilders have, but that's a system limitation where you only have three stats to represent the entire spectrum of physical capabilities. I really don't want to play a system where you can have different strength scores for your fingers relative to your wrists, legs, or back.




Well it's physical strength applied to an action that requires specific training. That training gave him the ability to perform a feat his overall body strength normally wouldn't allow.

If you had a skill in game called "Lift yourself with the fingertips" and added a Die (or a flat bonus if you like it better), it would represent just that.
You add the possibility of performing that particular task to a level you normally wouldn't be able to.

This way we don't need 17 different strength scores to represent each body part.

I think training should be in the form of a bonus to your check (either skill die or flat bonus), because it states that you're good at that particular task and can do more difficult things others might not even be able to.
It's like the other bonuses you get while leveling: To Hit, Damage, etc, because you're becoming better, more trained in combat, or magic or whatever.

And then, circumstancial modifiers should fall into the Adv/Disadv rules.
If you're prone, or fatigued, or maybe being helped with the Aid Another action, you will have a better or worse chance of accomplishing tasks you normally can do.
But you won't be able to, because of just that, do things you normally can't.
The circumstances are merely making it easier or harder for you to do what you know how to do. 
Define "relatively minor."  The difference between a character that receives a +5 bonus to attack rolls, ability rolls and saving throws (and perhaps AC) and a character that does not is enormous, all other things being equal.  Such a change would have ripple effects throughout the game, and you would find for example that monsters absolutely must be designed with this in mind.

One side effect of this change is that, by reducing the accuracy/AC of low level monsters compared to high level ones, you no longer need to scale HP/damage so quickly in order to differentiate them.  If a goblin has a 35% chance to hit you, regardless of your level, then you need to have a lot more HP at level 16 to make the goblin as ineffective as it would otherwise be if you'd scaled its chance to hit down to 15 percent.



Are you saying you're in favour of this suggestion or something similar, because it allows you to avoid the hit point inflation necessary to achieve level disparities similar to what occurs with this suggestion in place?

If so, then what makes you think the devs want to achieve those disparities? 

I fing rolling the additional die and adding it to the d20 fun, swingy sure, but fun. I could allowing players to choose a static modifier or chance it on a die roll. Epic success come with equal chance of Epic failure!
I would think a static bonus to skill and then apply the die modifier to ability. Natural talent is never as consistent as training.


Ability
10 no bonus
12 +1
14 +1d4
16 +1d6
18 +1d8
20 +1d10     

Disclaimer: Wizards of the Coast is not responsible for the consequences of any failed saving throw, including but not limited to petrification, poison, death magic, dragon breath, spells, or vorpal sword-related decapitations.

Are you saying you're in favour of this suggestion or something similar, because it allows you to avoid the hit point inflation necessary to achieve level disparities similar to what occurs with this suggestion in place?

If so, then what makes you think the devs want to achieve those disparities? 

The disparity already exists, at least as of the latest packet. Assuming that the designers actually want that much difference between high levels and low levels, then I'm in favor of small +hit/AC bonuses rather than placing the burden entirely on HP/damage.

If the designers don't want that disparity between high levels and low levels, then I'm still in favor of putting the difference on +hit/AC instead of HP/damage, but at that point it's just personal preference.

The metagame is not the game.

That's just physical strength, though. Granted, it's an entirely different sort of physical strength than bodybuilders have, but that's a system limitation where you only have three stats to represent the entire spectrum of physical capabilities. I really don't want to play a system where you can have different strength scores for your fingers relative to your wrists, legs, or back.




Well it's physical strength applied to an action that requires specific training. That training gave him the ability to perform a feat his overall body strength normally wouldn't allow.

If you had a skill in game called "Lift yourself with the fingertips" and added a Die (or a flat bonus if you like it better), it would represent just that.
You add the possibility of performing that particular task to a level you normally wouldn't be able to.

This way we don't need 17 different strength scores to represent each body part.

I think training should be in the form of a bonus to your check (either skill die or flat bonus), because it states that you're good at that particular task and can do more difficult things others might not even be able to.



I personally don't believe that skill works the way you believe it does.

Take darts for example. Let's say we have 3 individuals playing darts in a bar all with equal hand eye coordination. One just started playing, one is a bar regular who plays once a week, and the other plays darts everyday. Let's say they all hit bullseyes with a roll of 17+ or a specific number (lets say 20s) on 9+
 
With my method we have untrained guy hitting bullseyes 20% of the time, trained guy hitting bulls 36% of the time, and master guy hitting bulls 49% of the time. For the specific number we have untrained guy at 60%, trained guy at 84%, and master at 94%

If we had training grant a flat +3/+5 bonus instead, trained guy hits bulls 35% and master 45% The variance for this is really wild. Instead of clustering darts around the target, the masters throws are all over the place, just like the brand new dart throwes. For hitting the 20, the trained guy hits it 75% and the master only 85%. You can really see how skill doesn't narrow the variance with only a flat bonus.

The extra d20s thus seems more realistic to me. A trained individual is more likely to repeat good results, while an untrained one can get good results but is unable to reliably repeat them.

2) Problem with your system is you can't make a character who is "very good at climbing", for example, if skills don't raise the maximum DC of tasks you can reach.

Yeah you can make a character with high Str but that's saying "my character is Arnold Schwarzenegger" and not "he's really good at climbing."

No matter how many Advantage d20s you roll for being trained or for having skill mastery... you still can't beat a task that has a higher DC, you'll only score more high numbers in that same DC-margin you already had with 1 die.

With the system you propose, the only real way for a character to stand out in a particular task (say climbing for example) is to have a high base ability, which doesn't really represent being an expert in an area.
Even the +1 bonus every 4 levels would apply to all characters in all checks, so it wouldn't serve to differentiate your "expert climber" in that particular area from everyone else.

Specializing in a skill or task should allow your character to accomplish more difficult tasks (higher DCs) than those who don't train in that area.



That said, I do find the whole roll d20s plus other dice together somewhat odd and off.
Maybe I just need to get used to it, or maybe trade that for static bonuses.



Yes, you can.


It's called a skill trick.


They just need to add a feat to choose a skill trick.    



If you want to be an exceptionally good climber, you spend a feat and get Climb Sheer Surfaces* (which would of course have a different mechanic if there were no skill dice).

Carl


*On a tangentially related note:  I think that Climb Sheer Surfaces needs to be tweaked some anyway since as it stands it doesn't make them better at climbing, merely faster.    


I personally don't believe that skill works the way you believe it does.

Take darts for example. Let's say we have 3 individuals playing darts in a bar all with equal hand eye coordination. One just started playing, one is a bar regular who plays once a week, and the other plays darts everyday. Let's say they all hit bullseyes with a roll of 17+ or a specific number (lets say 20s) on 9+
 
With my method we have untrained guy hitting bullseyes 20% of the time, trained guy hitting bulls 36% of the time, and master guy hitting bulls 49% of the time. For the specific number we have untrained guy at 60%, trained guy at 84%, and master at 94%

If we had training grant a flat +3/+5 bonus instead, trained guy hits bulls 35% and master 45% The variance for this is really wild. Instead of clustering darts around the target, the masters throws are all over the place, just like the brand new dart throwes. For hitting the 20, the trained guy hits it 75% and the master only 85%. You can really see how skill doesn't narrow the variance with only a flat bonus.

The extra d20s thus seems more realistic to me. A trained individual is more likely to repeat good results, while an untrained one can get good results but is unable to reliably repeat them.



And statistics also don't work the way you think. The low variance dice systems you promote also restrict system granularity so that you lose control over the difficulty settings. It also means that the target range of DC 's has ordinality but no cardinality, undermining the ability for most GM's or designers to craft reliable challenges. Unfortunately there are tons of dice-pool and roll-keep systems out there and they almost all get this wrong.


I personally don't believe that skill works the way you believe it does.

Take darts for example. Let's say we have 3 individuals playing darts in a bar all with equal hand eye coordination. One just started playing, one is a bar regular who plays once a week, and the other plays darts everyday. Let's say they all hit bullseyes with a roll of 17+ or a specific number (lets say 20s) on 9+
 
With my method we have untrained guy hitting bullseyes 20% of the time, trained guy hitting bulls 36% of the time, and master guy hitting bulls 49% of the time. For the specific number we have untrained guy at 60%, trained guy at 84%, and master at 94%

If we had training grant a flat +3/+5 bonus instead, trained guy hits bulls 35% and master 45% The variance for this is really wild. Instead of clustering darts around the target, the masters throws are all over the place, just like the brand new dart throwes. For hitting the 20, the trained guy hits it 75% and the master only 85%. You can really see how skill doesn't narrow the variance with only a flat bonus.

The extra d20s thus seems more realistic to me. A trained individual is more likely to repeat good results, while an untrained one can get good results but is unable to reliably repeat them.



And statistics also don't work the way you think. The low variance dice systems you promote also restrict system granularity so that you lose control over the difficulty settings. It also means that the target range of DC 's has ordinality but no cardinality, undermining the ability for most GM's or designers to craft reliable challenges. Unfortunately there are tons of dice-pool and roll-keep systems out there and they almost all get this wrong.



I actually gave a significant amount of thought to the DCs which is why I created the convenient table on page 1. I paid attention to what these statistics mean for difficulties in general and what they do to the game overall. I included the "level bonus" to increase granularity without breaking the bounds of bounded accuracy.

For my system to work all the DM needs to do is choose a DC he feels appropriate for what he feels an untrained individual might face. The trained individuals will simply have an easier time accomplishing the task. We can even give the DC breakouts fancy names if you wish to make things easier on the DM.

DC 30 Godlike Difficulty - Only the very best could hope to accomplish such a task.
DC 26 Legendary Difficulty - Even seasoned adventurers with exceptional capabilities have a tough time accomplishing such tasks.
DC 22 Very Challenging - A task too difficult for most low level or untrained adventurers accomplish.
DC 18 Challenging - An untrained individual could sometimes accomplish such a task while a trained adventurer could accomplish such tasks a decent amount of time.
DC 14 Average - An untrained individual has a decent chance of success while a trained adventurer finds these tasks fairly easy. There is still a chance for failure however.
DC 10 Easy - Most people can accomplish such tasks given time while trained individuals rarely fail such tasks.

I find multiple dice and a narrower range of skill difficulty much preferable to tasks/skill so varied that one person only succeeds on a roll of 17+ while the other only fails on a 3 or less. It is much harder to make challenges that can equally affect a party with such a discrepancy. 5e has reduced this somewhat as compared to 4e and 3e, but a rogue rolling 2d20 (choose best) + 2d12 (choose best) + 4 vs a fighter with d20 + 2 is still too much of disparity. A DC 20 task for example would be trivially easy for the rogue, but the fighter fails 85% of the time. With my suggested method, DCs would be reduced across the board. The task may still be trivially easy for the rogue rolling 3d20, but the fighter rolling his single d20 would have more than twice the chance to contribute/succeed.


I personally don't believe that skill works the way you believe it does.

Take darts for example. Let's say we have 3 individuals playing darts in a bar all with equal hand eye coordination. One just started playing, one is a bar regular who plays once a week, and the other plays darts everyday. Let's say they all hit bullseyes with a roll of 17+ or a specific number (lets say 20s) on 9+
 
With my method we have untrained guy hitting bullseyes 20% of the time, trained guy hitting bulls 36% of the time, and master guy hitting bulls 49% of the time. For the specific number we have untrained guy at 60%, trained guy at 84%, and master at 94%

If we had training grant a flat +3/+5 bonus instead, trained guy hits bulls 35% and master 45% The variance for this is really wild. Instead of clustering darts around the target, the masters throws are all over the place, just like the brand new dart throwes. For hitting the 20, the trained guy hits it 75% and the master only 85%. You can really see how skill doesn't narrow the variance with only a flat bonus.

The extra d20s thus seems more realistic to me. A trained individual is more likely to repeat good results, while an untrained one can get good results but is unable to reliably repeat them.




Yes but that only works if you restrain the DC range inside the 1-20 range of the d20.
If you had a task that would require a virtual 21+ on the dice, rolling 2, 3 or more d20s would never get you there.

So the only way to reach DCs higher than 20, would be to have high ability scores, or levels (to get those general +X ability check bonus you suggested).

Which if you think only in terms of general mechanics is fine, but levels and pure ability bonus doesn't represent specific specializations of a character. So you would be taking away the possibility of a player to say "well, I would like my character to be especially good at that, so he can beat even those high DCs others can't."

To do that the player would have to focus on an entire base Ability, and that's changing a lot more in the concept of your character than simply focusing in a skill you want to be good at.
Or he would have to wait till a higher level when others were also supposedly be able to do the same because of the general bonus... and then it's not really something special he has because he focused on that aspect of the character.



You could set, I suppose, the highest possible DC for any task to be 20.
Then if you add no modifiers whatsoever but only more/less d20s to roll and choose (from abilities, mastery, etc), all that math would be neat and flawless in terms of synergy with character-concept building.


I'm not sure I've made myself too clear.
I'm not saying your system is bad or your math wrong. I'm just saying that if you take away the possibility of a player investing in one skill to reach those harder DCs you take away quite a bit of flavour-building in character creation.
It becomes more rigid: "If you want to ever even have the possibility to do that, then you need to have a high base ability score."





Well, see? I'm sure there's plenty of bodybuilders out there who are 10x stronger than him and can't even begin to hope accomplishing that.
Cause that guy's just really good at doing what you said... ;P
So he can do what for most is impossible.

That's just physical strength, though. Granted, it's an entirely different sort of physical strength than bodybuilders have, but that's a system limitation where you only have three stats to represent the entire spectrum of physical capabilities. I really don't want to play a system where you can have different strength scores for your fingers relative to your wrists, legs, or back.


it's a feat of strength
@Lawolf: While I generally don't like non-linear CDF's for game mechanics I respect the amount of consideration you have put into this. The chart you provided gives us needed transparency and indicates acceptable ranges. I think if this were for final use you might but want a third chart of probabilities for inbetween bonuses, but I respectfully rescind my concerns because this seems like a workable system.

I'm not sure I've made myself too clear.
I'm not saying your system is bad or your math wrong. I'm just saying that if you take away the possibility of a player investing in one skill to reach those harder DCs you take away quite a bit of flavour-building in character creation.
It becomes more rigid: "If you want to ever even have the possibility to do that, then you need to have a high base ability score."



Yep I think we just have a difference of opinion on what being skilled means that might be hard to compromise on.

Lets say underwaterbasketweaving is a very hard skill for even for a person with a 20 attribute. I would rather have the DC set around 22 so even an untrained person with a 20 attribute can still sometimes succeed by sheer luck at weaving a basket whilst underwater. For me, a trained individual succeeds at his underwater basketweaving more often and is less reliant on the luck of the die.

What it sounds like you want is underwater basketweaving to be set at DC 26. This way an untrained person should never even attempt to weave a basket underwater because they will always fail. Such a task is reserved only for those highly trained individuals. That is a perfectly valid view on what skills should be like, and I respect that.

It doesn't work for me though. I don't want to have to separate the "haves" and "have nots". I think it is better when every member of the group has the opportunity to participate (if they wish). I don't like that past editions made skills reach a point where one PC needed a 17+ while another needed a 3+. I also don't like that setting DCs artificially high lowers the chance of success too much. In general I like for D&D to be about heroes overcoming obstacles, not being stopped by them.
I agree with the OP.  A single skill die does not = training.  

The min result must increase along with the max DC a character can reach.   Alternatively,  you could also increase the chance of success by using more d20 dice.     

At the moment, a master fiction writer can fail just as hard as an unskilled writer.   I find that hard to accept.    You don't roll the dice with a task you're skilled at.    Entry level tasks at some point should become trivial and be an automatic success.     


@OP

For what it's worth, I for one think your rule suggestion is a masterstroke of simplicity and elegance that is mechanicaly sound and bound to satisfy 4e newbies as well as grumpy grognards. I really hope the designers get a look at this and get the same "lightbulb" moment I just had.

Add in a bit of specific tweaking (like additional attack bonus for fighters and some sort of skill edge for rogues) and we've got something here

 
Try radiance RPG. A complete D20 game that supports fantasy and steampunk. Download the FREE PDF here: http://www.radiancerpg.com
I definitely agree that with rogues the skill die and skill mastery gets really awkward, but I like the skill dice idea.  

I'd rather see MDD/Skill dice combine to give each PC a limited dice pool that can be used on extra damage, parry/protect, skill tricks, and other manevuers based on what they can learn through training in their class (or by picking feats a-la-carte).   Give clerics and wizards ways to use the dice too so that they can experience the "wow" factor of that additional die or dice roll.  Here is where clerics can affect more undead, or add to specialized domain specific attack or defense forms, or add to their skill attempts.   Here is where wizards can add meta-magic alterations to spells (enlarge, lengthen duration, add damage, quicken spells, etc.) and add damage (rather than having spells scale), allow wizards to block magical damage or energy damage, have wizard duels with other wizards, etc.   With some effort, WotC could give every PC a chance to gain more options and more chances to affect the game using the MDD/Skill dice as an easy way to scale the effectiveness of such options.

But back to the rogue, if using a trained skill is attribute plus skill dice, then skill mastery could allow rolling the d20 at an advantage rather than the skill die at the advantage, or we could just go back to having skill mastery set a result on the die roll no less than "8" (or 7 or 10 depending on how successful we want trained rogues to be) because the roll of the skill die would make it so that there would not be auto success (at least for moderate and difficult tasks).

A Brave Knight of WTF - "Wielder of the Sword of Balance"

 

Rhenny's Blog:  http://community.wizards.com/user/1497701/blog

 

 


What it sounds like you want is underwater basketweaving to be set at DC 26. This way an untrained person should never even attempt to weave a basket underwater because they will always fail. Such a task is reserved only for those highly trained individuals. That is a perfectly valid view on what skills should be like, and I respect that.



Actually it is the current system in the current 5ed packet that goes to 26 and higher DCs. (up to 30, is it not?)
I actually proposed that to use your system the DCs for "epic" checks (such as underwater basketweaving, lol) be lowered so that the many d20 rolls would be meaniful to represent a skilled character achieving such, and the skilled character wouldn't depend so much in level/base abilities to be able to even try those.

On a side note... I had too much wine today... I may not be sounding all too clear.
;P

 For me, a trained individual succeeds at his underwater basketweaving more often and is less reliant on the luck of the die.



I do agree with that. I was pointing out that with the current DCs going well above 20 for such tasks such as underwater basketweaving (again, lol) the extra d20 could fail to represent a character's skills when it came to such tasks, and only his level and base abilities could ever lead him to that (once the skill die is removed).
If we change the DCs, as I said, then things change as a whole.

Perhaps the better way to represent that skill would be the good old fixed "+X on skill check."
It would be, as you said, less reliant on luck and would guarantee the skilled character better checks, while also allowing him to reach higher DCs.

Maybe we just need to reduce the gap between the "very very skilled" and the average roller.
Say, if skill training instead of a die that went up to d12, gave you a fixed bonus that went up to +8 or +10 tops at high levels.
That would give the very very skilled on a max +50% better chance of succeeding at any related task... meaningful, but not so much that the untrained can't ever compete.
And even the untrained if he had a good base ability should be able to mitigate that margin when compared to the skilled who doesn't have so high an ability.

Then we shouldn't have those extremes like one needing 3+ and the other 17+.

Unless you're talking about a very very skilled one with high ability, and an untrained one with a very low ability.
But then I guess it's reasonable. The "weak" player made a choice of forfeiting that aspect of his character entirely.



2) Problem with your system is you can't make a character who is "very good at climbing", for example, if skills don't raise the maximum DC of tasks you can reach.

Yeah you can make a character with high Str but that's saying "my character is Arnold Schwarzenegger" and not "he's really good at climbing."

No matter how many Advantage d20s you roll for being trained or for having skill mastery... you still can't beat a task that has a higher DC, you'll only score more high numbers in that same DC-margin you already had with 1 die.

With the system you propose, the only real way for a character to stand out in a particular task (say climbing for example) is to have a high base ability, which doesn't really represent being an expert in an area.
Even the +1 bonus every 4 levels would apply to all characters in all checks, so it wouldn't serve to differentiate your "expert climber" in that particular area from everyone else.

Specializing in a skill or task should allow your character to accomplish more difficult tasks (higher DCs) than those who don't train in that area.



That said, I do find the whole roll d20s plus other dice together somewhat odd and off.
Maybe I just need to get used to it, or maybe trade that for static bonuses.



Yes, you can.


It's called a skill trick.


They just need to add a feat to choose a skill trick.    



If you want to be an exceptionally good climber, you spend a feat and get Climb Sheer Surfaces* (which would of course have a different mechanic if there were no skill dice).

Carl


*On a tangentially related note:  I think that Climb Sheer Surfaces needs to be tweaked some anyway since as it stands it doesn't make them better at climbing, merely faster.    



Yeah, I really wish skill tricks were a little more universal, much like the skill utility powers of 4e. Warrior type classes might focus on feats of strength, rogues on feats of skill or feats of cunning, casters could even gain a few magical ones.
As you haven't responded to any questions I posed, I'm going to be more explicit in some of my points and concerns.  It's not that I see such a system as unworkable or undesireable in some ways, but I see little evidence that it actually achieves the goals you seem to be striving for.

Take darts for example. Let's say we have 3 individuals playing darts in a bar all with equal hand eye coordination. One just started playing, one is a bar regular who plays once a week, and the other plays darts everyday. Let's say they all hit bullseyes with a roll of 17+ or a specific number (lets say 20s) on 9+
 
With my method we have untrained guy hitting bullseyes 20% of the time, trained guy hitting bulls 36% of the time, and master guy hitting bulls 49% of the time. For the specific number we have untrained guy at 60%, trained guy at 84%, and master at 94%

If we had training grant a flat +3/+5 bonus instead, trained guy hits bulls 35% and master 45% The variance for this is really wild. Instead of clustering darts around the target, the masters throws are all over the place, just like the brand new dart throwes. For hitting the 20, the trained guy hits it 75% and the master only 85%. You can really see how skill doesn't narrow the variance with only a flat bonus.

The extra d20s thus seems more realistic to me. A trained individual is more likely to repeat good results, while an untrained one can get good results but is unable to reliably repeat them.


Firstly, I think the example is unreasonable, in that D&D does not generally attempt to model such situations, simply because (with some exceptions) it does not measure degrees of success on tasks.  D20 rolls are generally boolean; you either succeed or fail.  If I need a 6 to succeed for example, there is no difference to me if I roll a 6 or a 16.  In that sense you could more reasonably think of a d20 as simply having the faces labelled with "S" for success and "F" for failure on d20 checks, and you simply select the die with the correct number of each for any particular roll.  Thus, I see no particular justification for saying that the trained darts player is "all over the place," as the system is not attempting to measure that; it only determines if the player hit the desired target or not.

For the same reason, I find your claim that a 3rd d20 is "only worth about a +2 bonus" to be dubious.  Firstly, simply saying that the difference in the mean roll between 2d20 and 3d20 is about +2 does not capture the notion of boolean results, since the mean value of the die is not the relevant quality.  Indeed, across much of the spectrum the 3rd d20 decreases the probability of failure more than a +2 bonus.  Secondly, you are evaluating that +2 bonus in a particular context.  That is, if the system gives a +2 bonus for flanking, it does so whether you have a 10% chance to hit or a 90% chance to hit; considering the real impact of that +2 bonus depends on the situation.  Were we to consider the 3rd d20 roughly equivalent to a +2 bonus, you are doing so in the context of having already substantially increased your probability of success, and that changes things significantly.

Further, saying that already using a d20 pool reduces the impact of adding an extra die (such as through advantage) is not immediately justifiable, as the 2nd die has exactly the same effect as the 10th die in terms of a proportional decrease in the probability of failure, which is arguably at least as significant a metric for determining the value of a bonus as anything else.

It doesn't work for me though. I don't want to have to separate the "haves" and "have nots". I think it is better when every member of the group has the opportunity to participate (if they wish). I don't like that past editions made skills reach a point where one PC needed a 17+ while another needed a 3+. I also don't like that setting DCs artificially high lowers the chance of success too much. In general I like for D&D to be about heroes overcoming obstacles, not being stopped by them.


I don't want to have to separate "haves" and "have nots" either, but I don't see any particular reason that what you're proposing accomplishes this better than the skill die. 

It is difficult to make a direct comparison, but it seems reasonable to think of your 2nd d20 as occupying approximately the same space as the skill die in the playtest (representing a limited character investment in terms of selecting a skill).  The 3rd d20 is a bit trickier as you're talking about greater character investment, and if there is a problem in the playtest here then it may have more to do with the rogue's skill mastery or the feats than the skill die mechanic.  This is further complicated by the fact you are talking about two distinct issues; one way to mirror your first suggestion within the context of skill dice is to have everyone add skill dice to all rolls all the time, and allow investment in skills in some other fashion.

However, there is no particular reason to compare the 2nd d20 to the effect of a d12 skill die, since that only occurs at the highest levels.  It is at least as relevant to compare the 2nd d20 to a d4 skill die, and in that sense you are actually creating a greater distinction between haves and have-nots under many circumstances. 

You could cherry-pick required rolls of 20 and higher to say the skill die is better, but why?  Only under extraordinary circumstances do such rolls become routine; even with d20 + highest of 2d12 you still fail more often than not when you need to roll 20 or higher, and placing characters in front of challenges in which they fail more often than not is hardly the norm.  The skill die system allows you to give characters tasks in which they require rolls higher than 20, but it doesn't restrict you from keeping DCs in a place where all characters have some chance of success if you wish, and the game tends to focus on those anyway.  The 2nd d20 is roughly equivalent to a d9 in terms of a boolean evaluation of success/failure across much of that spectrum, and so in that context you are not lessening the disparity between characters throughout the lower levels; you are increasing it.

So, as I said at the beginning of the post, I don't see how the system you're proposing really accomplishes what you are suggesting it does better than what we have now.
@Pyromantic: With a skill die the probability curve remains fairly similar to a flat bonus (except at the edges). This causes issues where even trained individuals still have a decent chance to fail average mundane tasks. At low levels the skill die is about as good as a flat +3 or so but it continues to grow to the point where those with and without skill shouldn't both attempt the same tasks. I also stated that I do not like rolling d4s, d6s, etc with my d20 as a preference. I just like they way 2+ d20s feel.

I think skill should decrease the chance of failure significantly, especially for easier tasks. This is something the skill die doesn't accomplish well. I also think that being skilled should not necessarily allow you do do things that an unskilled person couldn't already accomplish through luck. This way, everyone at the table is given the chance to contribute to any given conflict. By using multiple d20s, you can reduce the DCs across the board but still have a similar chance for success for a given task difficulty (see the table on the 1st page).

The skill die still has enough variance that being skilled just doesn't feel right to me. Rolling 2d20 and taking the better result is a significant benefit, but it doesn't increase the maximum die result, it only pushes the average roll up. There is still a chance for failure, but it is lessened. It also means that other party members, even unskilled ones can roll to accomplish tasks with you.

Take for example a dc 11 task under my method. Untrained guy has a 50% chance to succeed, trained guy has 75%. The DC for this task is bumped up to 14 in the current method, trained guy has a 50% chance while untrained guy has a 35% chance. (Remember: I lowered DCs across the board to account for the decrease in maximum value due to no skill die).

Also, the skill die is a slowdown. Roll 1 (or 2) d20, roll 1 (or 2) skill dice. Add the results. It is much easier to choose the highest from 1-3 d20s.

Now for the 2d20 => 3d20 is worth about +2. Just like 1d20 => 2d20 is worth around +5 (it increases the average result by about that much) going 2d20 to 3 is worth around +2. As such, if most people rolled 2d20 (for tasks they have proficiency with) going to 3d20 actually becomes a less significant jump.

This means that if a warrior rolls 2d20 normally and gains advantage for 3d20, advantage becomes worth slightly less. This means we can instead replace +1 and +2 bonuses entirely. If advantage has a reduced value, we can apply advantage more liberally. Flanking, ganging up, attacking prone enemies, etc could all simply grant advantage. The game would benefit from removing these micro-conditional bonuses that have been shown to cause clutter and slowdown. The rolling multiple d20s for attacks is more of a side thought though as the primary focus here is upon skills.
D20 rolls are generally boolean; you either succeed or fail.  If I need a 6 to succeed for example, there is no difference to me if I roll a 6 or a 16.

Is it unusual in any way to say that a higher check result is better than a low check result? I could have sworn that I'd seen exactly that line in at least one previous edition, but it was also the way d20 rolls have been explained ever since I started with AD&D. Maybe it is without direct support to say that a result of 16 on a check is better than a result of 6, when you only needed a 5 - I am away from my books at the moment - but I have never encountered a group which varied from that interpetation.

The metagame is not the game.

I really like the Advantage and Advantage+ for skill training. If people want their characters to be capable of extrodinairy feats then I think they should look to .. Feats.

I'm not a fan of an across the board bonus for advancement. I think there should be a few things that each character should never get better at. My stereotypical half-orc barbarian should never get better at socializing with high society, no matter his level and experience.
I really like the Advantage and Advantage+ for skill training. If people want their characters to be capable of extrodinairy feats then I think they should look to .. Feats.



I agree, I wish feats could represent this better. I would also like if feats could grant certain "skill tricks" or "feats of skill".

I'm not a fan of an across the board bonus for advancement. I think there should be a few things that each character should never get better at. My stereotypical half-orc barbarian should never get better at socializing with high society, no matter his level and experience.



It isn't necessarily that you barbarian is better at socializing, it is that high society is more likely to favor a "rock star" who acts brash/rude/drunk than the town hobo who acts brash/rude/drunk.  Afterall, by level 20 you are literally fighting gods. You should have a reputation that spans many kingdoms. You have also probably spent countless hours at these boring courtly dinners. Even if you don't necessarily train for it, you still pick up a thing here and there and your reputation does the rest. Worst comes to worst, just don't roll that diplomacy check and you will never notice that you are a little better at diplomacy at level 20 than you were at level 1.
The rockstar scenario is really setting based and shouldn't have anything to do with level and experience. A level 5 character could be a local hero and get better treatment then the visiting level 20 character from a far off land or plane with the exact same social capabilities.

And I'm not a fan of the "ignore it" idea as the rules should really represent my character's reality so I shouldn't need to ingnore something written right on the character sheet in order to play the character.

I think characters should see advancement across levels in many facets, just not universally.
I'm not going to comment on the math stuff for now, since I'm in kind of a hurry.

However, I oppose flat "across the board" bonuses. It detracts from the usefulness of focusing on a particular aspect of your character with actual resources, such as feats. A character -should- get better over time, but flat inflation strikes me as completely unnecessary, especially if you account for it on the other side of the screen. You start with good ability scores, honed class abilities, and features; you gain feats and improve features over time. Shouldn't you get better at things you actually care about, rather than just absorbing general all-purpose plotpower? xD
I'm not going to comment on the math stuff for now, since I'm in kind of a hurry.

However, I oppose flat "across the board" bonuses. It detracts from the usefulness of focusing on a particular aspect of your character with actual resources, such as feats. A character -should- get better over time, but flat inflation strikes me as completely unnecessary, especially if you account for it on the other side of the screen. You start with good ability scores, honed class abilities, and features; you gain feats and improve features over time. Shouldn't you get better at things you actually care about, rather than just absorbing general all-purpose plotpower? xD



I'm going to address this aspect right here. My opinion may differ from others so don't take offense to this. This is not how I see PCs. In books, movies, comics, etc the protagonist improves overtime. Yes the specifics they actually train in improve more pronouncedly, but with a lot of time and repetition they improve everywhere.

PCs are different than NPCs because they level. Leveling represents your skill and experience over time. Adventurers slowly improve overtime is just how I see things in fantasy fiction literature, media, etc. I want PCs to reflect this. In movies "high level" protagonists are almost always represented as more perceptive, more cunning, more skilled, more persuasive, and so on even though they are not always portrained as stronger, more intelligent, etc. Level directly translates into knowing how to use your latent ability.

This is very different than being highly skilled. Remember that I have a robust skill system given with training = 2d20, mastery = 3d20. So being skilled has a much higher impact than being high level for the majority of a PCs career. Remember at +1 per 4 levels even a level 10 PC on has a +2. The growth of the static "experience" bonus is very slow. The level bonus only becomes highly visible at level 20 when it reaches +5.

Also, the slow static increase in no way hinders specialization. Skill training, skill mastery, skill tricks, (and potentially grandmaster skill training), all allow for hyperspecialization and customization of your PC.

The level bonus basically is only there to show that yeah, after 20 years of adventuring, Joe the Adventurer is better at all things adventuring than some n00b, even if Joe never received formal training. Joe has spent his whole life in the dungeon, he can sneak better, climb better, search better, avoid traps better, fight better, survive better, and so on. If Joe wanted to further specialize he could easily go train some of his skills or even spend one of his 4! feats to become a master at them. Remember though skills are entirely optional and fighters only get a few of them so it is almost impossible to show Joe's years of adventuring providing some additional benefit over that of a 1st level adventurer in 5e.

Is it perfect...no.  Some things may not make sense (to you) such as a level 20 warrior being better at religion or diplomacy (despite uncovering 100s of religious artifacts and negotiating with countless dignitaries). But I feel that it makes more sense for Joe to generally be a better adventurer after 20 years of experience than for his strength to miraculously double, or his IQ to go up 40 points, or to have 20 years of adventuring leave Joe just as good at level 20 as he was at level 1. 
I feel that it makes more sense for Joe to generally be a better adventurer after 20 years of experience than for his strength to miraculously double, or his IQ to go up 40 points, or to have 20 years of adventuring leave Joe just as good at level 20 as he was at level 1. 



I mostly agree, and I think I'd be completely on-board if it was more like "At 4th level and every 4 levels thereafter, an adventurer increase his attribute bonus in 2 or 3 attributes of his choice."
I feel that it makes more sense for Joe to generally be a better adventurer after 20 years of experience than for his strength to miraculously double, or his IQ to go up 40 points, or to have 20 years of adventuring leave Joe just as good at level 20 as he was at level 1. 



I mostly agree, and I think I'd be completely on-board if it was more like "At 4th level and every 4 levels thereafter, an adventurer increase his attribute bonus in 2 or 3 attributes of his choice."



I thought about something like that, but thought it seemed much more clean and streamlined to simply say you gain +1 to all ability checks at 4th level and every 4 levels thereafter. This has the added benefit of improving your attack rolls (useful for MCd PCs with two primary attack attributes or even mundane PCs who use melee and ranged weaponry), saving throws across the board, and general ability checks. It does a much better job of showing the high level PC is generally more competent than the low level PC even in a game without feats and skills.
@Pyromantic: With a skill die the probability curve remains fairly similar to a flat bonus (except at the edges). This causes issues where even trained individuals still have a decent chance to fail average mundane tasks. At low levels the skill die is about as good as a flat +3 or so but it continues to grow to the point where those with and without skill shouldn't both attempt the same tasks. I also stated that I do not like rolling d4s, d6s, etc with my d20 as a preference. I just like they way 2+ d20s feel.


I'm aware of what the distribution looks like.  I think we need to be careful just moving things to the "edges" as if it's a corner case, since the size of these edges depends entirely on which skill die you're talking about.  Further, since it isn't until you get to a d10 skill die that it typically outweighs the benefit of a 2nd d20 for tasks that you will succeed on with high frequency, I'm not sure why you feel that your system is reducing the disparity between those that should or shouldn't attempt the same tasks.  It may be eliminating the gap between those can possibly succeed at a task and those that can't, but those are hardly the same thing.  Place characters in front of challenges that they fail more often than not and I think the game will break down very quickly.

I think skill should decrease the chance of failure significantly, especially for easier tasks. This is something the skill die doesn't accomplish well. I also think that being skilled should not necessarily allow you do do things that an unskilled person couldn't already accomplish through luck. This way, everyone at the table is given the chance to contribute to any given conflict. By using multiple d20s, you can reduce the DCs across the board but still have a similar chance for success for a given task difficulty (see the table on the 1st page).


This seems to me to be at odds with your first suggestion, in which you provide a +5 bonus to high-level characters for "increased capability through skill and experience," as it absolutely allows you to do things that you couldn't already accomplish through luck. 

I think, however, that I am starting to get a better sense of what you're trying to do.  I think that you are not only looking at 2d20 take highest as the "norm" for checks, but adjusting DCs to compensate so that probabilities of success are similar to what they were, which wasn't clear to me before.  If that's correct then I think you are not so much looking at lowering DCs as narrowing the range of DCs.  If that's the case then I need to evaluate something slightly differently...  
 
Now for the 2d20 => 3d20 is worth about +2. Just like 1d20 => 2d20 is worth around +5 (it increases the average result by about that much) going 2d20 to 3 is worth around +2. As such, if most people rolled 2d20 (for tasks they have proficiency with) going to 3d20 actually becomes a less significant jump.

This means that if a warrior rolls 2d20 normally and gains advantage for 3d20, advantage becomes worth slightly less. This means we can instead replace +1 and +2 bonuses entirely. If advantage has a reduced value, we can apply advantage more liberally. Flanking, ganging up, attacking prone enemies, etc could all simply grant advantage. The game would benefit from removing these micro-conditional bonuses that have been shown to cause clutter and slowdown. The rolling multiple d20s for attacks is more of a side thought though as the primary focus here is upon skills.


I was considering the case of an additional d20 for a particular DC, and in that sense I stand by the statement that there is no particular reason a 3rd die is any less significant than the 2nd; it still acts as an exponent on the probability of failure.  However, if we are speaking in the context of increasing the probability of failure for a single die in compensation for making a 2nd die the norm, then from a certain point of view the 3rd die becomes less significant.  Assuming the numbers are balanced to give you a chance to hit between 60% and 80%, a 3rd d20 is still better than a +2 bonus though.  This means that you might be able to be slightly more lenient in moving bonuses to additional dice, but it certainly doesn't negate the need for being careful.
 
Are you also talking about a substantial change to attribute bonuses?  If not, then you are further widening the gap between haves and have-nots.  The difference between an attribute bonus of +5 and +0 is greater under your proposal than the playtest.  For example, with a single d20 a DC 11 is at 50% for a person with +0 and 75% for a person with +5.  A close analog would be DC 15 under your proposal, which with 2d20 take highest is at 51% for a person with +0 and 79.75% with +5.  You can add the consideration of skill dice and find your proposal increases the potency of flat bonuses.  This is part of the issue for me; I don't see any consideration of context when you describe certain bonuses as relatively minor, which I think is entirely necessary given that not only does the in-game context of a bonus affect its significance, but you are also changing the nature of bonuses across the board.

Maybe I'm just not getting what you're trying to do, but it still seems to me that some of your goals contradict one another.
D20 rolls are generally boolean; you either succeed or fail.  If I need a 6 to succeed for example, there is no difference to me if I roll a 6 or a 16.

Is it unusual in any way to say that a higher check result is better than a low check result? I could have sworn that I'd seen exactly that line in at least one previous edition, but it was also the way d20 rolls have been explained ever since I started with AD&D. Maybe it is without direct support to say that a result of 16 on a check is better than a result of 6, when you only needed a 5 - I am away from my books at the moment - but I have never encountered a group which varied from that interpetation.


A higher number might feel better from a subjective point of view, and some DMs may interpret it as better as a result.  There are some examples of it being true.  Knowledge checks in 4E give you more information if you beat a hard DC than a moderate DC for example, and a cllimb check failed by 5 or more has a different effect than failing by 4 or less.  Diplomacy in 3.5 allows you to change the attitude of NPCs further with a higher check result.  Even in such cases though you see broad bands of effect rather than some kind of granular movement, and it's certainly not the norm the way it is in, say, Alternity (which broke down check results into Critical Failure, Failure, Ordinary Success, Good Success and Amazing Success).

The 3.5 PHb says about skill checks that "Either you're trying to match or exceed a certain Difficulty Class (DC), or you're trying to beat another character's check result" (61).  Similarly, the 4E Rules Compendium states "Typically, a creature either succeeds or fails at a skill check, meaning that the check result meets or exceeds (beats) the DC or else falls below it" (125).

The game isn't entirely consistent in this regard, but you could look at 3.5 balance and climb for example, for which you can accept a -5 penalty to increase your movement speed, which is quite a different idea than being given the option to increase your speed if you succeed by 5 or more.  The prototypical d20 roll, the attack roll, is pretty much success or failure.  How much you exceed the target's AC doesn't determine how solid a hit was landed; the damage roll does.  Even if you want to put critical hits into the mix, it has little (or sometimes effectively nothing) to do with your actual skill, and IIRC has been considered an optional rule on and off throughout D&D's history.
Are you saying you're in favour of this suggestion or something similar, because it allows you to avoid the hit point inflation necessary to achieve level disparities similar to what occurs with this suggestion in place?

If so, then what makes you think the devs want to achieve those disparities? 

The disparity already exists, at least as of the latest packet. Assuming that the designers actually want that much difference between high levels and low levels, then I'm in favor of small +hit/AC bonuses rather than placing the burden entirely on HP/damage.

If the designers don't want that disparity between high levels and low levels, then I'm still in favor of putting the difference on +hit/AC instead of HP/damage, but at that point it's just personal preference.


I hadn't responded to this yet in part because the math starts to get very tricky and I'm only just getting far enough in my analysis to begin testing this, but I put together a heavily simplified situation for some calculations:

Single player faces off against single monster.  Player gets initiative, and attacks at 65% hit rate for d10+4 damage.  Monster attacks at 50% hit rate for d8+2 damage.  Both have 25 hit points.

Probability monster wins: 10.4%

Probability monster wins if player is given 75% chance to hit: 5.4%

Probability monster wins if player is given 75% chance to hit and monster reduced to 40% chance to hit: 2.9%

For comparison's sake increasing the player's hit points to 34 gives the monster about the same chance to win as it has in the last case: around 2.8%. 

It gets pretty complicated here so it would take quite a while to sort through the data enough to get a good sense of things, but I think it's fair to say that in general small changes in hit rates in either direction have a much more significant impact on the probability of winning than most people think.  Then again, so do small changes in relative hit points and damage.  So, it isn't clear to me at the moment what effect relative level differences have on survivability in the last playtest, let alone whether it's what the devs are striving for.  Given their stated goal of keeping monsters relevant across more levels however, I think it unlikely that they are trying to bloat hit points and damage to maintain the status quo of such differences.