D&D 5E Skills - Does anyone actually like the way they're headed?

[Mistwell]

A 1 in 5 to an almost 1 in 3 chance of failure isn't significant?

That's an interesting outlook. Can't say I share it.

Try it some time.

 

[KidSnide]

When a player wants to take an action, it’s often appropriate to just let the action succeed. A character doesn’t normally need to make a Dexterity check to walk across an empty room, or a Charisma check to order a mug of ale in a tavern. Only call for a roll if you think it’s worth taking the time for the rules to come into the flow of the game. Ask yourself two questions to aid your decision. Is the action being taken so easy, so free of stress or conflict, or so appropriate to the situation that there should be no chance of failure? “So easy” should take into account the ability score associated with the intended action. It’s easy for someone with a Strength score of 18 to flip over a table, though not easy for someone with a Strength score of 9. Is the action being taken so inappropriate or impossible that it would never work? Hitting the moon with an arrow is, for instance, impossible in almost any circumstance. If the answer to both of these questions is no, some kind of roll is appropriate.

The DM Guidelines speak of setting a task DC based on the difficulty as compared to the abilities of the character. Flipping a large table over is a hard task for a weak wizard who hasn't done much physical activity, but it should be routine for the hulking barbarian who's used to brawling in taverns every other day. The DM is within his rights to set the DC at 15 for the wizard (who'd be hard pressed to succeed with his +0 modifier) and at 10 for the barbarian (who will likely flip the table over with his +5 modifier). Still, the DCs allow the wizard some chance at succeeding, and still gives a chance of failure to the barbarian.

The DM Guidelines actually do not state this. D&D has never had this philosophy. Let's say I have a tree that requires a DC15 skill check to climb. That tree in the world has a set DC, regardless of who you are. Every published adventure, furthermore, uses this method of set DCs. That is what skills are for; you get better at a skill to increase your chance of success. The DM is not supposed to assign different DCs to various characters for the same task. I'm not saying that such a system is bad or wrong, it's just not D&D.

I think Klaus and Repsesper are both correct in spirit, and this highlights the inconsistencies with the current skill rules. The DM guidelines clearly suggest that a roll on a given challenge could be appropriate for some characters and yet not for others. At the same time, the DM guidelines don't explicitly suggest putting different DCs on the same roll.

What this suggests is a concept where skill rolls are only used when there is doubt as to the outcome for particular characters. For example, a DM might decide that the meaning of magical glyph would be obvious to any arcane casters with an Int of 16 or higher and impossible for characters with an Int below 12 or with zero experience with magic (e.g. barbarians, low-level fighters, etc). For anyone else, they roll the dice. I don't have a problem with using a lot of DM judgment in skill checks (others might call it "DM fiat"). But it suggests a method of handling skills that is really different from how they are usually written - i.e. with fixed DCs that any PC can attempt.

I think the idea is that different groups are supposed to look at the rules and get different results. Groups that hate DM judgment calls just use the DCs at written. They get predictable results (i.e. the DCs mean what they say), but get abysmal "reality simulation" because the results are vastly more random than realistic. Folks who care more about probabilistic verisimilitude need to use a lot of DM judgment to push the results closer to plausible. D&DN basically takes the attitude that "reality simulation" isn't really a goal and is more of about subjective DM/group preference than it is something that a rules system can plausible get good enough with an acceptable level of complexity.

The other thing to keep in mind is that these skill rules are written to support adventures played by different groups of PCs. One of the 3.x complaints that D&DN is trying to address (as did 4e) is the issue that - in most parties - specialists automatically succeed and non-specialists had no chance of success. Personally, I think that is highly realistic. Most interesting tasks (a long jump, an appendectomy) are the types of things that specialists will almost always succeed at and non-specialists will almost always fail. Or, to be a little more accurate, there are very few tasks that both specialists and non-specialists have both a significant chance of success and failure. If the tasks is of a level of difficulty that it is uncertain for one, it will be guaranteed success or failure for the other. If a gap is small enough that I may or may not jump over it, a brilliant athlete is sure to succeed. If a surgery is hard enough that a professional surgeon might fail, I have no chance of success.

That's life, but it's not super interesting for a game. So, once you decide that "skill realism is a problem", you're going to get a system that produces very unrealistic results.

-KS

 

[Esper the Bard]

I think Klaus and Repsesper are both correct in spirit, and this highlights the inconsistencies with the current skill rules. The DM guidelines clearly suggest that a roll on a given challenge could be appropriate for some characters and yet not for others. At the same time, the DM guidelines don't explicitly suggest putting different DCs on the same roll.

What this suggests is a concept where skill rolls are only used when there is doubt as to the outcome for particular characters. For example, a DM might decide that the meaning of magical glyph would be obvious to any arcane casters with an Int of 16 or higher and impossible for characters with an Int below 12 or with zero experience with magic (e.g. barbarians, low-level fighters, etc). For anyone else, they roll the dice. I don't have a problem with using a lot of DM judgment in skill checks (others might call it "DM fiat"). But it suggests a method of handling skills that is really different from how they are usually written - i.e. with fixed DCs that any PC can attempt.

I think the idea is that different groups are supposed to look at the rules and get different results. Groups that hate DM judgment calls just use the DCs at written. They get predictable results (i.e. the DCs mean what they say), but get abysmal "reality simulation" because the results are vastly more random than realistic. Folks who care more about probabilistic verisimilitude need to use a lot of DM judgment to push the results closer to plausible. D&DN basically takes the attitude that "reality simulation" isn't really a goal and is more of about subjective DM/group preference than it is something that a rules system can plausible get good enough with an acceptable level of complexity.

The other thing to keep in mind is that these skill rules are written to support adventures played by different groups of PCs. One of the 3.x complaints that D&DN is trying to address (as did 4e) is the issue that - in most parties - specialists automatically succeed and non-specialists had no chance of success. Personally, I think that is highly realistic. Most interesting tasks (a long jump, an appendectomy) are the types of things that specialists will almost always succeed at and non-specialists will almost always fail. Or, to be a little more accurate, there are very few tasks that both specialists and non-specialists have both a significant chance of success and failure. If the tasks is of a level of difficulty that it is uncertain for one, it will be guaranteed success or failure for the other. If a gap is small enough that I may or may not jump over it, a brilliant athlete is sure to succeed. If a surgery is hard enough that a professional surgeon might fail, I have no chance of success.

That's life, but it's not super interesting for a game. So, once you decide that "skill realism is a problem", you're going to get a system that produces very unrealistic results.

-KS

Exactly. Your explanations make total sense.

It would seem that a game system would have to favor either realism or involvement for everyone. Or would it? Speaking in D&D Next terms, check this out:

Version:1.0 StartHTML:0000000167 EndHTML:0000004649 StartFragment:0000000457 EndFragment:0000004633 Level Proficiency Bonus
1 +4
3 +5
7 +6
11 +7
15 +8
19 +9


Example DCs


-Arcana-
Identify a relatively common elemental creature or magical beast: DC 10
Understand the basic nature of a spell, magic effect, or magic item: DC 15
Esoteric lore: DC 20
Read the runes in the ancient lich's spellbook: DC 25


-Athletics-
Climb a rope or a rugged, rocky incline: DC 10
Climb a tree: DC 15
Climb a stone wall: DC 20
Climb a surface with very limited handholds/footholds: DC 25


-Disable Device-
Disarm a simple trap (tripwire): DC 10
Disarm a standard trap (pressure plate): DC 15
Disarm a complex trap (complex mechanism): DC 20
Disarm a magic trap: DC 20 + spell level

-Medicine/Treat Wound-
Bandage wounds: DC 10
Stitch a deep cut: DC 15
Stabilize a dying character: DC 15
Treat a common disease: DC 15
Treat a rare disease: DC 20
Perform complicated surgery: DC 25



You can see that this sort of system, along with a +3 or +4 ability score, makes you feel trained. You are clearly better than other untrained characters, though you can still be challenged and those untrained still have a slim chance.

Furthermore, if a trained character is being rushed or threatened, he could take 10 on skill checks to grant automatic success.

 

[Warbringer]

"I sneak past the guard, climb up the wall, and then listen at the window."

That's three skill rolls, and they are linked, as you obviously cannot attempt the next if you failed the previous.

Let's look at the probabilities, making all those three rolls equal for simplicity's sake:

* If you assign each skill roll a 50% chance of success, the entire plan has only a 0.5^3 chance of success, which is 12.5%...
* If you assign them 70% chance, the entire plan stands at 0.7^3, which is 34% - still a lousy shot.
* If you assign them 90% chance, the entire plan ends up at 72%, where you may start to consider doing it...

This is why I prefer "skill challenges" (or at least the way I run them). This to me is a single objective that a rogue/thief would do on a daily basis. Failure doesn't end the attempt, but creates consequences in the chain or (preferably) a consequence to the overall success.

Failure here could be the guard hears the thief and gets a message to whomever is in the room, allowing them to pass false information, jump the thief when most vulnerable; or during the climb made noise; at the window is spotted; leaves something behind that allows him to be identified .... etc

 

[Esper the Bard]

I think Klaus and Repsesper are both correct in spirit, and this highlights the inconsistencies with the current skill rules. The DM guidelines clearly suggest that a roll on a given challenge could be appropriate for some characters and yet not for others. At the same time, the DM guidelines don't explicitly suggest putting different DCs on the same roll.

What this suggests is a concept where skill rolls are only used when there is doubt as to the outcome for particular characters. For example, a DM might decide that the meaning of magical glyph would be obvious to any arcane casters with an Int of 16 or higher and impossible for characters with an Int below 12 or with zero experience with magic (e.g. barbarians, low-level fighters, etc). For anyone else, they roll the dice. I don't have a problem with using a lot of DM judgment in skill checks (others might call it "DM fiat"). But it suggests a method of handling skills that is really different from how they are usually written - i.e. with fixed DCs that any PC can attempt.

I think the idea is that different groups are supposed to look at the rules and get different results. Groups that hate DM judgment calls just use the DCs at written. They get predictable results (i.e. the DCs mean what they say), but get abysmal "reality simulation" because the results are vastly more random than realistic. Folks who care more about probabilistic verisimilitude need to use a lot of DM judgment to push the results closer to plausible. D&DN basically takes the attitude that "reality simulation" isn't really a goal and is more of about subjective DM/group preference than it is something that a rules system can plausible get good enough with an acceptable level of complexity.

The other thing to keep in mind is that these skill rules are written to support adventures played by different groups of PCs. One of the 3.x complaints that D&DN is trying to address (as did 4e) is the issue that - in most parties - specialists automatically succeed and non-specialists had no chance of success. Personally, I think that is highly realistic. Most interesting tasks (a long jump, an appendectomy) are the types of things that specialists will almost always succeed at and non-specialists will almost always fail. Or, to be a little more accurate, there are very few tasks that both specialists and non-specialists have both a significant chance of success and failure. If the tasks is of a level of difficulty that it is uncertain for one, it will be guaranteed success or failure for the other. If a gap is small enough that I may or may not jump over it, a brilliant athlete is sure to succeed. If a surgery is hard enough that a professional surgeon might fail, I have no chance of success.

That's life, but it's not super interesting for a game. So, once you decide that "skill realism is a problem", you're going to get a system that produces very unrealistic results.

-KS

Exactly. Your explanations make total sense.

It would seem that a game system would have to favor either realism or involvement for everyone. Or would it? Speaking in D&D Next terms, check this out:

Level
1 (proficiency bonus +4)
3 (proficiency bonus +5)
7 (proficiency bonus +6)
11 (proficiency bonus +7)
15 (proficiency bonus +8)
19 (proficiency bonus +9)


Example DCs


-Arcana-
Identify a relatively common elemental creature or magical beast: DC 10
Understand the basic nature of a spell, magic effect, or magic item: DC 15
Esoteric lore: DC 20
Read the runes in the ancient lich's spellbook: DC 25


-Athletics-
Climb a rope or a rugged, rocky incline: DC 10
Climb a tree: DC 15
Climb a stone wall: DC 20
Climb a surface with very limited handholds/footholds: DC 25


-Disable Device-
Disarm a simple trap (tripwire): DC 10
Disarm a standard trap (pressure plate): DC 15
Disarm a complex trap (complex mechanism): DC 20
Disarm a magic trap: DC 20 + spell level

-Medicine/Treat Wound-
Bandage wounds: DC 10
Stitch a deep cut: DC 15
Stabilize a dying character: DC 15
Treat a disease: DC 20
Treat a rare, virulent disease: DC 25

Perform complicated surgery: DC 25



You can see that this sort of system, along with a +3 or +4 ability score, makes you feel trained. You are clearly better than other untrained characters, though you can still be challenged, and those untrained still have a slim chance.

Furthermore, as long as a trained character is not being rushed or threatened, he could take 10 on skill checks to grant automatic success.

 

[KidSnide]

You can see that this sort of system, along with a +3 or +4 ability score, makes you feel trained. You are clearly better than other untrained characters, though you can still be challenged and those untrained still have a slim chance.

I agree that the existing proficiency bonus doesn't do enough, but I think increasing it by a small amount only helps so much. The objective is to prevent silly disparities. I wonder if a better idea is to grant proficient characters an extra d20 (like advantage, but ideally stacking) on proficient skill checks? That will dramatically reduce the number of failures on easy checks.

Also, while we're talking about skills, three other points:

1) Ability checks are also a problem.

2) Ability contests are worse. Seriously, I have trouble thinking of a single example where two characters are matching a abilities in which the existing ability contest produces a plausibly realistic result. It's not just that there are corner cases. It's wrong almost all the time.

3) Expertise is too narrowly focused on a small number of classes. As it is now, the rogue 12 / wizard 1 is probably better at arcana than a wizard 20.

-KS

 

[MarkB]

Well, we already assume a success rate in attacks between 50%-75%, and skills are mostly in the same ballpark.

The DM Guidelines speak of setting a task DC based on the difficulty as compared to the abilities of the character. Flipping a large table over is a hard task for a weak wizard who hasn't done much physical activity, but it should be routine for the hulking barbarian who's used to brawling in taverns every other day. The DM is within his rights to set the DC at 15 for the wizard (who'd be hard pressed to succeed with his +0 modifier) and at 10 for the barbarian (who will likely flip the table over with his +5 modifier). Still, the DCs allow the wizard some chance at succeeding, and still gives a chance of failure to the barbarian.

I really dislike this concept. Surely the whole point of having a skill system which incorporates modifiers for both ability scores and training is that it takes into account differences in both basic aptitude and proficiency. The DM shouldn't be having to change the DC based upon which character is making the check - the differences between different characters' chances of success at a task should be inherent in their skill check modifier. That's the whole point in having those modifiers.

If those modifiers aren't doing that job properly, then the skill system is flawed. The solution should be to overhaul it, not to patch it over by modifying skill check difficulties on-the-fly to match perceptions of how easy it 'should' be for one character compared to another.

 

[lkj]

I agree, it is very easy. But, whether you had +1 or +5 at level 1, it's still just as easy. I am not arguing that it should be less easy, just that training should actually feel like training. This year, I practiced rock climbing, guitar playing, singing, and Spanish. I am definitely way more than 10% better at these skills now, and someone who has never trained in these skills is way more than 25% less capable than me. Again, none of this is complicating the system. It's just changing what the numbers are by a few points in order to evoke something that doesn't jar the sense of immersion.

Customizable? There really isn't any customization. You either are proficient or you are not. Unless, you're referring to the customization of selecting the skills in which you're proficient, which would have nothing to do with my post (as I never argued against skill selection/amount of skills known).

Been awhile since you posted this, but I don't check that often. In answer-- I shouldn't have used the word 'customizable'. What I meant was that the mechanic is very easy to use for creating a custom skill system. For example, I'm more of a mind to have skills be an 'area of expertise' rather than a set list. I find that set lists of skills often fail when trying to create a character concept. I'd prefer something along the lines of 'Sailor'-- and the given character is proficient at anything associated with sailing. So, for me to create that system under the current mechanic, I simply let the player apply the proficiency bonus whenever they are performing a task related to sailing or whatever. I can completely ignore the current list of skills without worrying about balance issues (beyond the increased need for increased DM adjudication which is not an issue for me and my group). So, it's easy to use for customization in that sense.

As for whether the bonus is enough-- well given that it increases with level it seems to me that the increases are quite reasonable. By high level, the +5 or +6 bonus seems plenty substantial given the DC set. I'll have a stronger feel for it after having played through enough levels.

Now, I will say it does assume that your base proficiency is low when you start adventuring and it also assumes that you continue to increase at a given skill with level. I've found that simplification works fine for me. But I can see circumstances and campaigns where it wouldn't fly. However, if the math is worked out for me already-- and it appears to be-- then adjusting that for a given game seems pretty easy. We know normal levels of skill progression ranges from +1 to +6 and that true expertise (ala the rogue) ranges between +6 and +11. I can easily imagine granting each character (in a particular campaign) high end proficiency in a given area-- sailing is a great example. Character could start the game with a high end bonus. I can also imagine some proficiencies never improving. I can easily imagine a 3e style system where you 'spend' a pool of ranks on skills you want to improve and let others stagnate. I think that's cool, but more complicated than I want for the base game. I think it would be pretty easy to include as a 'module' however.

Anyway, that's my more complete ramble. Hope it's useful.

AD

 

[Sadrik]

Scaling bonus for skills is wrong. I much preferred the 4e +5, and heck make and expert +10.

So,
Skill training +5
Skill mastery +5 more...

+5 and a stat of +3 = +8

DCs range from 10 to 35

Clearly the DCs need repair. They need repair in the current system too.

If they were better accommodated for low end bonuses. Perhaps range from 5 to 25, things would be more manageable. These are questions that appear like they have not been asked... Or perhaps were only superficially looked at... This is the core of the game however...

 

[Mistwell]

Scaling bonus for skills is wrong. I much preferred the 4e +5, and heck make and expert +10.

So,
Skill training +5
Skill mastery +5 more...

+5 and a stat of +3 = +8

DCs range from 10 to 35

Clearly the DCs need repair. They need repair in the current system too.

If they were better accommodated for low end bonuses. Perhaps range from 5 to 25, things would be more manageable. These are questions that appear like they have not been asked... Or perhaps were only superficially looked at... This is the core of the game however...

Except, it seems like the majority of people are not reporting this problem in their playtest responses. When you actually have played the playtest, are you finding in-play that the DCs are broken and the bonuses are not working? What have you found to be the results in your games?