Are Martials really that screwed? V2

[Xervous]

Previous thread https://forums.giantitp.com/showthread.php?613670-Are-Martials-really-that-screwed


In your experience, at the table, without signifigant homebrewing or buffing, have you ever felt that martials in 5e are actually suffering in any way?


Simple question to help solving the dilemma,
Would you go adventuring without a martial character, or without a full caster?


Is the skill system, when arbitrated by the GM as per the guidance set forth by the system, robust enough to allow Martials means to exercise comparable narrative authority to casters in the non combat pillars? (in absence of corrective bias assisting Martials)

To answer the matter of an absence of Martials or Casters I would have to say it depends on the GM. Underneath a gm who is heavily tailoring encounters to the party I would feel no need for Martials or Casters. However for a gm that is not tailoring things to the party, modules being a grand example, I would insist on having a caster and think nothing of lacking a martial. Narrative overrides are something martials generally lack, especially outside of combat. A reminder here that I am defining narrative override as an ability that allows a character to exert elevated influence on a scene or problem to a magnitude such that it likely becomes a foregone conclusion, or allows for additional avenues for approaching a dilemma. Casters provide more options, martials generally just provide more numbers. I’d rather fail due to slim margins on numbers, poor dice rolls and in the moment decisions than fail due to the inability to address a problem beyond the baseline means the system provides. Or to put it another way, I prefer to leave it up to my in session decisions and the dice to screw me over, I don’t want my demise preordained at character creation.

[OldTrees1]

1) The GM can, and in my experience does, try to patch any significant systemic issues. This does not mean those issues don't exist, but it does mean the GM takes the brunt of the blow.
2) 5E handles this better than 3E. 3E had problems with giving level appropriate features to martials by around Tier 2 (using 5E metrics). 5E delays this problem to around Tier 3. As always some have it a little before and some a little afterwards. Reliable Talent is nice at 11th.
3) Part of the problem is the independent problem caused by erroneously design ability checks using bounded accuracy's attack & damage model without providing the second half. This significantly shrinks the amount of progression available in that system. Combined with the anchoring effect of DC grades, WotC adventures' example DCs, and the tendency to start at 1st level, this readily results in DMs making ability checks risky maneuvers for non level appropriate results by Tier 2. But DMs can and do patch this issue since there are numerous ways to do so.

So no martials are not screwed in practice and it seems like WotC is improving in theory. Expect this to resurface every edition. Hopefully with a smaller and smaller problem each time.


Simple question to help solving the dilemma,
Would you go adventuring without a martial character, or without a full caster?
I am partial to abilities with faster to no recharge. Static abilities (Like Paladin Auras) are my favorite followed by at-will abilities (Like Cunning Action). So I am biased towards bringing a martial to the group.

However I found my healer specialized Paladin failed to handle conditions well enough. A PC was out of a session because I could not remove a curse. So I now doubt a group without a full caster will handle those issues to my satisfaction. So when faced with a group without a full caster, I would switch to a Cleric.

On the other hand if a group wanted to avoid martial characters, I could be convinced to not use a martial character despite my tendency.

So "yes without martials" and "no without a full caster", but it is nuanced and a close thing.

[heavyfuel]

In your experience, at the table, without signifigant homebrewing or buffing, have you ever felt that martials in 5e are actually suffering in any way?

A new thread seems as good a time as any to report an actual experience we had the previous weekend. Ie, not theorycraft.

Party is level 4, one player is an Rogue (Scout). We decide to go scout some mines which are obviously guarded.

The Rogue sneaks ahead and tells the DM he wants to shoot the guard non lethally. After a short debate about whether the Rogue could or not hit the guy with a crossbow bolt non-lethally, the DM says that it is impossible (which turns out to actually be RAW). So the rogue goes into melee.

Rogue is close by, everyone rolls initiative (including the guard), Rogue wins, hits the guard but doesn't kill him. An eldritch blast from the Warlock finishes the job. DM says that the blast made some noise, but that apparently no one heard it. We stabilize the guard before he dies. That went ok.
Next guard, this time, the Rogue lost initiative and went third (out four). Warlock wins init, and readies a spell for after the Rogue swings. Guard turn, he's surprised, does nothing. Rogue goes, hits for like 14 damage and one shots the guard. Only, because the guard was no longer surprised, that means he saw the Rogue swinging just in time to let out a scream of surprise.

The whole dungeon is now on alert.

This got me thinking about Rogues and the odds of them actually being able to do their typical job, which is sneaking.

Level 4 Rogue vs MM Guard (https://roll20.net/compendium/dnd5e/Guard#content)


1st - Stealth vs Passive Perception: With +4 Dex and Expertise, a lv 4 Rogue has +8 Stealth, so against a Guard with 12 PPercep, he has 85% chance of success.

2nd - Initiative: Rogue has +4, Guard has +1, so Rogue wins 66% of time.

3rd - Attack roll vs AC: Rogue has +6 and Advantage against 16. He needs a 10, which he has an 80% chance of getting.

4th - Dealing enough damage: He deals 3d6+4 with a SA he has a nearly 91% chance of KOing the guard.

All of this seems like good enough odds, but not really. The actual odds of all this hapenning is 40.7%. There's nearly a 60% chance that something will go wrong and the Guard will alert everyone.

A different subclass would have slightly different odds (Swashbuckler will have better initiative, Assassin has more damage), but no subclass tips the odds significantly.

This is a Challenge 1/8 Guard having a better chance of doing his job than the level 4 Rogue, who is a literal expert at sneaking about. Completely ridiculous.

[heavyfuel]

2) 5E handles this better than 3E. 3E had problems with giving level appropriate features to martials by around Tier 2 (using 5E metrics). 5E delays this problem to around Tier 3. As always some have it a little before and some a little afterwards. Reliable Talent is nice at 11th.

As a counterpoint, a lv 4 Rogue in 3e has 95% chance of KOing a level 1 Warrior, if we assume he's Taking 10 on spot and listen, which is the DMG recomendation. He always succeeds on Stealth due to the much higher modifier, he doesn't need to roll initiative due to the nature of a separate Surprise round, he always deals more than 4 damage, and he only misses the attack on a Nat 1.

The 5e Rogue doesn't have level appropriate features in Tier 1, let alone Tier 3

[KorvinStarmast]

This is a Challenge 1/8 Guard having a better chance of doing his job than the level 4 Rogue, who is a literal expert at sneaking about. Completely ridiculous. He sneaked just fine.

So the rogue goes into melee. He was out of his element in this case. What he didn't seem to be able to do was excel at one-on-one combat, which is more or less correct for the rogue who isn't a swashbuckler. Plus, the dice are fickle.

But I empathize with your frustration; the idea of sneaking up on the guard and getting surprise for being an unseen attacker should give him advantage, and thus the SA damage to allow for a non lethal knock out. Most of the time.

But guess what? The enemy gets a vote. That guard is supposed to be alert and not get sneaked up on. So the dice come into to play and sometimes, your rolls suck. That is a deliberate choice/feature of the d20 system. (Slope line distribution versus bell curve).

See the Introduction to the PHB by Mike Mearls.

You appear to be looking for an "I Win" button, which 3.5e appears to have had. Bounded accuracy tried to remove some of them. Feature, not bug.

[heavyfuel]

He sneaked just fine. What he didn't seem to be able to do was excel at one-on-one combat, which is more or less correct for the rogue who isn't a swashbuckler. Plus, the dice are fickle. That is a deliberate choice/feature of the d20 system. (Slope line distribution versus bell curve).

See the Introduction to the PHB by Mike Mearls.

It's not "one-on-one" combat. It's KOing a guard that hasn't seen him. Are you really going to argue that this isn't something that is expected of a Rogue?

40% chance of success isn't fickle, not for the character that's supposed to have an very good chance of success. It's so beyond fickle.

Not to mention, if every martial character has a ridiculous high chance of failuer due to a deliberate choice/feature of the d20 system, that is more than enough proof that Martials are screwed.

[KorvinStarmast]

It's not "one-on-one" combat. It's KOing a guard that hasn't seen him. Are you really going to argue that this isn't something that is expected of a Rogue?
No. (Please check my edit). IMO, the whole sneaking thing includes the bonus action "hide" (did you do that?) so that the guard does not see the Rogue, and then the Rogue gets the advantage attack - seems to me to be a great plan that should usually work.
Did the cleric or druid first cast Guidance on the rogue ahead of time to help his stealthing? Did the bard offer up bardic inspiration ahead of time to stack the deck a bit? Team success is the deal here, from where I sit.

But you still have to roll for it.

Also, vis a vis casters, monsters can roll saving throws or high perception checks. Any monster.

That's bounded accuracy. Casters can also get screwed.
My low level Sacred flames (Tempest Cleric) were incredible in how often they failed due to a monster making a dex saving throw during our first campaign. It drove me nuts. I still refer to that cantrip as "Suck red flame" to this day.

Last week, my Warlock cast Banishment on a salamander - and it failed. I burned a slot, one of two that I had, and nothing happened.
Are casters screwed?

40% chance of success isn't fickle, not for the character that's supposed to have an very good chance of success. It's so beyond fickle. Did rogue take expertise in Stealth? If not, why not?

Not to mention, if every martial character has a ridiculous high chance of failuer due to a deliberate choice/feature of the d20 system, that is more than enough proof that Martials are screwed. The enemy gets a vote.

[NaughtyTiger]

In your experience, at the table, without signifigant homebrewing or buffing, have you ever felt that martials in 5e are actually suffering in any way?
Having played almost exclusively AL, yes, i feel 5e melee-martials suffer compared to ranged-combatants or magical-controllers.

There are exceptions, but I find that the exceptions are cuz the player himself is really good tactically. So exceptional player skill can overcome it.


Would you go adventuring without a martial character, or without a full caster?
In my AL games, in the last 3 years: we have had magical players at every table, at least 1 full caster (cleric) at every table, often have 0 non-magical players.

In our homebrew game, we are all non-magical martials, combat was tough, including a near deadly gelatinous cube (that didn't attack, just advance)

Then a new player joined, Cleric, and the DM suddenly had to ramp up the difficulty a lot to account for the presence of magic.

[heavyfuel]

He sneaked just fine.
He was out of his element in this case.
being an unseen attacker should give him advantage
The enemy gets a vote.
That is a deliberate choice/feature of the d20 system.
You appear to be looking for an "I Win" button, which 3.5e appears to have had.

IMO, the whole sneaking thing includes the bonus action "hide" (did you do that?)
Did the cleric or druid first cast Guidance on the rogue ahead of time to help his stealthing? Did the bard offer up bardic inspiration ahead of time to stack the deck a bit?
Last week, my Warlock cast Banishment on a salamander - and it failed. I burned a slot, one of two that I had, and nothing happened.
Did rogue take expertise in Stealth? If not, why not?

You... so clearly didn't read my post. Most of your questions were answered there. But let me digest it for you.

He didn't sneak "just fine". The guard was walking with a lantern, so the DM said that if the Rogue tried to pass the guard, they would see him because he was no longer obscured. I didn't say that in the OP, maybe I should have. Fighting was the only option.

He wasn't "out of his element". Sneak attacking is very well within his element.

He did have advantage, he hit both times. Or did you not read my post before commenting on it?

The enemy should get a vote. I don't disagree with that. But if every enemy has a 60% chance of screwing the Rogue's plan, then it seems to me that the Rogue, a martial, is screwed.

I'm not looking for an "I win" button. My comment on 3E was a simple response that 5E martials don't have great features in Tiers 1 and 2. Which also supports the theory that they are screwed.

Every time you say "this is a feature of 5e", I read "5e made 'martials being screwed' a feature"

The Rogue was hidden. Why would he need to use his Bonus action hide if he was already hidden? Clearly you didn't read my post.

You seem to assume every party has a Cleric/Druid/Bard. We have none of those. Plus, relying on casters to help you succeed means you are screwed as a martial.

A creature making a save vs insta losing and still being able to be defeated through other means is so very very different than the Rogue having to succeed on 4 consecutive rolls to do anything expected of him. You still have another slot, we no longer have an opportunity to sneak.

Yes, he took Expertise in Stealth. I said that in my post. You would know that if you had bothered reading it.

[Spellweaver]

This is a Challenge 1/8 Guard having a better chance of doing his job than the level 4 Rogue, who is a literal expert at sneaking about. Completely ridiculous.

So the first part of the problem here is that your saying one thing and then doing another. To just sneak around is one thing. To auto kill or knock out any foe "guard" is a whole other thing. So the question here is why did the scout think that attacking a guard was a good idea? Simple logic would tell you that most of the time you can't "just kill or knock out a guard" on a whim. There is at least a 50% chance a typical guard can yell out. A sneaky scout avoids the guards, they don't take them out.

Now you could make a sneaky scout guard knockout character.....but here is the big rub: you would have to use resources to do it. Buy non lethal weapons. Maybe knock out poison?

Of course, 5E is Rules Lite....so in most cases the DM will have to homebrew everything(or barrow from 3.5). This is the sort of thing you bring up before the game though: "my sneaky character wants to be able to take out guards without killing them AND without them getting a warning shout out...how can my character do that?" And if the DM says "impossible! all guards will see you and alert the game world in .0001 seconds", then don't even bother trying. But if the DM is open to it, they might add some knock out rules or whatever else is needed.

THIS is the big point of 5E, much like 1E and 2E, the DM is free to make rules as needed.



It's not "one-on-one" combat. It's KOing a guard that hasn't seen him. Are you really going to argue that this isn't something that is expected of a Rogue?


Do you really feel this is a major rogue class feature? And if you do, would you be willing to give up current class features for it? For example, would you agree to lessen the effect of sneak attacks to get the KO ability?




He didn't sneak "just fine". The guard was walking with a lantern, so the DM said that if the Rogue tried to pass the guard, they would see him because he was no longer obscured. I didn't say that in the OP, maybe I should have. Fighting was the only option.

Um...fighting is never the only option. I can think of three other right now: You could have run away. You could have surrendered. You could have tried to bluff and trick the guard.

One of the biggest problems with martial characters is that the players feel they must be muderhobos.



then it seems to me that the Rogue, a martial, is screwed.

Does the guard count as a martial? Hummm




I'm not looking for an "I win" button. My comment on 3E was a simple response that 5E martials don't have great features in Tiers 1 and 2. Which also supports the theory that they are screwed.


Though again, THIS is the big point of 5E, much like 1E and 2E, the DM is free to make rules as needed.

[KorvinStarmast]

The enemy should get a vote. I don't disagree with that.
Good.
But if every enemy has a 60% chance of screwing the Rogue's plan, then it seems to me that the Rogue, a martial, is screwed. Maybe it's the team's plan that is at risk. Look, I have played a lot of thief/rogue characters over the years, and I get that what you were trying to do ought to be within reach.
What did his party do to help him? If there isn't someone who can cast guidance, well, the party, the team, will have troubles with a variety of ability checks, not just his checks.

I'm not looking for an "I win" button. OK, you are looking for ways to tip the odds in the Rogue's favor. Is that a fair assessment? Something like "the alert feat" would for example be a way to stack the deck in this rogue's favor, but I suspect that you'd rather not have a feat tax. Am I correct? Also, was there a party member who could have cast enhance ability? That would help with the sneak, right?

Cat’s Grace. The target has advantage on Dexterity checks. It also doesn’t take damage from falling 20 feet or less if it isn’t incapacitated.

how important to the party is this Rogue's Success? What assets will they expend to make that success happen?

Every time you say "this is a feature of 5e", I read "5e made 'martials being screwed' a feature"
The Rogue was hidden. Why would he need to use his Bonus action hide if he was already hidden? I guess that's a table to table thing. I have had DMs make one roll for the quiet movement, and one roll for "did you stay hidden" in cases similar to that. I agree with you there, it ought to all be one roll.

As to the hide and the sneak: this is IMO one of the clunky bits of stealth, hiding, and surprise that this edition has created as a problem. The Assassin sub class of Rogue in particular has a neat sounding feature that this clunky system seems to trip over. Surprise by itself has nuances that any of us who played previous editions will pull our hair out over until we get used to it. I am tempted to suggest that we return to rolling for surprise ...

You seem to assume every party has a Cleric/Druid/Bard.
I don't make that assumption: I was asking if that was added in since you didn't mention that you had or didn't have one of those three classes. Thank you for clearing that up. You did mention a warlock. My warlock got guidance at level 3, pact of the tome ...

Party is level 4, one player is an Rogue (Scout). We decide to go scout some mines which are obviously guarded. Was this a two person party?

Your story does raise a question for me that I am not sure that I can answer though: how does one provide the 'help' action for sneaking to offer a rogue or other scout advantage? There ought to be a way to do that, but at the moment, I am at a loss for how to mechanically apply that once the players describe what they are trying to do. (Hmm, is this a place where a subtle distraction provided by minor illusion might come into play? Yeah, your point on "takes a caster" is raised again).

[heavyfuel]

So the first part of the problem here is that your saying one thing and then doing another. To just sneak around is one thing. To auto kill or knock out any foe "guard" is a whole other thing.

Um...fighting is never the only option. I can think of three other right now: You could have run away. You could have surrendered. You could have tried to bluff and trick the guard.


I've already addressed that sneaking and KOing are one and the same in this situation because the DM said that only sneaking would mean being automatically detected once he entered the lantern radius and the guard turned around.

Allow me to rephrase that: "Fighting was the only option that allowed us to proceed with our goals, as retreating and surrendering would've accomplished nothing for reasons so obvious that I feel stupid for even having to explain this, and lack of Deception proficiency made bluffing even less optimal of a choice than sneaking".

[KorvinStarmast]

I've already addressed that sneaking and KOing are one and the same in this situation because the DM said that only sneaking would mean being automatically detected once he entered the lantern radius and the guard turned around.
You just gave me a thought about a way to resolve your situation, but I think that will depend a bit on the DM.

The Rogue sneaks ahead and tells the DM he wants to shoot the guard non lethally.
Does your DM allow you to try and stabilize or use "spare the dying" , or even a healing kit or a healing spell on enemies that you have just downed?

Order of operations: advantage with short bow/crossbow, hit, SA, guard drops, you or another party member immediately moves to stabilize or otherwise heal/stop being dead the guard. Guard is unconscious, but stable, your objectivce is achieved, it will be 1d4 hours before guard revives ...

I've seen it done both ways at various tables, but more DM's IME have allowed it than not. Varies by table.

Might have been something to discuss with the DM if your aim was to not kill the guard.

[MoiMagnus]

Simple question to help solving the dilemma,
Would you go adventuring without a martial character, or without a full caster?

I'm not sure this question really touch the problem. I'd reformulate it as

"Would you go adventuring with only full casters, or without any half/full caster at all?"
[So both alternatives exclude all kinds of half casters. And I'm unsure whether thrid-casters should be included in martials or not, because banning them out of martial subclasses probably means we should ban full caster subclasses that give martial features too...]

You could also add a third possibility of "only half-casters, third-casters, and martial-focussed subclasses of full-casters", and that would definitely be the one I choose.

[KorvinStarmast]

I've already addressed that sneaking and KOing are one and the same in this situation because the DM said that only sneaking would mean being automatically detected once he entered the lantern radius and the guard turned around. Hmm, way back in OD&D and AD&D 1e there were assassination tables based on level for assassins. Your dilemma gives me an idea with the whole cinematic "sneak up and knock 'em out" motif.

What if the rogue had a 'sap' skill that prevented the problem you ran into with that whole event needing 4 rolls? Thats' where the additive probability issue arose, it seems.

You'd have two rolls.
Roll 1: did you sneak effectively or not? If yes, then go to two ...
Roll 2: (from a table that goes up by rogue level). Did you knock 'em out?
Suggestion: Maybe base it off of something like the turn undead thing that clerics do.

Using a spell save DC versus your (something) CR X needs to roll a save versus "black jack to the head" or drop unconscious.

Is that feature too fiddly, or something along the lines of what you feel would support this kind of rogue operation better?

Here's a first draft. Base the save DC off of INT or DEX?


If you succeed on a stealthy aproach, or surprise an enemy, you can Sap them.

Sap
Rogue Level___ Knocks out a humanoid of CR
2d___________1/4 or lower
5th__________ 1/2 or lower
8th___________1 or lower
11th__________ 2 or lower
14th___________3 or lower
17th___________4 or lower
As many times per long rest as proficiency bonus, or maybe once per short rest ...

Anyway, it's an idea that I think fits into the theme and deals with the "why do I need to do four rolls to do this thing?"

Probably needs serious tweaking/adjustment ...
I am sure that someone will point out how this can be exploited ...
Maybe


If you succeed on a stealthy aproach, or surprise an enemy, you can Sap them.

Sap
Rogue Level___ Knocks out a humanoid of CR
2d___________1/8 or lower
5th__________ 1/4 or lower
8th___________1/2 or lower
11th__________ 1 or lower
14th___________2 or lower
17th___________3 or lower

[zinycor]

Having played almost exclusively AL, yes, i feel 5e melee-martials suffer compared to ranged-combatants or magical-controllers.

There are exceptions, but I find that the exceptions are cuz the player himself is really good tactically. So exceptional player skill can overcome it.

There isn't such a thing as a Melee martial. There are always ranged weapons that martials are proficient in and should be willing to use.

So in that regard, a martial character that doesn't do ranged attacks is just as sub-optimal as a caster who only prepares damaging spells.

[BloodSnake'sCha]

I never felt that martial class are lacking anything compered to casters.

[OldTrees1]

As a counterpoint, a lv 4 Rogue in 3e has 95% chance of KOing a level 1 Warrior, if we assume he's Taking 10 on spot and listen, which is the DMG recomendation. He always succeeds on Stealth due to the much higher modifier, he doesn't need to roll initiative due to the nature of a separate Surprise round, he always deals more than 4 damage, and he only misses the attack on a Nat 1.

The 5e Rogue doesn't have level appropriate features in Tier 1, let alone Tier 3

Hm. So in your case it boiled down to whether the party won enough initiative to down the guard before they recovered from surprise on their initiative. And you provide the math for that exact case.

Does your DM give the guard disadvantage on passive perception when you are stealthing in dim light? Or does only the PC get penalized for lighting conditions?

Did you notice Guards have 2 HD? Could they be better compared to a 2nd level Warrior from 3E (~11 hp)?

The guard could have been anything with PP 12, Dex 12, AC 16, HP 11. The CR 1/8 is more about the rest of the combat, not about the alpha strike. You were guaranteed to win the CR 1/8 fight. You had a 40% chance against the higher CR instant kill challenge and your DM ruled you had a 0% against the Lantern (a high enough stealth check should bypass blindsight, at some point a lantern should not be an instant loss, but that is DM dependant).

So there are 3 challenges. Defeat the foe. Sneak past the foe. Silently assassinate the foe. The circumstances in your case required the silent assassination in order to achieve sneaking past. However that does not mean that silent assassination should be as easy as sneaking past.

I am not really agreeing or disagreeing yet. I am still pondering when the Rogue should be able have better than 50% odds on killing a target before they recover from surprise.

[JFKsbatmobile]

Yea, martials are just fine. This is and always has been a people problem.

One of the great things about D&D is the Open Gaming. The rules cover combat, and maybe a thing or two that happens on an adventure....and that is it. Anything else the DM has to make up.

[Xervous]

Yea, martials are just fine. This is and always has been a people problem.

One of the great things about D&D is the Open Gaming. The rules cover combat, and maybe a thing or two that happens on an adventure....and that is it. Anything else the DM has to make up.

So in the absence of a favorable GM you would not argue with the statement that the system does not provide martials tools for social/exploration success in the same vein that it does for casters in spite of the two class archetypes being presented as equal and comparable? Or in short, martials are screwed out of reliable tools by the system.

[patchyman]

This is a Challenge 1/8 Guard having a better chance of doing his job than the level 4 Rogue, who is a literal expert at sneaking about. Completely ridiculous.

It seems that you are defining success extremely narrowly in your post. In order to qualify for success, the Rogue must successfully disable the Guard, without help, in the first turn, before the Guard acts.

That also depends on your GM ruling that the Guard is able to raise the alarm while surprised, rather than raising alarm only when they are able to act. (Not criticising your DM's ruling, just saying that many DMs might rule otherwise).

Even with those caveats, a generic Rogue succeeds 41% of the time. One shots the encounter, without taking damage or using any resources. In my books, that's a critical success, not a simple success.

Certain Rogues can push that chance of success even higher:
- Dual wielding Rogues get two chances to apply Sneak Attack to kill or knock out the Guard.
- Alert Rogues have a better chance of beating the Guard's initiative.
- Assassin Rogues get a critical hit if the Guard is surprised.
- Swashbuckler Rogues get Sneak Attack even if they don't beat the Guard's initiative.

Sidenote: What are a caster's chances of critically succeeding at that encounter (i.e. neutralising the guard without the guard raising the alarm)? If the character is alone, maybe the druid can avoid the guard by wildshaping into a rat (so the use of one short-term resource), but otherwise, it seems that the Rogue has the best chance of succeeding at this challenge.

[heavyfuel]

I'll make a more detailed post about it when I get home and I have access to my PC, but for now I'll just say this: The fact a lot people think it's perfectly okay that a Lv 4 Rogue gets shafted by a run of the mill guard over half the time is one of the reasons why martials are screwed.

[Willie the Duck]

I've already addressed that sneaking and KOing are one and the same in this situation because the DM said that only sneaking would mean being automatically detected once he entered the lantern radius and the guard turned around.

Allow me to rephrase that: "Fighting was the only option that allowed us to proceed with our goals, as retreating and surrendering would've accomplished nothing for reasons so obvious that I feel stupid for even having to explain this, and lack of Deception proficiency made bluffing even less optimal of a choice than sneaking".


So in the absence of a favorable GM you would not argue with the statement that the system does not provide martials tools for social/exploration success in the same vein that it does for casters in spite of the two class archetypes being presented as equal and comparable? Or in short, martials are screwed out of reliable tools by the system.

I've never really been bothered by the idea that a lot of the game is going to require a GM willing to work with you on how various things happen. However, heavyfuel is definitely showing how a GM who has it out for the players can really make things unfun (although I am unconvinced that a bunch of casters would have succeeded under this GM in this scenario). A GM who frames adventures in 'choose A or B' frameworks, especially if they haven't thought through the actual success percentiles they are then creating, are pretty hard for any rule system to reign in. Still, the one thing that I think 5e really does poorly is DM advice, training in new DMs, and the like.

[patchyman]

I'll make a more detailed post about it when I get home and I have access to my PC, but for now I'll just say this: The fact a lot people think it's perfectly okay that a Lv 4 Rogue gets shafted by a run of the mill guard over half the time is one of the reasons why martials are screwed.

Less than half the time. If the Rogue misses with his first attack (or the attack doesn’t kill the Guard), free object interaction to draw a dagger with his left hand and Bonus Action attack with the dagger with advantage and Sneak Attack.

And the fact that you consider an encounter where the opponent has a 50% chance to be knocked out or killed without getting any response, and a further 30% to end the turn with only a handful of hit points as “being shafted” is why it is hard to take you seriously when you say martials are screwed.

[JFKsbatmobile]

So in the absence of a favorable GM you would not argue with the statement that the system does not provide martials tools for social/exploration success in the same vein that it does for casters in spite of the two class archetypes being presented as equal and comparable? Or in short, martials are screwed out of reliable tools by the system.

The problem still is people. D&D has ability checks, skill checks and other abilities that people simply refuse to use. Plus there is the whole problem of over powering magic. The mageling cast a weak charm on the hill giant and the DM is all like "you now have a mind dominated slave forever!" and at the same time says for any ability or skill check "DC 100 you fail and suck!".

I guess some people won'y be happy until Martals get "abilities" exactly equal to spells...but just not called that. So there will be martial abilities of 1 to 9th level, and like at least 20 of each level. So a 1st level martal can pick the ability "Persuade Person", that does exactly what the spell Charm Person does, but it's all mundane.


I'll make a more detailed post about it when I get home and I have access to my PC, but for now I'll just say this: The fact a lot people think it's perfectly okay that a Lv 4 Rogue gets shafted by a run of the mill guard over half the time is one of the reasons why martials are screwed.

I see the problem as being from the other side:

The insistence of players that their characters just about always does whatever the character wants to do is the problem.

The game, any game, should be about equal and fair....not some lop sided mess where a character auto beats everything. That is not even really a "game".

Now, this would be fine for a game setting. Like if the players said "we want to play a game where the whole world is super, super weak so our characters are all powerful". Then the DM could make that game. But that in no way would be an average game.

In any average game, failure is part of the game. In basketball you won't always send the ball through the net every time. So the character failed...so what just keep on gaming.

[da newt]

If the rogue in question took the HIDE action successfully, why does initiative matter?

case 1: rogue wins initiative and strikes first from cover - if the attack is enough to incapacitate the guard this is a win.

case 2: guard wins initiative, on his turn he sees no enemy so he takes no action, rogue attacks from cover on his turn - if the attack is enough to incapacitate the guard this is a win.

What can we conclude - initiative is irrelevant in this scenario.

So with your provided %s,

85%*80%*91% = 62% chance of all three succeeding and the guard goes down with one short sword attack before raising the alarm.

Now any clever Rogue will have a weapon in each hand (otherwise that hand isn't doing anything useful). That certainly doesn't change the odds of a successful HIDE, but if the first attack misses you have a second chance to hit and do 3d6 damage. And what if you hit w/ first attack but don't take him out - still good chance to hit w/ offhand attack for d6 more - is that enough, almost certainly. I don't have the figures to do the math, but as you can see the odds of success just keep going up.

And this all assumes our rogue isn't clever enough to have poisoned his weapons or done anything to give himself a tactical advantage beyond trying to hide.

[Xervous]

I've never really been bothered by the idea that a lot of the game is going to require a GM willing to work with you on how various things happen. However, heavyfuel is definitely showing how a GM who has it out for the players can really make things unfun (although I am unconvinced that a bunch of casters would have succeeded under this GM in this scenario). A GM who frames adventures in 'choose A or B' frameworks, especially if they haven't thought through the actual success percentiles they are then creating, are pretty hard for any rule system to reign in. Still, the one thing that I think 5e really does poorly is DM advice, training in new DMs, and the like.

One of my underlying points is that the guidance necessary for GMs to find their way to the understanding of how martials need the red carpet rolled out while barriers are to be placed in front of casters is wholly absent from the rules documents. The system is at fault for presenting casters and martials as comparable when in fact there is a community consensus on addressing the differences that the system failed to do much as mention. Any one of us can put the proper handling of martials and casters in broad strokes sufficient to set a new GM down the right path, and do it in a paragraph or two. It’s not like this is startling new knowledge, two paragraphs could easily have been penned into the DMG. But they weren’t, and to me the absence seems to be either an intentional omission or far more likely a matter of ignorance.

[KorvinStarmast]

I'll make a more detailed post about it when I get home and I have access to my PC, but for now I'll just say this: The fact a lot people think it's perfectly okay that a Lv 4 Rogue gets shafted by a run of the mill guard over half the time is one of the reasons why martials are screwed. I think you made a decent point, and part of the problem seems to me to be "it takes four rolls to do this standard rogue thing!"

If you'll take a look at the idea I threw up there ... the "sap" sorta like 'turn undead' feature ... what do you think?

[Pex]

It seems that you are defining success extremely narrowly in your post. In order to qualify for success, the Rogue must successfully disable the Guard, without help, in the first turn, before the Guard acts.

That also depends on your GM ruling that the Guard is able to raise the alarm while surprised, rather than raising alarm only when they are able to act. (Not criticising your DM's ruling, just saying that many DMs might rule otherwise).




You make it sound like that's an unreasonable request.

Is it perfectly reasonable for a spellcaster to have done it with Sleep or Hold Person?

[MeimuHakurei]

I see the problem as being from the other side:

The insistence of players that their characters just about always does whatever the character wants to do is the problem.

The game, any game, should be about equal and fair....not some lop sided mess where a character auto beats everything. That is not even really a "game".

Now, this would be fine for a game setting. Like if the players said "we want to play a game where the whole world is super, super weak so our characters are all powerful". Then the DM could make that game. But that in no way would be an average game.

In any average game, failure is part of the game. In basketball you won't always send the ball through the net every time. So the character failed...so what just keep on gaming.

Let me run some numbers:

-A Level 4 Rogue, by himself, will have a 250 EXP budget for a Medium Encounter. Also, 125 EXP for Easy.

-A Guard is CR 1/8, which is worth 25 EXP.

-Only four such guards will be, by the numbers, a medium encounter for the Rogue (I don't expect him to beat them all in a straight fight, but to me it should be a decent challenge for him to sneak past all four)

So in result, you're saying the odds should be against a Rogue if trying to stealth takedown a single enemy way below their pay grade.

Also, here's how the math looks like if the Rogue is Level 20:

With 20 Dex, the rogue has a +17 to Stealth and easily enough damage to oneshot the guard every time, so those are guarantees.

Initiative is +5 vs. +1, a 77,25% chance to succeed.

Hitting the AC of 16 with a +11 bonus and advantage is a 96% chance.

The above adds up to a 74,16% chance for the stealth takedown to go off without a hitch. Still, more than a 1 and 4 chance for probably the best assassin in the world failing at a routine job does not sound very thrilling.

[Lupine]

While there’s discussion on how powerful or not casters of martial are, I personally have not seen the difference in the game I DM. By far, the most powerful classes are the monk and the paladin. (...aaaand to a lesser extent, the rogue. Helps when he actually pays attention.)

Perhaps I build my encounters in a way that makes this true. Perhaps my caster players are really bad at what they do (a hexblade warlock, and a swords bard, which strikes me now that they’re all basically martials)

But at my table, when the wheels fall off, and they need to lay out the hurt, they look to the monk and the paladin to do the hurting. That’s not to discount the others. The party punches well above their weight. But those two rule the battlefield.

[Composer99]

While there’s discussion on how powerful or not casters of martial are, I personally have not seen the difference in the game I DM. By far, the most powerful classes are the monk and the paladin. (...aaaand to a lesser extent, the rogue. Helps when he actually pays attention.)

Perhaps I build my encounters in a way that makes this true. Perhaps my caster players are really bad at what they do (a hexblade warlock, and a swords bard, which strikes me now that they’re all basically martials)

But at my table, when the wheels fall off, and they need to lay out the hurt, they look to the monk and the paladin to do the hurting. That’s not to discount the others. The party punches well above their weight. But those two rule the battlefield.

I can't speak for others, but the ways in which martials fall short, at least in my view, are as follows:
(1) With respect to their own features, a lack of interesting or strategic-level options and/or meaningful decisions to make in combat.
(2) With respect to their own features, a lack of mechanisms by which to more strongly and reliably affect the situations or problems they confront in the game outside of combat.

I state "with respect to their own features" because that is specifically where martials fall short. Both broad groups have interesting options or meaningful decisions to make when:
(1) using their ability scores and proficiencies;
(2) exploring a site or interacting with another creature;
(3) fighting other creatures and trying to use tactics or terrain or allies in a way that is not represented mechanically;
(4) they are playing in low-op games and can use their ASIs for feats that add interesting noncombat capabilities

However, spellcasters (especially those capable of casting up to 9th-level spells) also have spells, providing an additional layer of options and decisions both within combat and without, a layer that isn't as malleable as the above three circumstances or varying between DMs.

Where martials don't fall short is dealing damage to monsters.

I think it's important to note that there is nothing wrong with players who, preferring a simpler play experience, don't care to make use of martial features that extend their combat (and especially their noncombat) capabilities. But I think it's reasonable for martials to at least possess such features, even if they're ignored.

[heavyfuel]

As promised, here's a more detailed post.

In post #9, I made an offhand comment about how "this is a feature of 5e" basically means "5e made 'martials being screwed' a feature". I didn't give it a sencond thought at the time of writing it, but now I see that it is at the very core of the problem, that problem being:


Bounded Accuracy makes Martials suck.

Point 1 - It's impossible to be consistently good

Because of Bounded Accuracy, it's impossible for you to have a character that is consistently good at something. By "consistently good" I mean "at least a 90% chance of success". You are always at the mercy of the d20, and that makes any and all action you take a gamble. Wanna sneak? Wanna jump across a pit? Wanna scale a wall? Better hope you don't roll below a 7. This leads to point 2.

Point 2 - Multiple rolls are a death sentence

Now, you might be thinking that a 90% chance of success is WAY TOO HIGH! That would be broken! But that’s not the case, because a lot things Martials want to do require multiple rolls.

Climbing a wall requires an Athletics check to move half your speed. As does swimming. Sneaking also requires multiple rolls unless your goal is only to sneak (in which case you BETTER not roll low, because that means you’re probably alone in the middle of multiple enemies). So even with 90% chance of success, if you have to roll 4 times, you now only have a 65% chance of success. Without the ability to Take 10 or choosing to use your Passive score, rolling multiple times means you’re bound to fail.

This is drastically worsened by the fact that Bounded Accuracy means you probably don’t have 90% chance to begin with.

Point 3 - High-level martials are unnecessary due to Bounded Accuracy.

In previous editions, if you wanted or needed to kill a dragon, your best bet was to beg the adventurer that just stopped by your city, or at least keep a high-level character on retainer. Now, just hire an army of peasants to shoot the beast down before it even gets in range for a breath attack. The army of mooks is a far superior weapon than any Fighter, regardless of level.

Only creatures immune to non-magical BPS require high-level characters, and if this character happens to be a martial, they better have a magical weapon, which isn’t a guarantee in 5e. Well, they can still be dealt with an army of spell-casting mooks, but that’s a different story.

Conclusion

Martials sucking in 5e is a feature. Designers are on the record for making the classes suck on purpose and this will never change for 5e. I'm just happy to still play 3E on occasion, and I hope to continue to play it - and other systems like it - for the rest of my gaming life.

[Lupine]

I think it's important to note that there is nothing wrong with players who, preferring a simpler play experience, don't care to make use of martial features that extend their combat (and especially their noncombat) capabilities. But I think it's reasonable for martials to at least possess such features, even if they're ignored.

I respect that view, but I dislike your assumption that I run a “simpler” game, or that my players want a simpler experience.
Our Rogue comes up with the crazy plans. The Bard is a support character or face, as needed (player uncomfortable with RP). The Monk makes everything go right, and is face, as needed.

They may not have features that reward them for that activity, but that just means they have to be creative in their thinking. And if you think that Martials are “screwed” because they take extra creativity to play, then I think you have missed the point of D&D.

[heavyfuel]

There isn't such a thing as a Melee martial. There are always ranged weapons that martials are proficient in and should be willing to use.

So in that regard, a martial character that doesn't do ranged attacks is just as sub-optimal as a caster who only prepares damaging spells.

There are plenty of features that don't work with Ranged combat, so yes, there is such a thing as a Melee martial if you take it to mean "a martial that is far more effective in melee than they are at range". Throwing a couple javelins for 1d6+5 doesn't cut it as "effetive" at level 5, let alone at level 20.


Does your DM give the guard disadvantage on passive perception when you are stealthing in dim light? Or does only the PC get penalized for lighting conditions?

Could they be better compared to a 2nd level Warrior from 3E (~11 hp)?

Yes, this doesn't change anything because the lantern makes the area bright.

They could not. The lv 1 Warrior in 3E already has drastically higher Challenge than the Guard in 5e (1/2 CR vs 1/8 challenge).


However, heavyfuel is definitely showing how a GM who has it out for the players can really make things unfun (although I am unconvinced that a bunch of casters would have succeeded under this GM in this scenario).

To sneak you either need cover or be heavily obscured. In the abscence of cover, the DM rulled that a bright lantern made sneak impossible. That's not going against the players, that's following RAW.

Not every caster would have had success there, but a Druid would've fared much better than the Rogue, that's for sure.


I guess some people won'y be happy until Martals get "abilities" exactly equal to spells...but just not called that. So there will be martial abilities of 1 to 9th level, and like at least 20 of each level. So a 1st level martal can pick the ability "Persuade Person", that does exactly what the spell Charm Person does, but it's all mundane.


The insistence of players that their characters just about always does whatever the character wants to do is the problem.

Nobody has ever asked for a "Persuade Person" ability. Nobody has ever asked for "mundane Forcecage" or "mundane magic". Quit it with the strawman argument. All most people who argue that martials suck want is "hey, it would be nice if the book actually gave rules for stuff so I can point to a DC on a table and have my character do that"

Again with the strawman. Nor I nor anyone ever said martials should always succeed, but it would certainly be nice to have a decent chance of success.


I think you made a decent point, and part of the problem seems to me to be "it takes four rolls to do this standard rogue thing!"

If you'll take a look at the idea I threw up there ... the "sap" sorta like 'turn undead' feature ... what do you think?

I actually think it's great. I wouldn't even give it any restrictions. Just sap to your heart's content. I'd probably also give this feature to Rangers as well (with maybe a slightly slower scaling)


Also, here's how the math looks like if the Rogue is Level 20:

With 20 Dex, the rogue has a +17 to Stealth and easily enough damage to oneshot the guard every time, so those are guarantees.

Initiative is +5 vs. +1, a 77,25% chance to succeed.

Hitting the AC of 16 with a +11 bonus and advantage is a 96% chance.

The above adds up to a 74,16% chance for the stealth takedown to go off without a hitch. Still, more than a 1 and 4 chance for probably the best assassin in the world failing at a routine job does not sound very thrilling.

OMG that is so broken! I can't believe a level 20 Rogue can completely one shot someone nearly 75% of the time! #nerfrogues #roguestoostronk

[MoiMagnus]

Now, you might be thinking that a 90% chance of success is WAY TOO HIGH! That would be broken! But that’s not the case, because a lot things Martials want to do require multiple rolls.

Climbing a wall requires an Athletics check to move half your speed. As does swimming. Sneaking also requires multiple rolls unless your goal is only to sneak (in which case you BETTER not roll low, because that means you’re probably alone in the middle of multiple enemies). So even with 90% chance of success, if you have to roll 4 times, you now only have a 65% chance of success. Without the ability to Take 10 or choosing to use your Passive score, rolling multiple times means you’re bound to fail.

This is drastically worsened by the fact that Bounded Accuracy means you probably don’t have 90% chance to begin with.

That's assuming you need a perfect "all success" to succeed at a task, which will not be the case most of the time. [That's kind of the standard both in stories and reality to have something that goes wrong at the middle of a complex task, without triggering a complete failure of the task].

Maybe I was lucky with my DMs, but in sequences of skill checks, failed skill checks tend to be immediately followed by "what do you do to prevent failure?", which mean you usually need two failures in a row (or three in situations that were really at your advantage) to really suffer serious consequences.

[Also, climbing and swimming skill checks are explicitly not systematic. As for every skill check, it's up for the DM whether or not they are necessary, depending on circumstances, potentially depending on which character attempt the check. But I admit that 5e relies way to much on DM's discretion to determine whether or not a check is needed.]

[Sorinth]

This is a Challenge 1/8 Guard having a better chance of doing his job than the level 4 Rogue, who is a literal expert at sneaking about. Completely ridiculous.

The other way to look at it is hiring guards is completely pointless if a low level rogue can take them out 95% of the time.

Also since your 3.5 Rogue had to pass both his Hide and Move Silently check to sneak past his actual success rate is 90.25% of being hidden. A 5e rogue has a 85% chance, so not much different.

The 40% is because of combat and an unfavourable ruling. But it's intended that in combat even weak enemies are supposed to provide some sort of challenge to higher level characters. And the thing is this isn't martial specific, this is going to be a challenge to a spellcaster too. If instead of a Rogue you had a 4th level Wizard his odds aren't actually better.

[heavyfuel]

The other way to look at it is hiring guards is completely pointless if a low level rogue can take them out 95% of the time.

Also since your 3.5 Rogue had to pass both his Hide and Move Silently check to sneak past his actual success rate is 90.25% of being hidden. A 5e rogue has a 85% chance, so not much different.

Not really. Instruct your guards to have their backs against a wall, use the buddy system, and have well lit corridors. Boom. Guards are now useful again (although Martials are even more screwed in this situation)

You might be confusing 3.5 for a system where nat 1s auto-fail. It's not the case. A Rogue can beat a guard's Spot/Listen even if they roll a 1.

[porchdog]

All most people who argue that martials suck want is "hey, it would be nice if the book actually gave rules for stuff so I can point to a DC on a table and have my character do that"


So I noticed something about this: Magic use does not have a ton of rules for characters to do stuff. So if magic does not need them, why do mundane martials?

Sure spellcasters have additional layer of options and decisions both within combat and without, a layer that isn't as malleable as the above three circumstances or varying between DMs.

With No Rules.

And yet this is somehow a problem. So ask yourself why?



Not really. Instruct your guards to have their backs against a wall, use the buddy system, and have well lit corridors. Boom. Guards are now useful again (although Martials are even more screwed in this situation)


Is this not just nitpicking? Really? You will really say "oh if the game rules did just this one specific thing that I like then everything would be great?"

[heavyfuel]

So I noticed something about this: Magic use does not have a ton of rules for characters to do stuff.

Wat? Every single spell has a super strict rule. You can't use Fly to force a dex save versus fire damage, and you can't use a Fireball to gain a fly speed.

I honestly have no idea of what you're talking about.

[OldTrees1]

Bounded Accuracy makes Martials suck.

Point 1 - It's impossible to be consistently good

Because of Bounded Accuracy, it's impossible for you to have a character that is consistently good at something. By "consistently good" I mean "at least a 90% chance of success". You are always at the mercy of the d20, and that makes any and all action you take a gamble. Wanna sneak? Wanna jump across a pit? Wanna scale a wall? Better hope you don't roll below a 7. This leads to point 2.

Point 2 - Multiple rolls are a death sentence

Now, you might be thinking that a 90% chance of success is WAY TOO HIGH! That would be broken! But that’s not the case, because a lot things Martials want to do require multiple rolls.

Climbing a wall requires an Athletics check to move half your speed. As does swimming. Sneaking also requires multiple rolls unless your goal is only to sneak (in which case you BETTER not roll low, because that means you’re probably alone in the middle of multiple enemies). So even with 90% chance of success, if you have to roll 4 times, you now only have a 65% chance of success. Without the ability to Take 10 or choosing to use your Passive score, rolling multiple times means you’re bound to fail.

This is drastically worsened by the fact that Bounded Accuracy means you probably don’t have 90% chance to begin with.

Agreed. I can imagine a Tier 3 Barbarian wanting to do 15 ability/skill checks in a turn (with most but not all passing) as part of level appropriate behaviour. The only way that works is if the Barbarian can achieve consistency (I often use the word reliably) and even mastery (100+%) for the lower level checks.

I am not sure if I agree that this is a problem in Tier 1, I don't know if I could argue it about Tier 2, but it is definitely a problem in Tier 3+.


Point 3 - High-level martials are unnecessary due to Bounded Accuracy.

In previous editions, if you wanted or needed to kill a dragon, your best bet was to beg the adventurer that just stopped by your city, or at least keep a high-level character on retainer. Now, just hire an army of peasants to shoot the beast down before it even gets in range for a breath attack. The army of mooks is a far superior weapon than any Fighter, regardless of level.

Only creatures immune to non-magical BPS require high-level characters, and if this character happens to be a martial, they better have a magical weapon, which isn’t a guarantee in 5e. Well, they can still be dealt with an army of spell-casting mooks, but that’s a different story.

Conclusion

Martials sucking in 5e is a feature. Designers are on the record for making the classes suck on purpose and this will never change for 5e. I'm just happy to still play 3E on occasion, and I hope to continue to play it - and other systems like it - for the rest of my gaming life.

Interesting observation about outsourcing to lower level squads. It makes me think about what spells are not outsourced by squads of Tier 1 casters. And I can readily think of some. Then I think about it for martials. Expertise, when protected by Reliable Talent, is hard to outsource. But it does not scale as well as the spells I thought of on the other side.


Yes, this doesn't change anything because the lantern makes the area bright.

I am a bit confused on the logistics of the approach then. You attacked in melee with advantage which implied to me you were hidden on your approach. Your DM seemed to think bright light auto negates stealth. So that implied to me you stealthed through the dim light until you were behind the guard and then approached from the blind spot. So the guard would have disadvantage in my book. But, I also don't consider bright light to be as OP. This quibble / confusion did not impact the math much, I was just curious.


They could not. The lv 1 Warrior in 3E already has drastically higher Challenge than the Guard in 5e (1/2 CR vs 1/8 challenge).

Why not? The 5E Rogue is not facing a CR 1/8th challenge. Defeating the guard is a 1/8th challenge. Doing so before they have a chance to speak is a different challenge and would have a different CR. You are basically considering what if the guard had an reaction for +Yes ranged 1d6+Yes damage because if the guard gets a reaction the rogue lost the challenge, even if the fight was never in doubt. So the silent assassination CR is higher than 1/8th. When comparing their relevant abilities in this challenge space, a 2HD medium armored guard is roughly comparable to another 2HD medium armored guard.

The more I think about it, the more the guard seems to be a CR 2-3 for the silent assassination challenge (and a CR 1/8 for the actual fight). Now is that challenge too high for a "run of the mill guard"? That is less obvious.

[porchdog]

Wat? Every single spell has a super strict rule. You can't use Fly to force a dex save versus fire damage, and you can't use a Fireball to gain a fly speed.

I honestly have no idea of what you're talking about.

What I'm saying is there are no rules for magic of "stuff so I can point to a DC on a table and have my character do that" beyond the basic game rules, like saves and DCs.

Things that give a spellcaster "an additional layer of options and decisions both within combat and without".

Take your guard example:

The so called easy magic fix would be invisibility, right? Character turns invisible and NO lowly guard can detect them. Right? A great many players and DMs would say so, even if you don't.

So....where is the rule for that? Where does it say Invisibility equals perfect stealth and no foe can find you ever?

[OldTrees1]

What I'm saying is there are no rules for magic of "stuff so I can point to a DC on a table and have my character do that" beyond the basic game rules, like saves and DCs.

Things that give a spellcaster "an additional layer of options and decisions both within combat and without".

Take your guard example:

The so called easy magic fix would be invisibility, right? Character turns invisible and NO lowly guard can detect them. Right? A great many players and DMs would say so, even if you don't.

So....where is the rule for that? Where does it say Invisibility equals perfect stealth and no foe can find you ever?

Wait what?
1) Are you sure that is a common position? Unlikely to detect, sure. Can't detect, I am not sure. Most players and DMs have ears and can expect the guard to also have ears.
2) In 5E specifically there are rules that mention how to find them.

An invisible creature is impossible to see without the aid of magic or a Special sense. For the purpose of Hiding, the creature is heavily obscured. The creature’s location can be detected by any noise it makes or any tracks it leaves.

[heavyfuel]

I am a bit confused on the logistics of the approach then. You attacked in melee with advantage which implied to me you were hidden on your approach. Your DM seemed to think bright light auto negates stealth. So that implied to me you stealthed through the dim light until you were behind the guard and then approached from the blind spot. So the guard would have disadvantage in my book. But, I also don't consider bright light to be as OP. This quibble / confusion did not impact the math much, I was just curious.



Why not? The 5E Rogue is not facing a CR 1/8th challenge. Defeating the guard is a 1/8th challenge. Doing so before they have a chance to speak is a different challenge and would have a different CR. You are basically considering what if the guard had an reaction for +Yes ranged 1d6+Yes damage because if the guard gets a reaction the rogue lost the challenge, even if the fight was never in doubt. So the silent assassination CR is higher than 1/8th. When comparing their relevant abilities in this challenge space, a 2HD medium armored guard is roughly comparable to another 2HD medium armored guard.

The more I think about it, the more the guard seems to be a CR 2-3 for the silent assassination challenge (and a CR 1/8 for the actual fight). Now is that challenge too high for a "run of the mill guard"? That is less obvious.

I admit that my writing wasn't super clear on this. The Rogue wasn't in melee when the initiative was rolled. He was standing just outside the Bright Light radius of the Lantern, or "close by" as I put in my OP. Init was rolled, he won (the first time), moved in, and attacked.

DM rulled that, because the Rogue won Init, the Guard was still surprised even after he moved into Bright Light to attack. However, despite being surprised, the Guard would have seen the Rogue run past, because Bright Light would negate the Rogue's ability to use Stealth.



Besides my previous post, I also think they aren't comparable because they are both "run of the mill". A "run of the mill" guard in 3E is a lv 1 Warrior, a "run of the mill" guard in 5e is the MM Guard.

If run of the mill enemies are strong enough that you cannot adequately bypass them, then you are comparatively weaker. Run of the mill enemies shouldn't pose a threat to adventurers, who are supposed to be stellar at their jobs.

And even then, a level 2 Warrior in 3E still only has a slightly higher chance of surviving the attack, but will nonetheless be automatically surprised and will still be hit on anything but a nat 1. The chance of the level 2 Warrior surviving a successful SA from a mid OP Rogue (ie, Rogue with +1 Rapier and Craven feat and nothing else) is less than 10%. It only happens if the Rogue rolls 5 or less on 3d6.

[heavyfuel]

What I'm saying is there are no rules for magic of "stuff so I can point to a DC on a table and have my character do that" beyond the basic game rules, like saves and DCs.

Things that give a spellcaster "an additional layer of options and decisions both within combat and without".

Take your guard example:

The so called easy magic fix would be invisibility, right? Character turns invisible and NO lowly guard can detect them. Right? A great many players and DMs would say so, even if you don't.

So....where is the rule for that? Where does it say Invisibility equals perfect stealth and no foe can find you ever?

I've said that pobably only a Druid would be better than the Rogue here. Not the Wizard/Sorcerer.

I said this because Invisibility has very vague rules (though that is more to blame on skills, such as Stealth and Perception having vague rules). Some DMs will rule that Invis only lets you attempt to hide, but that you still have to make Stealth checks as normal (this actually seems to be the RAW)

However, even Invis has more rules than Stealth, and it's pretty much the only spell with super vague wording. Other spells are chock-full of rules, with super precise info in regards to duration, area, effects, etc.

[Telok]

Wait what?
1) Are you sure that is a common position? Unlikely to detect, sure. Can't detect, I am not sure. Most players and DMs have ears and can expect the guard to also have ears.
2) In 5E specifically there are rules that mention how to find them.

So, trying to ignore the invisibility radar & darkness is better derail, the point isn't that one rogue had an issue or wasn't char-opped to perform the action.

The point is that for things like taking out a guard before they raise the alarm non-casters have skill checks that often require multiple rolls to work, while the casters often have those same options plus magic that may autowin or make them better at it.

See, the challenge wasn't "take out one guard", it was "get the whole party past without killing the guard". Mr. Rogue can try to take out the guard (sneak, initative, hit, damage, 4 rolls to win), charisma the guard (very very DM & situation dependent), or distract the guard (convince DM, roll something, then 1 or more stealth checks). Mr. Caster can do all that too. Plus Mr. Caster may be able to hold person, silence, illusion distraction, familiar distraction, pass without trace, charm person, hypnotic pattern, disguise spell, etc., etc.

Now if your party is 4 non-casters and you need to get past 3 guards then you get your 40% to 60% chance try for each guard and have a ~20% or less chance of overall success. If your party is 4 casters you might spend 6 to 9 spell slots and have a 90% chance of success each time for ~73% chance of overall success.

As levels go up casters get bigger and better spells that do completely new and significantly better things. They go jump->levitate->fly->teleport. The non-casters go strength check + 2 -> strength check + 4 -> strength check + 6. That's what the complaint is about.

[KorvinStarmast]

I actually think it's great. I wouldn't even give it any restrictions. Just sap to your heart's content. I'd probably also give this feature to Rangers as well (with maybe a slightly slower scaling) Thanks, although I think there are some exploits in that rough idea, given that unconscious and incapacitated are such strong conditions to lay on someone. To keep it consistent with the way short rest or 'as many times as you have Stat Bonus' features work, it might be best as either once per short rest or 'per long rest based on stat bonus'

To fit within the rest of the game's basic structure.

[Ignimortis]

As promised, here's a more detailed post.

Good post. My friends who play 5e say that its' all irrelevant because it still works out narratively, and the fact that a thousand low-level archers can take down a dragon in one turn doesn't matter because they'd be too scared even before Frightful Presence came into play. Lack of high-level martial powers seems to be a "pro" for them, not a "con". I don't understand that, or at least can't express that without descending to namecalling. A game's rules are kind of like the game world's physics, and 5e's rules are severely grounded — until it comes to magic, that is.

People like to say that the disparity between martials and casters in 3.5 was way worse than in 5e. That is true on the PHB level. A Wizard/Fighter/Rogue/Cleric party is more balanced in 5e, sure. But 3.5 had so much more than the PHB. A Beguiler or Warmage/Warblade/Factotum/Divine Bard or Healer (that last one is a bit off-key, somehow you can't really replace the Cleric) party is much more balanced than 5e's default stuff.

Also, the game really lost a lot by not describing "take 10" and "take 20" as a mechanic, and instead giving it to the Rogue at level 11 and 20 respectively as class features (so only high-level rogues get to do stuff reliably). The only thing that seems to get to take 10 is Perception, by the dint of having a "Passive Perception" bar on charsheets.


So, trying to ignore the invisibility radar & darkness is better derail, the point isn't that one rogue had an issue or wasn't char-opped to perform the action.

The point is that for things like taking out a guard before they raise the alarm non-casters have skill checks that often require multiple rolls to work, while the casters often have those same options plus magic that may autowin or make them better at it.

<snip>

As levels go up casters get bigger and better spells that do completely new and significantly better things. They go jump->levitate->fly->teleport. The non-casters go strength check + 2 -> strength check + 4 -> strength check + 6. That's what the complaint is about.

In case of Stealth, Pass without Trace is a 2nd level Druid/Ranger/Trickery Cleric (Bard, if you really want to) spell. While under its' effect, a +2 DEX subject (i.e. most caster characters) gains a bonus to stealth that's better than having 17th level proficiency and maxed out DEX (+12 vs +11). Oh, did I mention that it affects the whole party at once and lasts up to an hour? Invisibility may allow you to forgo the Stealth roll in some cases, although it's DM-dependent and single-target.

[Pex]

As promised, here's a more detailed post.

In post #9, I made an offhand comment about how "this is a feature of 5e" basically means "5e made 'martials being screwed' a feature". I didn't give it a sencond thought at the time of writing it, but now I see that it is at the very core of the problem, that problem being:


Bounded Accuracy makes Martials suck.

Point 1 - It's impossible to be consistently good

Because of Bounded Accuracy, it's impossible for you to have a character that is consistently good at something. By "consistently good" I mean "at least a 90% chance of success". You are always at the mercy of the d20, and that makes any and all action you take a gamble. Wanna sneak? Wanna jump across a pit? Wanna scale a wall? Better hope you don't roll below a 7. This leads to point 2.

Point 2 - Multiple rolls are a death sentence

Now, you might be thinking that a 90% chance of success is WAY TOO HIGH! That would be broken! But that’s not the case, because a lot things Martials want to do require multiple rolls.

Climbing a wall requires an Athletics check to move half your speed. As does swimming. Sneaking also requires multiple rolls unless your goal is only to sneak (in which case you BETTER not roll low, because that means you’re probably alone in the middle of multiple enemies). So even with 90% chance of success, if you have to roll 4 times, you now only have a 65% chance of success. Without the ability to Take 10 or choosing to use your Passive score, rolling multiple times means you’re bound to fail.

This is drastically worsened by the fact that Bounded Accuracy means you probably don’t have 90% chance to begin with.



This is where I'm with you. This has always bothered me about 5E, tangentially related to my rant about skill use. You can never be just that good to autosucceed without DM fiat. The DM has to say no roll is needed, or the DC is whatever the DM feels like. "Mother May I". You as a player can never choose to be that good. Rogues get Reliable Talent so that's something, but only they get that. This is why I want Take 10/Take 20 back. Let the player choose to be just that good at something. No more depending on who is DM that day. I'm ok with this can mean Wizards autosucceed Knowledge checks to know about stuff. PCs know about the world they live in, and spellcasters are allowed to be good at things besides spellcasting.




Again with the strawman. Nor I nor anyone ever said martials should always succeed, but it would certainly be nice to have a decent chance of success.




For fairness I do want autosuccess for some things, but that's not the same thing as wanting autosuccess for everything nor wanting martials to punch holes in the multiverse to teleport or plane shift.

[OldTrees1]

So, trying to ignore the invisibility radar & darkness is better derail, the point isn't that one rogue had an issue or wasn't char-opped to perform the action.

Did you quote the wrong post?


I admit that my writing wasn't super clear on this. The Rogue wasn't in melee when the initiative was rolled. He was standing just outside the Bright Light radius of the Lantern, or "close by" as I put in my OP. Init was rolled, he won (the first time), moved in, and attacked.

DM rulled that, because the Rogue won Init, the Guard was still surprised even after he moved into Bright Light to attack. However, despite being surprised, the Guard would have seen the Rogue run past, because Bright Light would negate the Rogue's ability to use Stealth.

Ok. Makes sense. Thanks for clarifying. Seems strange that your DM gave you advantage in the same conditions that that ruled cancel your stealth. Surprise does not grant adv by itself normally. IIRC. But there are numerous variations between DMs.


Besides my previous post, I also think they aren't comparable because they are both "run of the mill". A "run of the mill" guard in 3E is a lv 1 Warrior, a "run of the mill" guard in 5e is the MM Guard.

If run of the mill enemies are strong enough that you cannot adequately bypass them, then you are comparatively weaker. Run of the mill enemies shouldn't pose a threat to adventurers, who are supposed to be stellar at their jobs.

And even then, a level 2 Warrior in 3E still only has a slightly higher chance of surviving the attack, but will nonetheless be automatically surprised and will still be hit on anything but a nat 1. The chance of the level 2 Warrior surviving a successful SA from a mid OP Rogue (ie, Rogue with +1 Rapier and Craven feat and nothing else) is less than 10%. It only happens if the Rogue rolls 5 or less on 3d6.

A 2HD guard is twice as hard to silently assassinate as a 1HD guard. Since you keep comparing to a level 1 warrior, perhaps your intuition would be better handled by a 1HD 5E guard? Also 5E guards, unlike 3E warriors, appear to actually be trained for the role of being a guard.

However you are also noting that the 5E Surprise rules are less one sided than the 3E Surprise rules. That decrease from 100% -> 66% chance of first action (followed by the 1HD -> 2HD) seems to be the biggest nerf to your low level Rogue. So let's redo that math real quick.

Stealth vs PP at Disadvantage is +8 vs 7. The 5E Rogue passes until they enter the bright light.
Initiative is still the 66% chance of going first.
Attack at advantage is still 80% but the 5E Rogue has 2 attacks. That is 96% of at least 1 hit.
I am assuming 2 daggers because I tend to not bother with bigger weapons on Rogues.
Expected chance of hitting and dealing at least 11 damage: 86.81%
After considering missing initiative: 57.87%

3E lvl 4 Rogue chance of hitting with 18 Dex, +1 Rapier, vs AC 16 is 65%.
5E AC 16 is higher than 3E AC 16 but 3E Rogue gets to hit flatfooted AC.
Why did you initially say a 3E 4th level Rogue would only miss a Warrior on a nat 1? Am I forgetting something major?
3E Rogue chance of hitting (3d6+5 right?) and dealing at least 11 damage: 37.51%

Seems like 5E Rogue is more accurate and more lethal but surprise was nerfed.

[Composer99]

I respect that view, but I dislike your assumption that I run a “simpler” game, or that my players want a simpler experience.
Our Rogue comes up with the crazy plans. The Bard is a support character or face, as needed (player uncomfortable with RP). The Monk makes everything go right, and is face, as needed.

They may not have features that reward them for that activity, but that just means they have to be creative in their thinking. And if you think that Martials are “screwed” because they take extra creativity to play, then I think you have missed the point of D&D.

I wasn't entirely clear, for which I apologise. I had meant for my remarks to serve not only as a specific reply to a point you raised with respect to the combat capabilities of characters at your table, but as a generally-addressed comment with respect to the topic of the thread as a whole.

Whereas, correct me if I'm wrong, but you were working from a less table-agnostic (as it were) position, and relying on a sort of "at my table" experience. Which is fine, as far as it goes, but feels a bit like it's talking past someone taking a more table-agnostic line.

As such, I intended to have such more generalised remarks that were not necessarily applicable to any given person's table.

Also, I have not stated that "martials are screwed", even if I do think they fall short in certain areas. I'll thank you to neither put words in my figurative mouth nor presume that you somehow know better what "the point" of D&D is, as if somehow "martials ought to have some more Nice Things, especially at higher levels" is somehow running contrary to whatever this "the point" might be.

As to creativity, your remarks quoted above seem closer, to my mind, to "missing the point" - not "the point" of D&D, but "the point" of what I am arguing. Martials can be played as creatively - or not - as their players desire. Well and good. But, if anything, it is casters who have a greater need for creativity: on average, they "have to be [more] creative in their thinking", because you have to either pick a spells known loadout that has the appropriate mix of supporting your desired primary role and offering some flexibility, or prepare spells that do whatever set of tasks you want to accomplish during the adventuring day, and then judiciously use your spell slots to support the party's combat or utility goals. Of course, you need not play that way as a caster - you can reserve your spells for blasting and pop a slot to that end once a fight (or more, depending on how many slots you have and how many fights you think you'll take part in) without doing anything else with them, if that's what you want.

"The point", as I stated upthread, is that either a fighter or a wizard can both be played with as much or as little creativity as players are willing or able to muster and DMs are willing to allow (viz. the four sets of activities that both groups can do in common), but casters have additional tools with which to leaven their creativity in each of the three pillars of play (but especially in the exploration and interaction pillars), tools that become vastly more powerful and flexible as they gain levels, while martials... just don't, not anywhere to the same extent - unless they take spellcasting features, that is.

I don't think it's too much to ask for fighters, rogues, and barbarians (and even, to a lesser extent, monks, rangers, and paladins) to have - for those players who want such things - access to more and better tools to leaven their creativity that are table-agnostic (to whatever extent that is possible), aren't spells, and offer mechanical support to make good on the two shortfalls I mentioned upthread. (They don't even have to be as powerful and flexible as spells, at least not in my view.)

[patchyman]

You make it sound like that's an unreasonable request.

Is it perfectly reasonable for a spellcaster to have done it with Sleep or Hold Person?

My point was that if the Rogue is required both to succeed a Stealth roll and an Initiative roll before incapacitating the Guard than so is the Wizard.

If the Wizard fails the Stealth roll, the Guard hears them and raises the alarm before the spell is cast. If they fail the initiative roll, the Guard raises the alarm before the spell is finished.

The Rogue’s higher Dex means he has a better chance of succeeding than the Wizard. Not only that, but he does’t use any resources, and at level 4, using a spell slot to incapacite a single guard leaves you with very few slots.

This particular case aside, in many cases you would be able to simply Stealth past the Guard.

[MeimuHakurei]

My point was that if the Rogue is required both to succeed a Stealth roll and an Initiative roll before incapacitating the Guard than so is the Wizard.

The Rogue’s higher Dex means he has a better chance of succeeding than the Wizard. Not only that, but he does’t use any resources, and at level 4, using a spell slot to incapacite a single guard leaves you with very few slots.

This particular case aside, in many cases you would be able to simply Stealth past the Guard.

A Wizard would've been able to stealth past because invisibility does let you sneak in open terrain and bright light, with advantage to boot. With a +5 to Stealth and advantage (I assume 16 Dex since it's a Wizard wanting to fill the stealth scout role), it's a 91% chance to do so. You wouldn't even have to cast anywhere near the guard as invisibility lasts for 1 hour. 91% vs. 80% chance seems like not that much but it'll add up when multiple checks are to be made.

And even if Expertise is that good, all it takes is one level dip for the Wizard to get it - Rogues will have to dip at least 3 levels if they want to cast Invisibility.

[Willie the Duck]

One of my underlying points is that the guidance necessary for GMs to find their way to the understanding of how martials need the red carpet rolled out while barriers are to be placed in front of casters is wholly absent from the rules documents. The system is at fault for presenting casters and martials as comparable when in fact there is a community consensus on addressing the differences that the system failed to do much as mention. Any one of us can put the proper handling of martials and casters in broad strokes sufficient to set a new GM down the right path, and do it in a paragraph or two. It’s not like this is startling new knowledge, two paragraphs could easily have been penned into the DMG. But they weren’t, and to me the absence seems to be either an intentional omission or far more likely a matter of ignorance.

The game has flaws. It also has things I think were deliberate, that I don't think serves our or starting out DMs purposes perfectly. I don't think we know what is intentional, done through ignorance, or done supporting a notion of how they were supposed to do things to which we are not privy.


To sneak you either need cover or be heavily obscured. In the abscence of cover, the DM rulled that a bright lantern made sneak impossible. That's not going against the players, that's following RAW.

Not every caster would have had success there, but a Druid would've fared much better than the Rogue, that's for sure.

I am not debating the RAW-ness of the DM making sneak impossible, I am suggesting that the setup where the only two options you had were sneak (which you cannot do), or one-shot the guard before they can act is the DM setting you up for failure (of there being plenty of third options about which the rest of us don't know, but we're assuming you are correct here).


OMG that is so broken! I can't believe a level 20 Rogue can completely one shot someone nearly 75% of the time! #nerfrogues #roguestoostronk

It is entirely reasonable for a system to not be set up to promote a character from being able to unceremoniously one-shot the opposition. It tends to make it extremely difficult to make reasonably-challenged encounters if any given creature can be taken out before acting, as well as make the game a series of rocket-tag events that really don't make long-term survival for the PCs tenable (excluding the DM effectively choosing not to use the PC's toolbox against them).

D&D, for most of its life, and including 5e, is predicated on the notion that combat is going to be a series of back-and-forth tests (mostly attack versus AC and damage against HP) before one side is victorious. In general, that is to the PCs benefit, as they are subjected to vastly more of these tests than any given opposition. In this very specific exchange, where you get exactly one round (if you win initiative) to determine the outcome of the situation, it does not serve you well. To that point, there should be more options for a rogue to be able to grapple-and-silence a single guard, such that they can get in those multiple attacks needed (or some other avenue towards success in this type of scenario). However, that is where I place the blame -- a rather limited amount of subsystems that support situational actions -- not a game designed towards making oneshotting opponents unlikely.


Martials sucking in 5e is a feature. Designers are on the record for making the classes suck on purpose and this will never change for 5e. I'm just happy to still play 3E on occasion, and I hope to continue to play it - and other systems like it - for the rest of my gaming life.

I do think that spellcasting really dominates in the upper tiers of the game. I also agree that the rather nebulous skill system is a turnoff for many. I am surprised by the notion that 3e D&D is being used as an example of a system that does martials right (especially since the instant you step outside of D&D, games where magic is either a fool's errand or costs exactly as much build resources as it benefits you), but if you find 3e to be the game that fits your needs, more power to you.

[heavyfuel]

It is entirely reasonable for a system to not be set up to promote a character from being able to unceremoniously one-shot the opposition.

Sure it's reasonable. If martials sucking is something you consider reasonable, that is. Because Wizards, Sorcerers, Clerics, Bards, and Warlocks of lv 20 are all capable of unceremoniously one-shot the opposition, especially when the opposition is a Challenge 1/8 Guard.

I never said Rogues should be able able to one-shot challenge appropriate threats, but when your level is 32 times higher than the creature's challenge, maybe they shouldn't be able to cause you so much trouble.

[Sorinth]

Not really. Instruct your guards to have their backs against a wall, use the buddy system, and have well lit corridors. Boom. Guards are now useful again (although Martials are even more screwed in this situation)

You might be confusing 3.5 for a system where nat 1s auto-fail. It's not the case. A Rogue can beat a guard's Spot/Listen even if they roll a 1.

You maybe right, it was probably a house rule that 1s would auto-fail skill checks. Anyways the difference between 3/3.5e and 5e is that they wanted challenges to stay challenges even as you leveled up. So it's normal that the chances of success are much less for 5e. But that's not caster/martial specific.

Getting back to martials vs caster, how are martials screwed in this situation? How is the level 4 caster knocking out the guard(s) and not alerting the rest of the complex? The caster still has to hide and unless they are spending a spell slot on Invisibility or Pass Without Trace they don't have nearly as good a chance of success as the Rogue. They still have to beat the guard initiative to not have him raise the alarm but again have a worse chance of doing that compared to the rogue. And then they have to take out the guard silently which might be tricky. KOing the guards at ranged risks the noise of the guard falling to the ground and making a big noise especially since he's armored. Whereas the melee rogue will grab the guard and gently lowers his body to the ground to not make a sound. Granted that may or may not matter, though if the complex is alerted by a surprised guard yelling when it's not his turn, I think the crash of metal from the guards armored body falling to the ground would probably alert anyone nearby. Which might mean you have to use a spell with a save.

So at the end of the day a Wizard using Invisibility and Sleep will accomplish the same thing at about the same chance of success but will have used 2 spell slots. In your example you had to do this twice, so the wizard had to spend over half his spell slots to get past a couple guards which puts him at a pretty big disadvantage once past the guards.

Also since this was a real example, and you think martials are screwed whereas a caster could breeze through this encounter why didn't the caster step in and "handle" it, especially after the near miss the first time shouldn't the caster have said, wait a minute why don't I handle this next one?

[heavyfuel]

So it's normal that the chances of success are much less for 5e. But that's not caster/martial specific.

Getting back to martials vs caster, how are martials screwed in this situation? How is the level 4 caster knocking out the guard(s) and not alerting the rest of the complex?

Which brings us back to my point that Bounded Accuracy makes martials suck.

Wildshape / Rat or Spider familiar to scout for us intead of the Rogue.

[Sorinth]

A Wizard would've been able to stealth past because invisibility does let you sneak in open terrain and bright light, with advantage to boot. With a +5 to Stealth and advantage (I assume 16 Dex since it's a Wizard wanting to fill the stealth scout role), it's a 91% chance to do so. You wouldn't even have to cast anywhere near the guard as invisibility lasts for 1 hour. 91% vs. 80% chance seems like not that much but it'll add up when multiple checks are to be made.

And even if Expertise is that good, all it takes is one level dip for the Wizard to get it - Rogues will have to dip at least 3 levels if they want to cast Invisibility.

By RAW you don't have advantage on your stealth roll. You are trying to move silently which is a straight stealth vs perception roll.

I also don't think it's wise to assume the Wizard has a 16 Dex and Stealth proficiency. Yes you can build a Wizard who is specialized in stealth, but specialized builds are irrelevant for a general martial vs caster debate. But since this was a real example, ask the OP if the party had a spellcaster with Invisibility prepared and what their actual stealth was.

Also the fact that Invisibility last 1hr is irrelevant, it would drop as soon as he cast a spell to deal with the guard. Which presumably had to be done to allow the whole party to get past the guard. Not too mention if the guard was guarding say a door you couldn't sneak past even with invisibility. So since there were two guard encounters this would actually force the Wizard to use 4 spell slots which is over half his total. Better hope these two guards were the only thing you needed the wizard to do that whole day.

[KorvinStarmast]

Also 5E guards, unlike 3E warriors, appear to actually be trained for the role of being a guard. Yep.

Seems like 5E Rogue is more accurate and more lethal but surprise was nerfed. The Assassin Rogue in particular feels that nerf.


I am not debating the RAW-ness of the DM making sneak impossible, I am suggesting that the setup where the only two options you had were sneak (which you cannot do), or one-shot the guard before they can act is the DM setting you up for failure (of there being plenty of third options about which the rest of us don't know, but we're assuming you are correct here).
I suggested that the rogue shoot from hidden (Sneak Attack, light cross bow) and then quick stabilize guard (Medicine check or healer kit medicine check... or spare the dying) but that relies on DM being good with PC attempts to try that. At the tables whre I play, more DMs than not go along with that ... so there's your third option. It's fiddly, but it fits the current mechanics more or less.

It is entirely reasonable for a system to not be set up to promote a character from being able to unceremoniously one-shot the opposition. The 4e minion is gone, the bag of HP monsters/NPCs are here.


D&D, for most of its life, and including 5e, is predicated on the notion that combat is going to be a series of back-and-forth tests (mostly attack versus AC and damage against HP) before one side is victorious. In general, that is to the PCs benefit, as they are subjected to vastly more of these tests than any given opposition. In this very specific exchange, where you get exactly one round (if you win initiative) to determine the outcome of the situation, it does not serve you well.
well said.


To that point, there should be more options for a rogue to be able to grapple-and-silence a single guard, such that they can get in those multiple attacks needed (or some other avenue towards success in this type of scenario). However, that is where I place the blame -- a rather limited amount of subsystems that support situational actions -- not a game designed towards making oneshotting opponents unlikely.
what did you think of my idea on the rogue "sap 'em" feature (https://forums.giantitp.com/showsinglepost.php?p=24610612&postcount=15)? Good, bad, or too open to exploits?

[Sorinth]

Which brings us back to my point that Bounded Accuracy makes martials suck.

Wildshape / Rat or Spider familiar to scout for us intead of the Rogue.

Except Bounded Accuracy also impacts casters.

Sure a Druid can wildshape into something and probably get past the guard (Far from guaranteed). Though what is the rest of the party going to do in order to get past the guard?

A familiar can scout, but is limited to 100ft, so assuming you are at least 60ft away from the guard to stay out of his lantern light you won't scout much with it. And again how is the party going to get past the guard.

And again since this was a real example from actual play, why did you and your party decide to go with the Rogue knocking them out? Did your party not have any casters in it?

[Xervous]

By RAW you don't have advantage on your stealth roll. You are trying to move silently which is a straight stealth vs perception roll.

I also don't think it's wise to assume the Wizard has a 16 Dex and Stealth proficiency. Yes you can build a Wizard who is specialized in stealth, but specialized builds are irrelevant for a general martial vs caster debate. But since this was a real example, ask the OP if the party had a spellcaster with Invisibility prepared and what their actual stealth was.

Also the fact that Invisibility last 1hr is irrelevant, it would drop as soon as he cast a spell to deal with the guard. Which presumably had to be done to allow the whole party to get past the guard. Not too mention if the guard was guarding say a door you couldn't sneak past even with invisibility. So since there were two guard encounters this would actually force the Wizard to use 4 spell slots which is over half his total. Better hope these two guards were the only thing you needed the wizard to do that whole day.

Do note that the original goal is to infiltrate the compound. A silent takedown was necessary because lantern beats rogue 100%. Invisibility users just need to stealth past.

Once we start scaling up the guard even a little bit the viability of an alpha strike plummets, leaving invisibility as the runaway winner. This makes the case of low grade guard and middling rogue an outlier, silent alpha strikes are the exception rather than the norm. The main winning move is just to not play (the quiet alpha strike game).

[KorvinStarmast]

And again since this was a real example from actual play, why did you and your party decide to go with the Rogue knocking them out? Did your party not have any casters in it? That was, I think, his point. "Martials are screwed, you have to have a caster do this ..." and then he illustrates the game mechanics issue of needing four rolls to get this done, with each additional roll of the dice reducing the probability of overall success by some amount. The issue with the caster is that if the spell slot is burned, that resource is gone.

In a "DMG based" adventure day of 6 encounters, with two short rests, that may be a steep resource price to pay. In the 5 MWD, not so much.

Of course, there are a number of other ways to deal with a guard; not all of them require the sneaky approach. But the sneaky approach is all over fiction, movies, books, video games, etc. So there's an expectation of that kind of move being viable.

[Xervous]

That was, I think, his point. "Martials are screwed, you have to have a caster do this ..." and then he illustrates the game mechanics issue of needing four rolls to get this done, with each roll of the dice reducing the probability of overall success by some amount.

Four different rolls when the caster has a feature that removes 3 of said rolls not with an auto success, but by letting the wizard make more use of the skill system (stealthing past).

[KorvinStarmast]

Four different rolls when the caster has a feature that removes 3 of said rolls not with an auto success, but by letting the wizard make more use of the skill system (stealthing past). Which costs the resource of a 2d level spell, while the rogue approach costs no resources. This takes me back to looking at the scenario from a team perspective, per my first response.

[Sorinth]

Do note that the original goal is to infiltrate the compound. A silent takedown was necessary because lantern beats rogue 100%. Invisibility users just need to stealth past.

Once we start scaling up the guard even a little bit the viability of an alpha strike plummets, leaving invisibility as the runaway winner. This makes the case of low grade guard and middling rogue an outlier, silent alpha strikes are the exception rather than the norm. The main winning move is just to not play (the quiet alpha strike game).

So again why didn't the caster in your party simply cast invisibility if that was all that was needed?

[Ignimortis]

Which costs the resource of a 2d level spell, while the rogue approach costs no resources. This takes me back to looking at the scenario from a team perspective, per my first response.

The further up you go, the less that 2nd level spell slot is going to matter in the grand scheme of things. The rogue's inability to consistently beat the guard at perception stays a thing until level 11, far past the point where a level 2 spell slot is a huge thing. A non-rogue's inability to consistently beat a guard's Perception by sneaking, even if they invest into it, stays a thing forever.

[Frozenstep]

Which costs the resource of a 2d level spell, while the rogue approach costs no resources. This takes me back to looking at the scenario from a team perspective, per my first response.

Sometimes I think this is actually kind of a problem. The rogue's big moment comes up, the moment where having someone good at stealth is going to pay off...but actually, it's to save the wizard a spell slot. That was your contribution. Of course that's valuable (though less so as the party levels up and spellcasters have more spells in general to spare), but it's not quite as exciting. Worse if it costs a spellcaster something they won't need before the next long rest (wild shape). It's also even worse if the spell is guaranteed to get you past a situation and the skill check isn't, because for any desperate, tense situation, you'll end up using the spell rather then take the risk.

Of course, you could help fix this by having a lot of situations requiring checks, so that it's a huge resource drain to try and solve them all, but then you're asking multiple checks of the martial and as discussed, that often leads to a failure. There's just a lot of things that need to be done to make the system really work. It's certainly possible, but it's so easy to create a situation that kind of steps on a martial's moment.

[KorvinStarmast]

Sometimes I think this is actually kind of a problem. The rogue's big moment comes up, the moment where having someone good at stealth is going to pay off...but actually, it's to save the wizard a spell slot. That was your contribution. Of course that's valuable (though less so as the party levels up and spellcasters have more spells in general to spare), but it's not quite as exciting. IMO, and your tastes may differ, if you approach this from a team perspective, it isn't a problem, but I do understand the desire for spotlights to shine on each player as a matter of table style.

I am all about team success; that's what excites me. I find that if the team does well, the DM will tend to ramp up the challenge and make us dig a little deeper to over come tougher situations. That may not be true for all tables.

For the record, in D&D 5e, I have mostly played support themed characters, with a Champion Fighter and a Gloom Stalker Ranger being outliers. Heck, my Shadow Sorcerer was mostly a support ... my sun soul monk got as much joy out of stunning enemies for the rest of the party to demolish as anything else ... and since I am a boring and predictable person, I always make my monks with a criminal background (something about a reformed criminal becomeing lawful pleases me as a concept) so that they can be the primary or the back up lock picking PC with thieves tools proficiency.


So again why didn't the caster in your party simply cast invisibility if that was all that was needed? At 4th level, burning a 2d level slot may be better spent in combat. There are not infinite spell slots, are there? All we know for sure is that there was a warlock in the party. Not sure what other casters were there.

And let's go with full disclosure on my perspective: I have an ingrained desire (from OD&D and AD&D 1e days) to not use spell slots unless you have to. I discovered in 5e during our first campaign, when our DM basically neglected the short rest, that it was all too easy to run out of spell slots way too fast and have nothing left. That, and some crap die rolls, is part of what led to KorvinStarmast's death; our party had a very bad habit, all of us, of using spells early and often. That, and the fact that turn undead resets on a short rest, blew up in our face.

[heavyfuel]

Except Bounded Accuracy also impacts casters.

Sure a Druid can wildshape into something and probably get past the guard (Far from guaranteed). Though what is the rest of the party going to do in order to get past the guard?

And again since this was a real example from actual play, why did you and your party decide to go with the Rogue knocking them out? Did your party not have any casters in it?

Far less than it impacts martials.

I'd say a Druid IS guaranteed to sneak past, unless the guards just happen to kill every random animal they see. And, not really. The goal was to investigate what was happening in the mines without causing a disturbance. Only the Rogue didn't want to go in alone for the obvious reasons.

Only a Warlock with Phatasmal Force (Archfey) and Darkness known.

[Sorinth]

That was, I think, his point. "Martials are screwed, you have to have a caster do this ..." and then he illustrates the game mechanics issue of needing four rolls to get this done, with each additional roll of the dice reducing the probability of overall success by some amount. The issue with the caster is that if the spell slot is burned, that resource is gone.

In a "DMG based" adventure day of 6 encounters, with two short rests, that may be a steep resource price to pay. In the 5 MWD, not so much.

Of course, there are a number of other ways to deal with a guard; not all of them require the sneaky approach. But the sneaky approach is all over fiction, movies, books, video games, etc. So there's an expectation of that kind of move being viable.

The thing is, he made it harder for him and his party. He decided to go melee instead of using ranged attacks. Yes he wanted to not kill the guard, but they could have used range then moved up and made a medicine check to stabilize the guard. Note that they basically did this in the first encounter where he didn't KO the guard with his first attack the Warlock used Eldritch Blast and then they stabilized the guard. So ranged was a viable option from the start which decreases the risk to almost nothing if they have a couple of ranged attackers.

Second, the goal was infiltration, so it's likely they could have tried entering the complex along a different path like say climbing up to a window and sneaking in that way. Made a small distraction that caused the guard to momentarily leave his post. They could have also tried to use social skills, after they took out the first guard, they could have dressed up in his uniform and bluffed their way past the 2nd guard. The list goes on, there were plenty of ways to do this that were probably less risky then combat.

Third, going invisible works great so long as you don't have to open a door. Because if the guard is at a door invisibility doesn't help sneak past anymore then the Rogue using stealth. It's also not an auto-success because you still have to make a straight stealth roll to beat the perception in order to move silently past the guard.

But I'm still waiting on the reason for the party not using magic to infiltrate the place to begin with. The OP is saying martials are screwed because my chance of success is so small, yet in game they still decided to go with the "martial" approach. It makes no sense to go with a plan where you are "screwed" and yet is easily solvable with magic. So it's likely that they didn't have anyone with magic that could actually bypass this which is the often overlooked aspect of martial vs caster debate, there's an opportunity cost associated with the magic solution so often times you won't have it available.

[patchyman]

A Wizard would've been able to stealth past because invisibility does let you sneak in open terrain and bright light, with advantage to boot. With a +5 to Stealth and advantage (I assume 16 Dex since it's a Wizard wanting to fill the stealth scout role), it's a 91% chance to do so. You wouldn't even have to cast anywhere near the guard as invisibility lasts for 1 hour. 91% vs. 80% chance seems like not that much but it'll add up when multiple checks are to be made.

Except we are assuming a Wizard filling a super stealth role, because we were not assuming a Rogue doing the same. So instead, your Wizard has a +2 (14 Dex and no prof).

Note that the Rogue incapacitating the Guard allows the rest of the party to pass as well.

So in your example, the Wizard uses one of his 3 level 2 spells to avoid notice by a single guard, does not help the party pass, and still has a 20% chance of failure.

[Sorinth]

Far less than it impacts martials.

I'd say a Druid IS guaranteed to sneak past, unless the guards just happen to kill every random animal they see. And, not really. The goal was to investigate what was happening in the mines without causing a disturbance. Only the Rogue didn't want to go in alone for the obvious reasons.

Only a Warlock with Phatasmal Force (Archfey) and Darkness known.

Well attacking rats/spiders is not crazy, but sure let's say the Druid can easily sneak past. Why didn't your Druid do that? Why was the plan for the Rogue and Warlock to go in by themselves?

[Sorinth]

At 4th level, burning a 2d level slot may be better spent in combat. There are not infinite spell slots, are there? All we know for sure is that there was a warlock in the party. Not sure what other casters were there.

And let's go with full disclosure on my perspective: I have an ingrained desire (from OD&D and AD&D 1e days) to not use spell slots unless you have to. I discovered in 5e during our first campaign, when our DM basically neglected the short rest, that it was all too easy to run out of spell slots way too fast and have nothing left. That, and some crap die rolls, is part of what led to KorvinStarmast's death; our party had a very bad habit, all of us, of using spells early and often. That, and the fact that turn undead resets on a short rest, blew up in our face.

Which is overlooked part of the whole martial vs caster debate. If the best course of action is to go with mundane non-magic solution then you can't claim martials are screwed.

[KorvinStarmast]

Which is overlooked part of the whole martial vs caster debate. If the best course of action is to go with mundane non-magic solution then you can't claim martials are screwed. It's not me who claims that .. might want to head back to the original thread. I find the position to be poorly supported, but YMMV.

The case being made for "why does the rogue need to roll 4 times to do this one thing" addresses things from a mechanical standpoint to illustrate a difficulty in an expectation versus mechanical implementation of that motif.

[OldTrees1]

It's not me who claims that .. might want to head back to the original thread. I find the position to be poorly supported, but YMMV.

The case being made for "why does the rogue need to roll 4 times to do this one thing" addresses things from a mechanical standpoint to illustrate a difficulty in an expectation versus mechanical implementation of that motif.

I think the 4 rolls comes from the fact that martials rarely want to do "one thing". Instead they want to do this compound thing.

Here the Rogue wanted to sneak up on the guard and incapacitate them before they could call out. In english that sounds like a 3 part compound plan. In D&D incapacitate is split into 2 parts (hit and deal enough damage to incapacitate).

At higher level the Rogue might want to:
Sneak around a bunch of guards. Silence the first guard with a locking garrote, knock out the second guard, cover the mouth of a 3rd guard, all without the 4th guard noticing. Not to mention these might be better trained guards with special senses.

[heavyfuel]

Well attacking rats/spiders is not crazy, but sure let's say the Druid can easily sneak past. Why didn't your Druid do that?

Because... we don't have a Druid?

[Sorinth]

Of course, you could help fix this by having a lot of situations requiring checks, so that it's a huge resource drain to try and solve them all, but then you're asking multiple checks of the martial and as discussed, that often leads to a failure. There's just a lot of things that need to be done to make the system really work. It's certainly possible, but it's so easy to create a situation that kind of steps on a martial's moment.

One way to handle the multiple skill checks situation is how you view failure. For instance a Rogue sneaking past a guard but fails to beat the perception check, maybe the guard heard something but isn't sure what it was. He'll investigate the noise he heard but if he can't find the source after a minute he gives up. This gives the Rogue a second chance of hiding (Assuming there is a place to get out of sight) and not getting caught. So the odds of failing both checks is much less likely so the Rogue is more likely to succeed overall but now there's a tense moment in there.

[Xervous]

One way to handle the multiple skill checks situation is how you view failure. For instance a Rogue sneaking past a guard but fails to beat the perception check, maybe the guard heard something but isn't sure what it was. He'll investigate the noise he heard but if he can't find the source after a minute he gives up. This gives the Rogue a second chance of hiding (Assuming there is a place to get out of sight) and not getting caught. So the odds of failing both checks is much less likely so the Rogue is more likely to succeed overall but now there's a tense moment in there.

See: rolling out the red carpet for Martials, DC to climb a tree. We can’t assume a favorable GM. Why not three failed rolls in a row? Above 60% you don’t have to roll. Plot is this way you don’t have to roll. Such distractions from the point are completely arbitrary and do not reflect upon the system so much as the community’s evolved understanding of which pegs don’t go in which holes when they’re all uniform shape and nothing indicates the colors matter.

[OldTrees1]

One way to handle the multiple skill checks situation is how you view failure. For instance a Rogue sneaking past a guard but fails to beat the perception check, maybe the guard heard something but isn't sure what it was. He'll investigate the noise he heard but if he can't find the source after a minute he gives up. This gives the Rogue a second chance of hiding (Assuming there is a place to get out of sight) and not getting caught. So the odds of failing both checks is much less likely so the Rogue is more likely to succeed overall but now there's a tense moment in there.

Can you apply this to the guard crying out?
When the rogue passed the initiative check, the surprised guard did not cry out when attacked.
When the rogue failed the initiative check, the once surprised but no longer surprised guard cried out when attacked.
The example heavyfuel gave seemed to hinge on the change in the surprise rules and this one 66% check in particular.

[Doug Lampert]

Sometimes I think this is actually kind of a problem. The rogue's big moment comes up, the moment where having someone good at stealth is going to pay off...but actually, it's to save the wizard a spell slot. That was your contribution.

And you know what's better for the team than having a higher failure chance to save a single level 2 slot? Being a full caster and having several level 2 slots of your own.

That's the problem with the teamwork argument. If the rogue's value of occasionally saving a spell slot at the cost of a higher failure chance is good because of teamwork, then the character bringing additional slots that are consistently available and more reliable when used is a better team player.

Admittedly, the BEST case (and what the rules should aim for), is for the rogue to be so much better at sneaking that the obvious "win" button is to cast invisibility on the rogue and by comparison putting it on the wizard or having the druid try to sneak past in animal form is clearly far worse. That actually encourages team play. Rogues get to be that much better at skills around level 11, by which time the slot is a trivial cost and no one cares that you've saved it (although a proper level 11 adventure may have things you want to sneak past that need the invisible rogue).

But till level 11, how is another caster with invisibility or wildshape NOT better for the party with a stealth mission than another rogue? And this is the rogue's alleged strong point.

[Kane0]

Something that may be pertinent: Clerics at level 5 with their CR 1/2 Destroy Undead. If undead fail save against this short rest aoe, they die.

Seems roughly comparable to ‘if you win inititive, single-roll for one CR 1/8 creature to die’

[KorvinStarmast]

Rogues get to be that much better at skills around level 11, by which time the slot is a trivial cost and no one cares that you've saved it (although a proper level 11 adventure may have things you want to sneak past that need the invisible rogue).

But till level 11, how is another caster with invisibility or wildshape NOT better for the party with a stealth mission than another rogue? And this is the rogue's alleged strong point. How many games last that long?

[Sorinth]

It's not me who claims that .. might want to head back to the original thread. I find the position to be poorly supported, but YMMV.

The case being made for "why does the rogue need to roll 4 times to do this one thing" addresses things from a mechanical standpoint to illustrate a difficulty in an expectation versus mechanical implementation of that motif.

Sorry I was actually agreeing, I meant the general you, not you specifically.

[Eldariel]

Something that may be pertinent: Clerics at level 5 with their CR 1/2 Destroy Undead. If undead fail save against this short rest aoe, they die.

Seems roughly comparable to ‘if you win inititive, single-roll for one CR 1/8 creature to die’

Honestly, if you can sneak up on someone, you should be able to do way better than 1d8+Dex+Sneak Attack damage in any case. At the very least it should be an automatic crit (but that's behind Assassin's niche protection and as a ****e ability, so that isn't sadly happening) though I could make an argument for just a simple Coup de Grace (autocrit and then Con-save with a DC of...damage dealt vs. death).

[Sorinth]

Because... we don't have a Druid?

Which just goes to show how martials aren't screwed/worse compared to casters because in actual gameplay you didn't have a caster capable of doing better.

That's the thing about this whole caster vs martial debate. Casters look great if they can be any caster and have prepared/know whichever spell is helpful to deal with the situation. But that completely ignores the real game play where there are opportunity costs with which spells you have and so often you won't have the go to magic solution.

[Sorinth]

Can you apply this to the guard crying out?
When the rogue passed the initiative check, the surprised guard did not cry out when attacked.
When the rogue failed the initiative check, the once surprised but no longer surprised guard cried out when attacked.
The example heavyfuel gave seemed to hinge on the change in the surprise rules and this one 66% check in particular.

Personally I didn't find that ruling particularly fair on the DMs part. I would've only allowed the guard to cry out on his turn. Even though he wasn't "surprised" anymore, he still couldn't even take reactions until the start of the next turn.

It's also worth noting the initiative problem is equally problematic for casters, fail the initiative and when the guard hears you casting he yells out. So for a martial vs caster debate it's irrelevant.

[Frozenstep]

IMO, and your tastes may differ, if you approach this from a team perspective, it isn't a problem, but I do understand the desire for spotlights to shine on each player as a matter of table style.

I am all about team success; that's what excites me. I find that if the team does well, the DM will tend to ramp up the challenge and make us dig a little deeper to over come tougher situations. That may not be true for all tables.

For the record, in D&D 5e, I have mostly played support themed characters, with a Champion Fighter and a Gloom Stalker Ranger being outliers. Heck, my Shadow Sorcerer was mostly a support ... my sun soul monk got as much joy out of stunning enemies for the rest of the party to demolish as anything else ... and since I am a boring and predictable person, I always make my monks with a criminal background (something about a reformed criminal becomeing lawful pleases me as a concept) so that they can be the primary or the back up lock picking PC with thieves tools proficiency.

At 4th level, burning a 2d level slot may be better spent in combat. There are not infinite spell slots, are there? All we know for sure is that there was a warlock in the party. Not sure what other casters were there.

And let's go with full disclosure on my perspective: I have an ingrained desire (from OD&D and AD&D 1e days) to not use spell slots unless you have to. I discovered in 5e during our first campaign, when our DM basically neglected the short rest, that it was all too easy to run out of spell slots way too fast and have nothing left. That, and some crap die rolls, is part of what led to KorvinStarmast's death; our party had a very bad habit, all of us, of using spells early and often. That, and the fact that turn undead resets on a short rest, blew up in our face.

It does come down to taste to some degree, but I think having a generalist who can do things better than a specialist, only limited by spell slots, is going to cause issues because it's a strange design. There are quite a few cases where a spell just works, but a skill check risks failure and consequences. And so when a situation is dramatic because there are visible consequences for failure, the spellcaster is the one who the party turns to, because those situations deserve a spell slot.

At some point, you risk the ideal party members for that limited-use generalist not being unlimited-use specialists, but rather more limited-use generalists. Of course, the problem may not really exist for the lower levels, but every extra level makes those low-level utility spells relatively cheaper.

If you can understand the desire for spotlights, then please try to understand how this kind of design can leave a bad taste.


One way to handle the multiple skill checks situation is how you view failure. For instance a Rogue sneaking past a guard but fails to beat the perception check, maybe the guard heard something but isn't sure what it was. He'll investigate the noise he heard but if he can't find the source after a minute he gives up. This gives the Rogue a second chance of hiding (Assuming there is a place to get out of sight) and not getting caught. So the odds of failing both checks is much less likely so the Rogue is more likely to succeed overall but now there's a tense moment in there.

Yes, there are ways to help solve the problem, but this is an intense amount of extra thought and set-up, which is why I say it takes a lot of work on the DM's part to help deal with these issues.

[Sorinth]

But till level 11, how is another caster with invisibility or wildshape NOT better for the party with a stealth mission than another rogue? And this is the rogue's alleged strong point.

Well a Rogue sneaking up and knocking a guard out provides a lot more options then one party member sneaking past. One party member sneaking past is good for scouting, but knocking out the guard allows someone to dress up as a guard and talk to other guards gaining different info, or allowing him into areas an invisible person can't enter like a closed door in sight of a guard, etc...

[Willie the Duck]

Sure it's reasonable. If martials sucking is something you consider reasonable, that is.
I hope you understand that comments like this would tend to make us believe that you're not interested in actually having a discussion at this point. I do not consider martials sucking to be reasonable. I disagree that this situation shows it to be the case. Please try to show your fellow thread participants a reasonable level of curtesy in assuming that they aren't mustachio-twirling cartoon villains.


Because Wizards, Sorcerers, Clerics, Bards, and Warlocks of lv 20 are all capable of unceremoniously one-shot the opposition, especially when the opposition is a Challenge 1/8 Guard.
I wholeheartedly agree that the martial-caster balance breaks down massively in the upper two tiers of play, but it has nothing to do with the situation at hand. A 20th level caster could use a spell. However, a 20th level fighter could run up and attack them 8+ times using action surge, almost guaranteeing success. In that instance it is the weird combat system rogues have been saddled with -- never more than 1 attack (2 with 2wf), with a huge (and situationally overkill) damage -- that makes this problematic for a rogue.

Also hamstringing the rogue is that they are the low-expendable resource class -- yes even a 3rd level caster could almost guarantee success on this one guard (if they happen to have silenceor invisibility memorized), but only by expending (what is supposed to be*) an huge portion of their resources (after which the fighters and rogues had better we willing to take point on the rest of the day's challenges). A rogue will over-contribute in the situations where one has to sneak past two separate guards at a reasonable interval, plus maybe having to scale a mineshaft for a bit, and then get past a locked door (and maybe a combat somewhere in the midst of it all, where the caster, if they have even gotten that far, will have nothing left but cantrips).
*yes, there is the whole 5mwd issue, to which there is actual guidance in the DMG if it is a problem in one's group.


I never said Rogues should be able able to one-shot challenge appropriate threats, but when your level is 32 times higher than the creature's challenge, maybe they shouldn't be able to cause you so much trouble.

I think, at some point, there is also a boundary issue/edge case going on. A creature's CR never gets so low that they do not get to roll for initiative, nor so low that you don't have to roll to hit to hit them (and CR 1/8 guards don't have ACs of 10-, as other low CR things tend to have), nor so low that they can't yell for other guards (who really are part of the challenge that the rogue is trying to bypass, something we should take into account).

Fundamentally, I think there is a point buried in all of this. And that would be that there are a bunch of things in D&D where casters just have to spend the spell slot, and then it just works, no check required, while everything a martial does requires a check (leading to potential issues of multiplicative checks). The general argument holds water. I just don't see this specific scenario as having the far-reaching implications you are giving it.


what did you think of my idea on the rogue "sap 'em" feature (https://forums.giantitp.com/showsinglepost.php?p=24610612&postcount=15)? Good, bad, or too open to exploits?
It's absolutely what people should be doing, and similar-to-which the DMG ought to have had as examples. There was language about letting people try to attempt real-world actions, but the examples pretty much boiled down to disarms and wrestling and the like, which clearly was insufficient. I know they didn't want a whole bunch of subsystems (or new cases of AD&D's unarmed combat mechanics, or the like), but guidance, suggestions, examples, goodness they could have included more.

[porchdog]

"The point", as I stated upthread, is that either a fighter or a wizard can both be played with as much or as little creativity as players are willing or able to muster and DMs are willing to allow (viz. the four sets of activities that both groups can do in common), but casters have additional tools with which to leaven their creativity in each of the three pillars of play (but especially in the exploration and interaction pillars), tools that become vastly more powerful and flexible as they gain levels, while martials... just don't, not anywhere to the same extent - unless they take spellcasting features, that is.

I would call into question the tools and their power and flexibility, used with creativity.

At the most basic, a RPG is where a GM presents a problem or obstacle or task that the player must solve. The easy way to do this is for the player to simply look down at their character sheet and pick a tool there to use. The hard way is for a player is look into their imagination and find a tool to use.

Though in both cases the use of a tool needs creativity and imagination. And most games have this big problem:

The big problem is Magic, specifically spells, gets a free pass. Both because gamers like magic more then mundane and magic is not real so it can "do anything". There is also the game problem that has the magic spell is printed in the game rules it gets a free pass...well...for no real reason other then the gamers can say that they are however vaguely using the game rules.

Anything else, other then spells, does not get a free pass. In fact, anything else get a hard core near lock down of being impossible. Any mundane activity is seen to be nearly useless and can only do some small things. And there is the problem that anything else is not printed in the rule book, so gamers will feel like they are not playing using the rules to play the game.

Take for example two near identical "DM call" situations and what most often happens. The group needs to distract a guard and get them away from a door they are in front of.

Mundane way: The halfling fighter throws a rock into some trees on make a noise. The DM, not giving any mundane action any chance and not liking that the player did a non rules action, just rules "ha, the guard just ignores the silly noise"

Magic Way: The elf wizard uses Mage Hand to throw a rock into some trees on make a noise. The DM being a big fan of magic and liking that the player (sort of) used the game rules, says "Wow...the guard runs away from the door at full speed and goes over into the trees to look for the source of the noise!"

The only difference is one is mundane and one is magic, but the reaction is very different. And this is pure DM Whim, no rolls or rules: just what the DM says happens.

Like I said, this is a Big Problem.

[Sorinth]

See: rolling out the red carpet for Martials, DC to climb a tree. We can’t assume a favorable GM. Why not three failed rolls in a row? Above 60% you don’t have to roll. Plot is this way you don’t have to roll. Such distractions from the point are completely arbitrary and do not reflect upon the system so much as the community’s evolved understanding of which pegs don’t go in which holes when they’re all uniform shape and nothing indicates the colors matter.

Deciding what failed check actually represents is always going to be arbitrary. And yes he can make it super tough on you like the fail initiative and the alarm goes off, or they can be generous. Spellcasting is irrelevant to that. The wizard cast Invisibility on himself and tries to sneak past silently, well he's probably only got a 50/50 chance to get by without making a sound since it's still a straight stealth roll vs passive perception. If he fails that roll how the guard reacts is in the hands of the DM (Raise the alarm or not) same as with the martial character.

[prabe]

Take for example two near identical "DM call" situations and what most often happens. The group needs to distract a guard and get them away from a door they are in front of.

Mundane way: The halfling fighter throws a rock into some trees on make a noise. The DM, not giving any mundane action any chance and not liking that the player did a non rules action, just rules "ha, the guard just ignores the silly noise"

Magic Way: The elf wizard uses Mage Hand to throw a rock into some trees on make a noise. The DM being a big fan of magic and liking that the player (sort of) used the game rules, says "Wow...the guard runs away from the door at full speed and goes over into the trees to look for the source of the noise!"

The only difference is one is mundane and one is magic, but the reaction is very different. And this is pure DM Whim, no rolls or rules: just what the DM says happens.

Like I said, this is a Big Problem.

That's just bad DMing. Like, so bad it almost feels like a strawman argument, though I have no reason to believe it is--it in fact is depressingly plausible as an at-the-table experience.

[Xervous]

I would call into question the tools and their power and flexibility, used with creativity.

At the most basic, a RPG is where a GM presents a problem or obstacle or task that the player must solve. The easy way to do this is for the player to simply look down at their character sheet and pick a tool there to use. The hard way is for a player is look into their imagination and find a tool to use.

Though in both cases the use of a tool needs creativity and imagination. And most games have this big problem:

The big problem is Magic, specifically spells, gets a free pass. Both because gamers like magic more then mundane and magic is not real so it can "do anything". There is also the game problem that has the magic spell is printed in the game rules it gets a free pass...well...for no real reason other then the gamers can say that they are however vaguely using the game rules.

Anything else, other then spells, does not get a free pass. In fact, anything else get a hard core near lock down of being impossible. Any mundane activity is seen to be nearly useless and can only do some small things. And there is the problem that anything else is not printed in the rule book, so gamers will feel like they are not playing using the rules to play the game.

Take for example two near identical "DM call" situations and what most often happens. The group needs to distract a guard and get them away from a door they are in front of.

Mundane way: The halfling fighter throws a rock into some trees on make a noise. The DM, not giving any mundane action any chance and not liking that the player did a non rules action, just rules "ha, the guard just ignores the silly noise"

Magic Way: The elf wizard uses Mage Hand to throw a rock into some trees on make a noise. The DM being a big fan of magic and liking that the player (sort of) used the game rules, says "Wow...the guard runs away from the door at full speed and goes over into the trees to look for the source of the noise!"

The only difference is one is mundane and one is magic, but the reaction is very different. And this is pure DM Whim, no rolls or rules: just what the DM says happens.

Like I said, this is a Big Problem.

Yes, I agree bad GMs can be a big problem. Look at the wealth of horror stories about terrible GMs. This post doesn’t support much else with a lonely anecdote.

[MeimuHakurei]

That's just bad DMing. Like, so bad it almost feels like a strawman argument, though I have no reason to believe it is--it in fact is depressingly plausible as an at-the-table experience.

It's a made-up strawman of an argument nobody made.

I'd rather not be burdened to run a game where can never allow the party a day off, never let them get away with fighting less than 6 encounters in a day, need to homebrew in plenty of caster monsters/NPCs, many of which pack counterspell, make up an entire subsystem of gathering followers whole-cloth, harshly restrict access to spell components, splurge out significant amounts of magic items I need to be careful to not be better used by a caster, avoid most kinds of challenging terrain and ever placing anything at a position not accessible via foot travel just to have a balanced game.

Also, did you notice how often 5th or higher level spells are considered irrelevant because "nobody plays at that level", yet we assume that Rogues get to frequently use Reliable Talent and Fighters get their four attacks a turn?

[Xervous]

Deciding what failed check actually represents is always going to be arbitrary. And yes he can make it super tough on you like the fail initiative and the alarm goes off, or they can be generous. Spellcasting is irrelevant to that. The wizard cast Invisibility on himself and tries to sneak past silently, well he's probably only got a 50/50 chance to get by without making a sound since it's still a straight stealth roll vs passive perception. If he fails that roll how the guard reacts is in the hands of the DM (Raise the alarm or not) same as with the martial character.

So, by its very nature of being arbitrary, we clearly have no foundation for debating such resolutions with respect to the mechanics of the game. Either we can debate the mechanical resolutions or fuss about which version of calvinball is the right one.

[prabe]

It's a made-up strawman of an argument nobody made.

I'd rather not be burdened to run a game where can never allow the party a day off, never let them get away with fighting less than 6 encounters in a day, need to homebrew in plenty of caster monsters/NPCs, many of which pack counterspell, make up an entire subsystem of gathering followers whole-cloth, harshly restrict access to spell components, splurge out significant amounts of magic items I need to be careful to not be better used by a caster, avoid most kinds of challenging terrain and ever placing anything at a position not accessible via foot travel just to have a balanced game.

Also, did you notice how often 5th or higher level spells are considered irrelevant because "nobody plays at that level", yet we assume that Rogues get to frequently use Reliable Talent and Fighters get their four attacks a turn?

Huh. I don't do any of those things, and I haven't noticed a major imbalance in either of the parties I'm DMing. Now, one of those parties is at 12th level and the other is at (I think) 6th, so whatever imbalances arise at higher levels, I haven't encountered much yet. YMMV, obviously.

[MeimuHakurei]

Huh. I don't do any of those things, and I haven't noticed a major imbalance in either of the parties I'm DMing. Now, one of those parties is at 12th level and the other is at (I think) 6th, so whatever imbalances arise at higher levels, I haven't encountered much yet. YMMV, obviously.

Truth is, making it so that everyone contributes isn't a tall order, especially with very casual/inexperienced players. Not to mention that all you really need is guide players towards a certain problem solving tool for any given campaign obstacle and the vast majority of the imbalance (out of combat) is more or less obviated.

It's usually more pronounced with more experienced players who are more aware of what you can do with magic, as well as more challenging modules where you don't have a path to victory laid out for you or where environmental obstacles become more frequent and challenging.

[KorvinStarmast]

If you can understand the desire for spotlights, then please try to understand how this kind of design can leave a bad taste. I see it, and understand the point. The quadratic wizard linear fighter has been with us since 1974. :smallwink: You can say that it still with us somewhat, though perhaps not as pronounced as in some previous editions. What has changed in 5e is that the wizard is nowhere nears as fragile 'out of the box' as when I made my first magic user in 1975. Whether that's a good design implementation or not is also a matter of taste.

[Morty]

The further up you go, the less that 2nd level spell slot is going to matter in the grand scheme of things. The rogue's inability to consistently beat the guard at perception stays a thing until level 11, far past the point where a level 2 spell slot is a huge thing. A non-rogue's inability to consistently beat a guard's Perception by sneaking, even if they invest into it, stays a thing forever.

Even if a rogue can reliably sneak at level 1, it's as good as it gets. A caster always has new, shiny spells to look forward to, but a non-caster mostly gets incrementally better at stuff they could do already - there's some subclass abilities that are more exciting, but rogues sure as heck don't get them.

[Pex]

Which costs the resource of a 2d level spell, while the rogue approach costs no resources. This takes me back to looking at the scenario from a team perspective, per my first response.

Using the spell slot is a cost, but for this concept the use of the resource is irrelevant to the matter. The wizard wants to use it, so it is worth the slot. By using his class feature he bypasses the problem with ease. The issue is the rogue doesn't have a button to push to bypass the problem with ease. He's not spending resources, but he relies on the skill system with its inherent issues of DM fiat on how that works. He has to make the rolls with each one contributing to the chance of failure. That he doesn't have a resource to spend to solve the problem is the angst. The resource doesn't have to be a button to push. It could have been something spent in character creation to enable having an easy time to the solve the problem, at least as easy as the wizard who gets to press a button/use a spell slot.

[Sorinth]

Even if a rogue can reliably sneak at level 1, it's as good as it gets. A caster always has new, shiny spells to look forward to, but a non-caster mostly gets incrementally better at stuff they could do already - there's some subclass abilities that are more exciting, but rogues sure as heck don't get them.

Out of curiosity what should a high level rogue be able to do that a lower level rogue can't even attempt to do? Because it seems normal that without magic you can attempt to do something regardless of level and that level mainly impacts how successful you'll be.

[KorvinStarmast]

Using the spell slot is a cost, but for this concept the use of the resource is irrelevant to the matter. The wizard wants to use it, so it is worth the slot. By using his class feature he bypasses the problem with ease. The issue is the rogue doesn't have a button to push to bypass the problem with ease. Did you read my proposed Sap feature? (https://forums.giantitp.com/showsinglepost.php?p=24610612&postcount=15) If so, what do you think of it? (I am sure that it has at least one loophole that need to be closed, since knocking an enemy unconscious makes them massively vulnerable due to Conditions ...)

When you position this like "I want an Easy Button to do This" we are maybe back to "Wait, you want an I Win button but don't want to expend a resource to do it?" Hence my idea on the "sap" feature. The spell caster expends a resource. That has an impact on the whole party, since that spell slot isn't there for what happens next.
Under the presumption of the six or so encounters for the adventure day, that's a non trivial cost.
No hold person when you need it
No lesser restoration when you need to unparalyze an ally
No misty step when you need it
No Phantasmal Force (in this case) to distract a foe ...

[Sorinth]

Using the spell slot is a cost, but for this concept the use of the resource is irrelevant to the matter. The wizard wants to use it, so it is worth the slot. By using his class feature he bypasses the problem with ease. The issue is the rogue doesn't have a button to push to bypass the problem with ease. He's not spending resources, but he relies on the skill system with its inherent issues of DM fiat on how that works. He has to make the rolls with each one contributing to the chance of failure. That he doesn't have a resource to spend to solve the problem is the angst. The resource doesn't have to be a button to push. It could have been something spent in character creation to enable having an easy time to the solve the problem, at least as easy as the wizard who gets to press a button/use a spell slot.

The wizard doesn't bypass the guard with ease even with Invisibility. He has to be quiet as he moves past to not alert the guard, that's a straight up stealth roll vs his passive perception. Which for most wizards is going to be something like a 50/50 chance.

Also it's not an apples to apples comparison. A KOed guard is much different end result then an invisible wizard who has snuck past said guard. The odds of a wizard getting the same end result of a KOed guard without making any noise is actually probably much less then that of a Rogue.

[OldTrees1]

Out of curiosity what should a high level rogue be able to do that a lower level rogue can't even attempt to do? Because it seems normal that without magic you can attempt to do something regardless of level and that level mainly impacts how successful you'll be.

Have you tried to hide in bright light? I expect most rogues can by mid level.

Have you ever snuck up on a Dragon? Blindsight is nice and all, but the mid level Rogue understands how it works and how to hide from it.

As for high level Rogues, who even knows you exist? You are out of sight and out of mind. Traces of your presence have been disguised to be coincidences or the actions of your comrades. You can fool even those using divination by leading their assumptions elsewhere.

The main insight is that D&D can and has treated skills as EXtraordinary, not mundane. And the best way to do that is to have progression.

[prabe]

Truth is, making it so that everyone contributes isn't a tall order, especially with very casual/inexperienced players. Not to mention that all you really need is guide players towards a certain problem solving tool for any given campaign obstacle and the vast majority of the imbalance (out of combat) is more or less obviated.

It's usually more pronounced with more experienced players who are more aware of what you can do with magic, as well as more challenging modules where you don't have a path to victory laid out for you or where environmental obstacles become more frequent and challenging.

I wouldn't say the players at either table I'm DMing are particularly inexperienced, but I would say they're not overly interested in ruthlessly maximizing their characters--I think that's where the discussions of martials being screwed come from. Of course, I'm not guiding them toward anything, and I'm not laying out any paths to victory for them, either.

So, I guess your expectations aren't entirely correct. Or even particularly at all, from where I'm sitting.

[heavyfuel]

Which just goes to show how martials aren't screwed/worse compared to casters because in actual gameplay you didn't have a caster capable of doing better.

Wait. So your argument is that, because in a party of 3 people, no one decided to play Druid, then the Druid isn't an amazing class?

Really?

That's the most ridiculous argument I've heard. Your sample size is one single group out of countless groups.

[Doug Lampert]

SNIP thrown rock example

Note that in 3.x RAW diplomacy discussions, I was repeatedly told, "It isn't charm person".

I agreed, Charm Person gives a save, allows spell resistance, only works on some targets, and makes the target FRIENDLY, which is a defined term and the description was 100% in line with that term.

Diplomacy gave no save, allowed no spell resistance, worked on absolutely anything you could communicate with (up to and including gods, and it wasn't hard to get telepathy by mid levels), and it gave the better and stronger HELPFUL condition.

There was absolutely no way I could see to read the rules, and claim that diplomacy couldn't do far more more effectively and more reliably than charm person.

Yet people still claimed charm person was better because it was a spell and the diplomacy simply wasn't that good. They had no rules whatsoever to back up this claim. It was simply a rock hard conviction that magic HAD TO BE better than mundane means even when there was rules text describing the effects of both and making the mundane mean's superiority (when you got to the point you could use it) obvious.

Yeah, there's a bad GM problem as well as a game designer problem.

[Aimeryan]

...[Snip]...This got me thinking about Rogues and the odds of them actually being able to do their typical job, which is sneaking.

Level 4 Rogue vs MM Guard (https://roll20.net/compendium/dnd5e/Guard#content)

1st - Stealth vs Passive Perception: With +4 Dex and Expertise, a lv 4 Rogue has +8 Stealth, so against a Guard with 12 PPercep, he has 85% chance of success.

2nd - Initiative: Rogue has +4, Guard has +1, so Rogue wins 66% of time.

3rd - Attack roll vs AC: Rogue has +6 and Advantage against 16. He needs a 10, which he has an 80% chance of getting.

4th - Dealing enough damage: He deals 3d6+4 with a SA he has a nearly 91% chance of KOing the guard.

All of this seems like good enough odds, but not really. The actual odds of all this hapenning is 40.7%. There's nearly a 60% chance that something will go wrong and the Guard will alert everyone.

A different subclass would have slightly different odds (Swashbuckler will have better initiative, Assassin has more damage), but no subclass tips the odds significantly.

This is a Challenge 1/8 Guard having a better chance of doing his job than the level 4 Rogue, who is a literal expert at sneaking about. Completely ridiculous.

I agree with you that a common single guard (CR 1/8) should not have ~60% chance to sound the alert against a level 4 Rogue; that is pathetic. Not sure why you are getting so much push back on this, I feel you.

However, I would say the issue here is more 5e's surprise attacking - lets go through them;


1st - This seems fine.

2nd - Here, right here. There should be no Initiative roll - the guard shouldn't know to roll for Initiative. 5e screws up here, in my view - 3.5e had this right. Lets dismiss this for the math and see what happens.

3rd - Again, here. The guard should be flat-footed. Another thing 3.5e got right and 5e got wrong, as I see it. Lets take this up to 95% (nat 1 fail) and see what happens.

4th - Seems fine-ish. Maybe the guard has a little too much HP for being a common guard, however, we'll keep it like this.


Ok, so now the math works out at ~73.5% chance to succeed, or 26.5% chance to fail. Still a little high for a single common guard, but it is a lot better than ~60% chance to fail. If we take the HP down a little so that its 100% enough damage then the math works out at 80.75% chance to succeed, or 19.25% chance to fail.

Does this help other martials? Nope, but at least the Rogue could look somewhat competent at level appropriate stealth take-downs.

[heavyfuel]

2nd - Here, right here. There should be no Initiative roll - the guard shouldn't know to roll for Initiative. 5e screws up here, in my view - 3.5e had this right. Lets dismiss this for the math and see what happens.

Completely agree. I made a thread a while ago on how the I though the Surprise rules were really really poorly implemented in 5e, but, alas, I'm not the DM this time.

[porchdog]

Yet people still claimed charm person was better because it was a spell and the diplomacy simply wasn't that good. They had no rules whatsoever to back up this claim. It was simply a rock hard conviction that magic HAD TO BE better than mundane means even when there was rules text describing the effects of both and making the mundane mean's superiority (when you got to the point you could use it) obvious.

Yeah, there's a bad GM problem as well as a game designer problem.

This is my point. A great many DMs make magic spell the "Easy/Win Button" and make anything else the "Epic Fail Buzzer".

Sure, it's bad Dming, but that does not make it any less of a problem. And I'm talking more around just the DM judgment and abdication, I'm not really trying to compare rules.

Just think of any games you have been in: do both magic and mundane have the same abdication?

Example 2: the character want to climb up on a wall past two guards(and remember I'm only talking about adjudication, so don't worry about the dice roll for the climbing):

The spellcaster uses spider climb and moves right past the guards no problem.

The mundane uses athletics to climb...and the DM says "oh the guards look up and see your character".

The exact same action, but magic gets the free pass.

I guess there is one more part of this too, the DM guilt. The DM will feel bad and guilty if a player has a character "waste" a spell that does not work: so this is a big reason why some DMs give magic a free pass.

[Ignimortis]

Out of curiosity what should a high level rogue be able to do that a lower level rogue can't even attempt to do? Because it seems normal that without magic you can attempt to do something regardless of level and that level mainly impacts how successful you'll be.

If the game was fair to both casters and martials in fluff magnitude of actions they do, it would be something like this:


Level 1-4 Rogue: sneak by a few guards, steal some purses, open a normal lock without much difficulty, backstab a normal human, talk to people well
Level 5-10 Rogue: sneak through a non-alerted large camp/shadowstep between shadows not further apart than 30 feet, steal anything not bolted down or actively wielded, open a magical lock without much difficulty, backstab an orc general, talk to people and be able to persuade them of really outrageous-but-technically-believable lies or true things, evade minor detection spells (levels 1-2)
Level 11-16 Rogue: sneak through an active city at midday/shadowstep across a few hundred feet, steal things actually bolted down without removing the bolts, or armor/weapons off enemies, open any sort of lock, backstab a smaller dragon, talk people into anything, play riddles with dragons and win, evade magical mind-reading/detection
Level 17+ Rogue: sneak like Greater Invisibility, but True Seeing doesn't see you/shadowstep across the world, steal souls/luck/love/other concepts, open anything metaphysically similar to a door (including the gates between life and death to slip out while Hades/Hel/Anubis isn't watching), backstab a god, actually mind control people with honeyed words or successfully lie to gods, evade any sort of detection that isn't supported by someone being about as good with Perception

About the only thing Rogues actually mechanically do from that list is backstabbing, which does get progressively better (well, it keeps up mostly) with levels. Everything else remains in the 1-4 or 5-10 category at best.

[JNAProductions]

If the game was fair to both casters and martials in fluff magnitude of actions they do, it would be something like this:


Level 1-4 Rogue: sneak by a few guards, steal some purses, open a normal lock without much difficulty, backstab a normal human, talk to people well
Level 5-10 Rogue: sneak through a non-alerted large camp/shadowstep between shadows not further apart than 30 feet, steal anything not bolted down or actively wielded, open a magical lock without much difficulty, backstab an orc general, talk to people and be able to persuade them of really outrageous-but-technically-believable lies or true things, evade minor detection spells (levels 1-2)
Level 11-15 Rogue: sneak through an active city at midday/shadowstep across a few hundred feet, steal things actually bolted down without removing the bolts, or armor/weapons off enemies, open any sort of lock, backstab a smaller dragon, talk people into anything, play riddles with dragons and win, evade magical mind-reading/detection
Level 16+ Rogue: sneak like Greater Invisibility, but True Seeing doesn't see you/shadowstep across the world, steal souls/luck/love/other concepts, open anything metaphysically similar to a door (including the gates between life and death to slip out while Hades/Hel/Anubis isn't watching), backstab a god, actually mind control people with honeyed words or successfully lie to gods, evade any sort of detection that isn't supported by someone being about as good with Perception

About the only thing Rogues actually mechanically do from that list is backstabbing, which does get progressively better (well, it keeps up mostly) with levels. Everything else remains in the 1-4 or 5-10 category at best.

Minor nitpick-the Tiers of 5E go 1-4, 5-10, 11-16, 17-20.

So that'd work better in terms of breakpoints.

Otherwise, I do like your train of thought.

[Pex]

Did you read my proposed Sap feature? (https://forums.giantitp.com/showsinglepost.php?p=24610612&postcount=15) If so, what do you think of it? (I am sure that it has at least one loophole that need to be closed, since knocking an enemy unconscious makes them massively vulnerable due to Conditions ...)

When you position this like "I want an Easy Button to do This" we are maybe back to "Wait, you want an I Win button but don't want to expend a resource to do it?" Hence my idea on the "sap" feature. The spell caster expends a resource. That has an impact on the whole party, since that spell slot isn't there for what happens next.
Under the presumption of the six or so encounters for the adventure day, that's a non trivial cost.
No hold person when you need it
No lesser restoration when you need to unparalyze an ally
No misty step when you need it
No Phantasmal Force (in this case) to distract a foe ...


The Sap idea is fine. Works similar to Turn Undead. Granted I do want autosuccess as a player build choice with skill use, but putting that aside if there was a non-skill use button it's fine it's a resource and also ok not a guaranteed thing. Even spells have saving throws. Add this to the list of hypothetical 5.5E or 6E to give to rogues as stuff has been discussed to give fighters and barbarians.

Naturally spellcasters have to manage their resources. I do agree just because a spellcaster can use a spell to solve a problem doesn't mean he has to or should. In threads past the fuss was about the Knock spell instead of the rogue picking locks. However, for some people spending the spell slot resource is a non-issue. They're bothered that the spellcaster can, and that's all that matters.


The wizard doesn't bypass the guard with ease even with Invisibility. He has to be quiet as he moves past to not alert the guard, that's a straight up stealth roll vs his passive perception. Which for most wizards is going to be something like a 50/50 chance.

Also it's not an apples to apples comparison. A KOed guard is much different end result then an invisible wizard who has snuck past said guard. The odds of a wizard getting the same end result of a KOed guard without making any noise is actually probably much less then that of a Rogue.

Or Sleep, Hold Person, Phantasmal Force, even Charm Person if caught to at least convince the guard not to raise the alarm. The issue is the spellcaster only needs one roll to worry about, the saving throw, with defined rules to what happens where as the rogue has to roll many d20s against DM whim interpretations.

[Ignimortis]

Minor nitpick-the Tiers of 5E go 1-4, 5-10, 11-16, 17-20.

So that'd work better in terms of breakpoints.

Otherwise, I do like your train of thought.

Changed it for clarity, but...

Personally, I would probably do tiers as 1-4, 5-10, 11-14, and 15+ by the simple dint of those levels being meaningful for spell level progression. Level 3 spells are a major jump up for casters, as are level 6 spells, and level 8+ spells put godlike power into casters' hands. There are, of course, exceptions, but generally speaking, it sounds better to me than 5e tiers.

[OldTrees1]

If the game was fair to both casters and martials in fluff magnitude of actions they do, it would be something like this:


Level 1-4 Rogue: sneak by a few guards, steal some purses, open a normal lock without much difficulty, backstab a normal human, talk to people well
Level 5-10 Rogue: sneak through a non-alerted large camp/shadowstep between shadows not further apart than 30 feet, steal anything not bolted down or actively wielded, open a magical lock without much difficulty, backstab an orc general, talk to people and be able to persuade them of really outrageous-but-technically-believable lies or true things, evade minor detection spells (levels 1-2)
Level 11-16 Rogue: sneak through an active city at midday/shadowstep across a few hundred feet, steal things actually bolted down without removing the bolts, or armor/weapons off enemies, open any sort of lock, backstab a smaller dragon, talk people into anything, play riddles with dragons and win, evade magical mind-reading/detection
Level 17+ Rogue: sneak like Greater Invisibility, but True Seeing doesn't see you/shadowstep across the world, steal souls/luck/love/other concepts, open anything metaphysically similar to a door (including the gates between life and death to slip out while Hades/Hel/Anubis isn't watching), backstab a god, actually mind control people with honeyed words or successfully lie to gods, evade any sort of detection that isn't supported by someone being about as good with Perception

About the only thing Rogues actually mechanically do from that list is backstabbing, which does get progressively better (well, it keeps up mostly) with levels. Everything else remains in the 1-4 or 5-10 category at best.

I like this breakdown. Especially the detail.

[JNAProductions]

I like this breakdown. Especially the detail.

I too liked it! Which is why this (https://forums.giantitp.com/showthread.php?615835-Rogue-Mk-II&p=24613773#post24613773) now exists!

[Wizard_Lizard]

I mean to expand upon the idea of rogues not getting shiny new cool things, my wizard character only needed a one level dip in rogue, and boom, expertise, thieves cant, light armour, weapons. Same goes with fighter I guess, one level will get ya a lot of things like armour and weapons, and two levels in Paladin go a long way. Note that all of these things are really early on.

[Ignimortis]

I mean to expand upon the idea of rogues not getting shiny new cool things, my wizard character only needed a one level dip in rogue, and boom, expertise, thieves cant, light armour, weapons. Same goes with fighter I guess, one level will get ya a lot of things like armour and weapons, and two levels in Paladin go a long way. Note that all of these things are really early on.

Meanwhile dipping a caster isn't that impressive, yes, and dipping out of caster doesn't hurt you as much. There are several solutions to this:
1) Make multiclassing hurt more. How, though, is a good question. I don't know how to make it less powerful than now without making it unpleasant/annoying/plain bad.
2) Make class abilities come online way later, at level 3 or even 5 for things that make the class what it is. Also a poor solution because people want to play their character now, not three levels later.
3) Reinstate some barriers that meant that class features aren't as good as they seem. Medium Armor prof doesn't mean you can cast Wizard spells in it, unless you're specifically an Eldritch Knight, that sort of thing. Playtest 5e had a great idea about making weapon attack and spell attack separate stats, so a wizard 10 with 19 STR isn't as good at swinging a sword as a fighter 10 with 19 STR.

[MeimuHakurei]

Here's a simple behavioral guide that makes you realize the martial isn't screwed:
1. Always have three words on hand whenever somebody points out your shortcomings: Sustained, Reliable damage. These are the things you hold up against those filthy casters. Doesn't matter if you're built for defense or if you're consistently not in range. Also doesn't matter if your damage isn't high enough to take out the monsters before they down your party, either.

2. If you have full plate and shield, Defense Fighting Style and spend every combat action on Dodge, and your DM is attacking other party members, he is metagaming. Or if he incaps you with save-based effects. Keep pumping your AC and HP though - you're the tank, after all!
2a. If you're specced for offense, it's obviously not metagaming to identify the vulnerable targets and take them out.

3. Wizard, Bard, Druid, whatever - if it's a caster, they are there for one thing and one thing only: buff the everloving hell out of you. Even if they do something else, it's only for the sake of setting up the kill for you or to place you in attack range of the monsters. Should a spell ever be used for anything not directly benefitting you, that's an obvious sign of the caster being a selfish powergamer munchkin who only cares about winning.
3a. As a collorary to the above, your own abilities are sacrosanct. If you ever do anything out of combat, casters are never allowed to contribute in that area in any way. Obviating their abilities to fly, teleport or revive the dead (or whatever else they can do) is completely fine, however.

4. Mental stats are for chumps. Don't give in to the urge to invest into them for out of combat skills. Either you insist that you can use your strength without issue, or if that doesn't work, insist that the DM is trying to screw you over. That is, if you actually care about doing crap out of combat. You can just browse your favourite social media until the DM says "roll initiative" instead.

5. An ally falling in battle is a clear sign that they are terrible at the game. It was definitely them going nova and not having anything for the remaining fights of the day. Or they didn't let you tank (ignore the fact that you don't actually draw in attacks). However, if you're the one going down, it must be because your allies refused to buff/heal you, or because the DM went out of their way to screw you over. If you're a rogue/monk and don't want to tank for the team, an ally's death is from them overextending or picking a fight with something you're obviously supposed to run from.

6. Ignore how powerful the higher level spells are - no campaign ever goes that far. No, that doesn't mean you won't get Reliable Talent, Stroke of Luck, Survivor, four attacks, Persistent Rage, Primal Champion, Diamond Soul or Empty Body - you should definitely use them frequently in whatever game you play.

[Sorinth]

Wait. So your argument is that, because in a party of 3 people, no one decided to play Druid, then the Druid isn't an amazing class?

Really?

That's the most ridiculous argument I've heard. Your sample size is one single group out of countless groups.

No, the point I was making is that just because there exists a magic solution to a problem doesn't mean that it is actually available in real gameplay. So in actual gameplay martials aren't screwed because magic users will often not have the spell/ability and so end up relying on martials.

[Willie the Duck]

I agree with you that a common single guard (CR 1/8) should not have ~60% chance to sound the alert against a level 4 Rogue; that is pathetic. Not sure why you are getting so much push back on this, I feel you.

I don't think many people disagree with this statement as a specific issue about what a 4th level rogue should be able to do. I think the pushback is with regards to how much it proves about the system in general. At least for me, this specific instance is a bad flooring issue, not dissimilar to the every-edition situation of a guy at 1 hp being able to be killed by a rat/housecat (because the system can't handle damages below 1). Every creature, regardless of CR, gets an initiative roll, an armor class, and the ability to alert other local monsters. Does that suck?, should there perhaps be further guidance on when a situation should be a 'shoe-in' (i.e. "don't bother rolling, of course your rogue can take out one guard before they can raise the alarm")?, and does the system favor casters? Yes to all. That does not mean that this case is a indicator of the initial premise.


Naturally spellcasters have to manage their resources. I do agree just because a spellcaster can use a spell to solve a problem doesn't mean he has to or should. In threads past the fuss was about the Knock spell instead of the rogue picking locks. However, for some people spending the spell slot resource is a non-issue. They're bothered that the spellcaster can, and that's all that matters.

Certainly a reasonable position to have. That spells don't have a native success roll is an artefact of how the game evolved without a specific rhyme or reason (although, to be fair, bitd the number and strength of the limitations such as spells per day, spell fizzling, magic resistance, and so on used to be higher). 4E's knock -- where IIRC the wizard effectively got to make a Int-based instead of Dex-based pick locks check and nothing more -- is a good model for how it could be done differently.



Ignore how powerful the higher level spells are - no campaign ever goes that far. No, that doesn't mean you won't get Reliable Talent, Stroke of Luck, Survivor, four attacks, Persistent Rage, Primal Champion, Diamond Soul or Empty Body - you should definitely use them frequently in whatever game you play.

This is the second time I've seen this, yet I have not seen the actual instance of someone doing so such that it deserves derision. So far as I can tell, people have been very clear that Reliable Talent comes online at the same time as L6+ spells.

[Sorinth]

Or Sleep, Hold Person, Phantasmal Force, even Charm Person if caught to at least convince the guard not to raise the alarm. The issue is the spellcaster only needs one roll to worry about, the saving throw, with defined rules to what happens where as the rogue has to roll many d20s against DM whim interpretations.

Except that's not true.

The spellcaster still needs to stealth because they have to get within range of the spell quietly. That's a straight roll with a likely +2 bonus against a DC of 12, so 50% chance.
The spellcaster still needs to roll initiative because if they fail to beat the guard, the guard will hear the spellcaster start casting and raise the alarm. Again at a +2 vs +1 he's barely above 50%
And finally there is the saving throw assuming DC 14, so 70% chance.

All together the spellcaster is probably at something like 20% chance of succeeding. So the Rogue is twice as likely to succeed as the spellcaster.

It's ironic that people claim martials need the DM to roll out the red carpet in order to be competitive when that's exactly what the DM is doing for the spellcaster if he lets them boil it down to one roll.

[Sorinth]

Have you tried to hide in bright light? I expect most rogues can by mid level.

Have you ever snuck up on a Dragon? Blindsight is nice and all, but the mid level Rogue understands how it works and how to hide from it.

As for high level Rogues, who even knows you exist? You are out of sight and out of mind. Traces of your presence have been disguised to be coincidences or the actions of your comrades. You can fool even those using divination by leading their assumptions elsewhere.

The main insight is that D&D can and has treated skills as EXtraordinary, not mundane. And the best way to do that is to have progression.


If the game was fair to both casters and martials in fluff magnitude of actions they do, it would be something like this:


Level 1-4 Rogue: sneak by a few guards, steal some purses, open a normal lock without much difficulty, backstab a normal human, talk to people well
Level 5-10 Rogue: sneak through a non-alerted large camp/shadowstep between shadows not further apart than 30 feet, steal anything not bolted down or actively wielded, open a magical lock without much difficulty, backstab an orc general, talk to people and be able to persuade them of really outrageous-but-technically-believable lies or true things, evade minor detection spells (levels 1-2)
Level 11-16 Rogue: sneak through an active city at midday/shadowstep across a few hundred feet, steal things actually bolted down without removing the bolts, or armor/weapons off enemies, open any sort of lock, backstab a smaller dragon, talk people into anything, play riddles with dragons and win, evade magical mind-reading/detection
Level 17+ Rogue: sneak like Greater Invisibility, but True Seeing doesn't see you/shadowstep across the world, steal souls/luck/love/other concepts, open anything metaphysically similar to a door (including the gates between life and death to slip out while Hades/Hel/Anubis isn't watching), backstab a god, actually mind control people with honeyed words or successfully lie to gods, evade any sort of detection that isn't supported by someone being about as good with Perception

About the only thing Rogues actually mechanically do from that list is backstabbing, which does get progressively better (well, it keeps up mostly) with levels. Everything else remains in the 1-4 or 5-10 category at best.

Hiding in the open, shadow step, greater invisibility, protection from divination magic, aren't all these things just adding magic?

[Xervous]

Hiding in the open, shadow step, greater invisibility, protection from divination magic, aren't all these things just adding magic?

Only if you define them as such. Or does that sound too much like weeaboo magic for you to accept that there can be nonmagical, but nonetheless extraordinary means for accomplishing such feats?

[Eldariel]

Hiding in the open, shadow step, greater invisibility, protection from divination magic, aren't all these things just adding magic?

Only one of those that's probably obviously magic is shadow step, and even there that depends. With certain ontology of shadows (plane of shadows and all that), this is simply slipping through gaps in dimensions.

Protection from divination is simply knowing how the magic works and having mental and physical skills to subvert it.

Hiding in the open is just extraordinary skill and perhaps camouflage/slipping into the environment/whatever. Very roguey anyways. Same goes for "greater invisibility" (you are so good at being inconspicuous/stealthy/disguised/canouflaged that you can only be noticed if you want to).

[Ignimortis]

Hiding in the open, shadow step, greater invisibility, protection from divination magic, aren't all these things just adding magic?

Unless you insist that everything supernatural/unrealistic can only be done by magic, no. If you're saying that every one of those effects can be simulated by magic - then yes, but that's because D&D magic is so broad it includes everything, plus the kitchen sink, and the kitchen sink's Ice Assassin Shadow Simulacrum for good measure.

[Aimeryan]

I don't think many people disagree with this statement as a specific issue about what a 4th level rogue should be able to do. I think the pushback is with regards to how much it proves about the system in general. At least for me, this specific instance is a bad flooring issue, not dissimilar to the every-edition situation of a guy at 1 hp being able to be killed by a rat/housecat (because the system can't handle damages below 1). Every creature, regardless of CR, gets an initiative roll, an armor class, and the ability to alert other local monsters. Does that suck?, should there perhaps be further guidance on when a situation should be a 'shoe-in' (i.e. "don't bother rolling, of course your rogue can take out one guard before they can raise the alarm")?, and does the system favor casters? Yes to all. That does not mean that this case is a indicator of the initial premise.

I think I agree and somewhat partially disagree;

Agree: The system is not designed to cater to the issue. This, I would say, applies to pretty much all of martial 2/3 pillar support. The system, instead, tries to push that on to the DM to solve. Mean while, it does give 3/3 pillar support to casters.

Partially disagree: The specific case here with the Rogue, your solution is to apply a patch to cover it. This I don't fully disagree with (at times that is the best solution), however, the solution I would prefer for the specific case here is to go back to 3.5e Surprise mechanics, which had better results for this sort of thing and made more sense. In 5e's attempt to simplify things it has once again gone for less-gives-less rather than less-gives-more.

[Necrosnoop110]

3) Part of the problem is the independent problem caused by erroneously design ability checks using bounded accuracy's attack & damage model without providing the second half. This significantly shrinks the amount of progression available in that system. Combined with the anchoring effect of DC grades, WotC adventures' example DCs, and the tendency to start at 1st level, this readily results in DMs making ability checks risky maneuvers for non level appropriate results by Tier 2. But DMs can and do patch this issue since there are numerous ways to do so.

Can someone help me unpack this, I'm having trouble understanding it. And what would be these numerous solutions to that problem be?

[Xervous]

Can someone help me unpack this, I'm having trouble understanding it. And what would be these numerous solutions to that problem be?

Combat is a bar you’re filling through successive rolls. It scales across levels in the rate at which you can fill the bar (do damage). Bounded accuracy on the attack roll means your chance of an individual attempt at contributing wont change much, but you scale in other ways to contribute more on each success, and/or get more attempts. You expect to face bigger bars over time, and as you grow more powerful the smaller bars are easier to deal with.

Skills are a yes/no roll, end of story. They operate under bounded accuracy but the magnitude of the effect nor the number of attempts allowed never scale.

[heavyfuel]

No, the point I was making is that just because there exists a magic solution to a problem doesn't mean that it is actually available in real gameplay. So in actual gameplay martials aren't screwed because magic users will often not have the spell/ability and so end up relying on martials.

So, your point is that because people don't always play a super optimized party of spellcasters, then there's no problem at all?


Can someone help me unpack this, I'm having trouble understanding it. And what would be these numerous solutions to that problem be?

I *think* what OldTrees mean is that Bounded Accuracy makes it so that you can never truly progress in your skills. An expected modifier at lv 1 is +5, and expected modifier at lv 20 is +11. It's not a huge difference, and it's certainly possible to have +11 at level 1.

[Composer99]

Partially disagree: The specific case here with the Rogue, your solution is to apply a patch to cover it. This I don't fully disagree with (at times that is the best solution), however, the solution I would prefer for the specific case here is to go back to 3.5e Surprise mechanics, which had better results for this sort of thing and made more sense. In 5e's attempt to simplify things it has once again gone for less-gives-less rather than less-gives-more.

I just read through the 3.5 srd rules on surprise, and have to disagree that they're better or make more sense. By my reading they have an almost identical effect except are more clumsily worded.

5e is a much cleaner implementation. Where I think the 5e surprise rules could be patched is denying surprised creatures the chance to take reactions until their second turn (that is, the turn in which they are able to take actions). In a nod to the old flat-footed condition, I would also give rogues the ability to sneak attack a surprised creature.

[Aimeryan]

I just read through the 3.5 srd rules on surprise, and have to disagree that they're better or make more sense. By my reading they have an almost identical effect except are more clumsily worded.

5e is a much cleaner implementation. Where I think the 5e surprise rules could be patched is denying surprised creatures the chance to take reactions until their second turn (that is, the turn in which they are able to take actions). In a nod to the old flat-footed condition, I would also give rogues the ability to sneak attack a surprised creature.

Just in case you have an incorrect source, here are the 3.5e and 5e Surprise rules:

The Surprise Round: If some but not all of the combatants are aware of their opponents, a surprise round happens before regular rounds begin. Any combatants aware of the opponents can act in the surprise round, so they roll for initiative. In initiative order (highest to lowest), combatants who started the battle aware of their opponents each take a standard action during the surprise round. You can also take free actions during the surprise round. If no one or everyone is surprised, no surprise round occurs.

Unaware Combatants: Combatants who are unaware at the start of battle don’t get to act in the surprise round. Unaware combatants are flat-footed because they have not acted yet, so they lose any Dexterity bonus to AC.
If you're surprised, you can't move or take an action on your first turn of the combat, and you can't take a reaction until that turn ends.

They are quite similar, except the parts you pointed out - which is the exact point:


In 5e, Initiative is rolled before the surprise is sprung, which makes no sense; if the attacker is yet to do anything and has remained undetected, what is the surprised they taking initiative on??
In 5e, the surprised can take a reaction on the attacker's turn. In 3.5e, they are surprised; they don't get to react.
In 5e, the surprised apply their dexterity to avoiding the hit. In 3.5e, they are surprised; they don't get to use their dexterity.

[porchdog]

Or Sleep, Hold Person, Phantasmal Force, even Charm Person if caught to at least convince the guard not to raise the alarm. The issue is the spellcaster only needs one roll to worry about, the saving throw, with defined rules to what happens where as the rogue has to roll many d20s against DM whim interpretations.

This is the Big Problem. The spellcaster uses a spell and only needs to worry about the save roll...AND the DM rolls out the Red Carpet and alters everything in the game to make the spell work and everything be alright. The DM whim is present here, except it's "magic always works in just the way the player wants it too".

This even goes as far as the set up: The guard at the door is a human, an easy target for a spell like hold person. If the door guard is a fey, celestial, fiend, construct, elemental or undead; then suddenly hold person won't work anymore. And, in fact, plenty of such monsters are perfect guards: Awakened Shrubs, Awakened Trees, Animated Armors, flying swords and skeletons. Note the spellcaster can't just sleep or hold or charm most of these guard monsters....but note the martial can still act against each one.




It's ironic that people claim martials need the DM to roll out the red carpet in order to be competitive when that's exactly what the DM is doing for the spellcaster if he lets them boil it down to one roll.

Truer words were never posted.

[Ignimortis]

This is the Big Problem. The spellcaster uses a spell and only needs to worry about the save roll...AND the DM rolls out the Red Carpet and alters everything in the game to make the spell work and everything be alright. The DM whim is present here, except it's "magic always works in just the way the player wants it too".

This even goes as far as the set up: The guard at the door is a human, an easy target for a spell like hold person. If the door guard is a fey, celestial, fiend, construct, elemental or undead; then suddenly hold person won't work anymore. And, in fact, plenty of such monsters are perfect guards: Awakened Shrubs, Awakened Trees, Animated Armors, flying swords and skeletons. Note the spellcaster can't just sleep or hold or charm most of these guard monsters....but note the martial can still act against each one.


And most DMs don't have level 2 characters acting against fiends, constructs and fey, well, at level 2. Furthermore, a lot of DMs (not me, but I know quite a few) tend to populate their worlds mostly with humanoids, and make non-humanoid threats rarer - well, aside from the nicely generic low-level undead, I suppose. I can't call that rolling out the red carpet, since nothing is actually altered, the game works as expected when the spellcaster does their thing, and nobody catered to them specifically.

[Sorinth]

So, your point is that because people don't always play a super optimized party of spellcasters, then there's no problem at all?

Even a super optimized party of spellcasters are limited by spells know/prepared and so won't always have the spell needed at the ready.

[Frozenstep]

Except that's not true.

The spellcaster still needs to stealth because they have to get within range of the spell quietly. That's a straight roll with a likely +2 bonus against a DC of 12, so 50% chance.
The spellcaster still needs to roll initiative because if they fail to beat the guard, the guard will hear the spellcaster start casting and raise the alarm. Again at a +2 vs +1 he's barely above 50%
And finally there is the saving throw assuming DC 14, so 70% chance.

All together the spellcaster is probably at something like 20% chance of succeeding. So the Rogue is twice as likely to succeed as the spellcaster.

It's ironic that people claim martials need the DM to roll out the red carpet in order to be competitive when that's exactly what the DM is doing for the spellcaster if he lets them boil it down to one roll.

The sleep spell has a range of 90 feet (110 feet is the furthest you can get the AoE to reach). It's pretty hard to sell a guard really having much chance at detecting someone trying to move quietly at that kind of distance in the dark (especially if there is any other sound in the area, like other guards). Hold person is much more reasonable, at 60 feet.

I don't remember any rules on how loud verbal components need to be. This is something the books really should have fleshed out. They also could have fleshed out how recognizable spells are...because is your average guard, hearing a voice that isn't a whisper but isn't loud either from a long distance going to be able to identify it as a spell being cast within 6 seconds, rather then the sound of another guard talking somewhere he can't see?

Depending on how you run your table, maybe the DM is strict on stealth and has very alert guards, which maybe favors the martial. But also a few reasonable lines of logic can swing it around. Honestly though, the whole guards+cave example is actually favorable for martials to begin with, given the implication of multiple guards to deal with and the low level it takes place at, and the strictness of the situation (not killing the guards).

[Xervous]

This is the Big Problem. The spellcaster uses a spell and only needs to worry about the save roll...AND the DM rolls out the Red Carpet and alters everything in the game to make the spell work and everything be alright. The DM whim is present here, except it's "magic always works in just the way the player wants it too".

This even goes as far as the set up: The guard at the door is a human, an easy target for a spell like hold person. If the door guard is a fey, celestial, fiend, construct, elemental or undead; then suddenly hold person won't work anymore. And, in fact, plenty of such monsters are perfect guards: Awakened Shrubs, Awakened Trees, Animated Armors, flying swords and skeletons. Note the spellcaster can't just sleep or hold or charm most of these guard monsters....but note the martial can still act against each one.




Truer words were never posted.

I fail to see how the simple fact of a guard being susceptible to a narrow scope spell like hold person is a GM rolling out the red carpet.

As you’ve so conveniently brought us back to the carpet analogy do note the context including the part about putting walls in front of casters.

The latter portion of your post supports the red carpeted martial criticism as you are pointing out that the GM needs to actively tailor things to deny casters.

Again, please shelve the distraction of GMs favoring magic. We are not discussing bad GMs here, we are discussing whether or not the system provides a good basis for Martials to interact with the world assuming a neutral GM guided by said system.

[Sorinth]

The sleep spell has a range of 90 feet (110 feet is the furthest you can get the AoE to reach). It's pretty hard to sell a guard really having much chance at detecting someone trying to move quietly at that kind of distance in the dark (especially if there is any other sound in the area, like other guards). Hold person is much more reasonable, at 60 feet.

I don't remember any rules on how loud verbal components need to be. This is something the books really should have fleshed out. They also could have fleshed out how recognizable spells are...because is your average guard, hearing a voice that isn't a whisper but isn't loud either from a long distance going to be able to identify it as a spell being cast within 6 seconds, rather then the sound of another guard talking somewhere he can't see?

Depending on how you run your table, maybe the DM is strict on stealth and has very alert guards, which maybe favors the martial. But also a few reasonable lines of logic can swing it around. Honestly though, the whole guards+cave example is actually favorable for martials to begin with, given the implication of multiple guards to deal with and the low level it takes place at, and the strictness of the situation (not killing the guards).

You're right that if the caster can whisper then it's unlikely to alert the guard, however how is that not a case of the DM rolling out the red carpet for for casters? RAW it should alert everyone to your presence as you are "chanting mystic words". And in a world where magic is a thing pretty much everyone is going to recognize that weird chanting is probably part of a spell.

Hearing movement 90-110 ft when everything is supposed to be quiet isn't difficult unless there is ambient noise. But I'll grant you that a failed stealth at that range shouldn't necessarily cause the guard to raise the alarm but he should become suspicious at the least.

And for the record Sleep has the side effect of making a lot of noise when the armoured guard falls to the ground. And you also better hope the guard isn't an Elf or Half-Elf, which might be hard to tell when you decide to use the spell.

[Sorinth]

I fail to see how the simple fact of a guard being susceptible to a narrow scope spell like hold person is a GM rolling out the red carpet.

As you’ve so conveniently brought us back to the carpet analogy do note the context including the part about putting walls in front of casters.

The latter portion of your post supports the red carpeted martial criticism as you are pointing out that the GM needs to actively tailor things to deny casters.

Again, please shelve the distraction of GMs favoring magic. We are not discussing bad GMs here, we are discussing whether or not the system provides a good basis for Martials to interact with the world assuming a neutral GM guided by said system.

The balance between martials and casters is based on the idea that you will encounter a variety of challenges. The martials are supposed to provide that steady/reliable ability whereas the casters will vary between great and not so great. Pointing out that the caster tactics are limited isn't a case of actively tailoring things to deny casters, it's establishing the fact that the caster isn't as reliable as some assume it to be.

[MeimuHakurei]

How to DM for casters:
1. Never include any humanoid enemies. Casters will run away with the game using Charm Person, Hold Person and similar.

2. Throw everything and the kitchen sink at a party trying to rest before 6 to 8 encounters happened, no matter how peaceful and quiet the situation is and how well-guarded they are against monsters. If you cannot make it work, instead make all NPCs in the world hate them for the audacity to take a day off.

3. All campaign places must be fully accessible via foot travel or, at the very least, services that cost little to nothing for the party. It's unfair and unreasonable for DMs to have locales underwater, in the air or on other planes of existence.

4. All enemies are to have either regular magic resistance, legendary resistance or both. After all, we can't have enemies get incapacitated. The only valid interaction with monsters is hit point damage.And don't even think about letting casters use Invisibility, Misty Step and their ilk to evade an encounter entirely!

5. Everyone hates magic. No matter the situation, a caster must be exterminated if they are ever seen casting spells or doing anything that looks that way. Of course, spellcasting NPCs you run yourself are exempt from this rule.

6. All encounters must include at least one spellcaster capable of casting Counterspell. No getting cheeky with trying to buff past the encounter math.

7. Every dungeon and major villain location is fully covered in Forbiddance zones or dead magic zones. Likewise, every named NPC is mindblanked and beyond the knowledge of higher entities. Knowing about your enemies is metagaming.

8. If all of the above fails and casters are still contributing, it's because they're powergamer munchkins who only care about winning. Let rocks fall sort it out.

[Man_Over_Game]

Just in case you have an incorrect source, here are the 3.5e and 5e Surprise rules:

The Surprise Round: If some but not all of the combatants are aware of their opponents, a surprise round happens before regular rounds begin. Any combatants aware of the opponents can act in the surprise round, so they roll for initiative. In initiative order (highest to lowest), combatants who started the battle aware of their opponents each take a standard action during the surprise round. You can also take free actions during the surprise round. If no one or everyone is surprised, no surprise round occurs.

Unaware Combatants: Combatants who are unaware at the start of battle don’t get to act in the surprise round. Unaware combatants are flat-footed because they have not acted yet, so they lose any Dexterity bonus to AC.
If you're surprised, you can't move or take an action on your first turn of the combat, and you can't take a reaction until that turn ends.

They are quite similar, except the parts you pointed out - which is the exact point:


In 5e, Initiative is rolled before the surprise is sprung, which makes no sense; if the attacker is yet to do anything and has remained undetected, what is the surprised they taking initiative on??
In 5e, the surprised can take a reaction on the attacker's turn. In 3.5e, they are surprised; they don't get to react.
In 5e, the surprised apply their dexterity to avoiding the hit. In 3.5e, they are surprised; they don't get to use their dexterity.



Why are you rolling initiative if nobody initiated it?
The Defender only gets to use their Reaction if their Initiative is higher than the Offender. They still don't get to act any sooner, they just get to use a Reaction (if one is relevant to the situation). Someone who is faster should be able to use their Reaction better than someone else who was slower and surprised, no?
That's more of a difference between the AC rules, not really a direct reflection of Surprise rules. Unless you want to start doing Touch AC again. They took that out because simplicity was more valuable than realism.


Although, small correction on bullet 1, the order goes:


Determine surprise.
Establish positions.
Roll initiative.


As a result, Initiative should never be rolled until after the Defenders notice a threat, or when the Offender becomes one.

[Xervous]

The balance between martials and casters is based on the idea that you will encounter a variety of challenges. The martials are supposed to provide that steady/reliable ability whereas the casters will vary between great and not so great. Pointing out that the caster tactics are limited isn't a case of actively tailoring things to deny casters, it's establishing the fact that the caster isn't as reliable as some assume it to be.

1. The assumption of balance
2. Supposing is an assumption
3. The sky is blue, everyone’s tactics are limited. You need hard quantitative evidence to convince others of your position that casters are comparably or more limited than Martials in absence of GM bias.

[Frozenstep]

You're right that if the caster can whisper then it's unlikely to alert the guard, however how is that not a case of the DM rolling out the red carpet for for casters? RAW it should alert everyone to your presence as you are "chanting mystic words". And in a world where magic is a thing pretty much everyone is going to recognize that weird chanting is probably part of a spell.

Hearing movement 90-110 ft when everything is supposed to be quiet isn't difficult unless there is ambient noise. But I'll grant you that a failed stealth at that range shouldn't necessarily cause the guard to raise the alarm but he should become suspicious at the least.

And for the record Sleep has the side effect of making a lot of noise when the armoured guard falls to the ground. And you also better hope the guard isn't an Elf or Half-Elf, which might be hard to tell when you decide to use the spell.

How quietly can you chant? If it isn't very loud, how recognizable is it from 110 feet away? As far as the book tells you, you just need to clearly get certain tones right, so I don't think a whisper would work, but a quiet voice might. Depending on the game, that guard might have never have actually seen magic being cast before, or be able to identify it instantly.

It's pretty difficult to imagine there being no ambient noise. The guard can become suspicious all he wants, unless he's skittish enough to yell at the first sign of maybe an animal or another guard making a tiny sound more then a hundred feet away (rather then, you know, trying to approach so his lantern can light it up, like a normal person would), he'll be asleep before he understands what's happening.

Is it any more noise than a rogue slamming something into the back of the guards head? Maybe it is, but is the situation such that the noise a rogue makes knocking this guard out isn't enough to cause alarm, but the armor falling is?

And now we're hoping the guard is at least partly elven?

Every DM is different, but it's pretty easy to create situations that are easier and less risky to solve with magic even if they didn't mean to. They aren't rolling out the red carpet, they just presented a scenario, then came to reasonable rulings (you don't have to shout spells for everyone to hear, guard doesn't instantly scream upon hearing a quiet voice from more then 100 feet away) and it just happens sometimes.

[heavyfuel]

Even a super optimized party of spellcasters are limited by spells know/prepared and so won't always have the spell needed at the ready.

My bad, I didn't know druids had to prepare Wild Shape

[KorvinStarmast]

This thread has me thinking: was "legendary saves" an attempt to balance casters versus martials?

[Frozenstep]

This thread has me thinking: was "legendary saves" an attempt to balance casters versus martials?

That might be part of why they exist, but even non-casters have plenty of ways to completely cripple a single enemy if they fail a saving throw (most famously, monk stunning strike). They're more about making sure bosses at least get a few turns of non-restrained actions to actually threaten the party.

[Man_Over_Game]

This thread has me thinking: was "legendary saves" an attempt to balance casters versus martials?

I think that was more about giving bosses the means of shedding off cheap single target effects that cripple the target, like Hideous Laughter or Hold Person.

Given, most of these are spells (the Monk's Stunning Strike being literally the only notable nonmagic exception I can think of), but otherwise Bosses would be rendered useless from level 1-2 spells. Part of this is because Bosses have Legendary Actions that would otherwise be lost if they lost a round. Normally a spell like that stops 1 turn per round, but each Legendary Action could be worth half of a turn, so crippling a boss for just a single round could mean you could remove roughly 3 turns' worth of high value actions with a single action and a level 1 spell.

It's also a lot better than just increasing save bonuses or magic resistance, as Legendary Resistance at least provides a future reward for the investment you spent (similar to how damage doesn't do anything until the creature hits 0 HP).

Personally, though, I like the idea of afflicting the boss with Exhaustion to force a Save instead, as it is a lot more gradual and is a lot easier to telegraph to your players than some pseudo-Counterspell effect.

[Pex]

This is the Big Problem. The spellcaster uses a spell and only needs to worry about the save roll...AND the DM rolls out the Red Carpet and alters everything in the game to make the spell work and everything be alright. The DM whim is present here, except it's "magic always works in just the way the player wants it too".

Only because the DM doesn't have to make anything up. The spell description says what happens and the DC is specified how it's determined. For skill use the DM makes it up. Your example earlier of a wizard using Spider Climb was bad form. It's not the simple case the guard autofails to see the wizard because he used the spell where the guard autosucceeds seeing the rogue because he climbed. The problem is wizard casts Spider Climb so autosucceeds on climbing the wall and out of sight having rolled no dice. The rogue has to roll to climb. If he succeeds on an Athletics check of DC whatever the DM feels likes he gets to climb to the top of the wall out of sight like the wizard did who didn't have to roll a die. His success chance is totally dependent on DM whim of how difficult the DM thinks it is to climb a wall. If he fails the check then it's DM whim again. Maybe the guard autosucceeds seeing the rogue, Maybe the DM will allow the rogue a chance not to be detected by rolling a stealth check. The DC for the stealth check is DM whim. It could be against the passive Perception of the guard or the DM may decide to give the guard a Perception roll to make this an opposed roll check.


This even goes as far as the set up: The guard at the door is a human, an easy target for a spell like hold person. If the door guard is a fey, celestial, fiend, construct, elemental or undead; then suddenly hold person won't work anymore. And, in fact, plenty of such monsters are perfect guards: Awakened Shrubs, Awakened Trees, Animated Armors, flying swords and skeletons. Note the spellcaster can't just sleep or hold or charm most of these guard monsters....but note the martial can still act against each one.



Right, so the spellcaster uses another spell if he prepared it. It's possible a spellcaster won't have a needed spell. Point goes to the martials. Hooray. However, Hold Person is not the spellcaster's only spell. It behooves a spellcaster player to diversify, so while there's no guarantee the spellcaster will have the perfect spell he can have good enough. If just walking past a guard he could cast Silent Image of the passageway to make it look like it remains empty. There we have some DM whim. Maybe it's autosuccess to walk by. Maybe it's a Stealth check at Advantage. There's still Phantasmal Force or even Minor Illusion for sound as a distraction to get Advantage. It is only a Cantrip so not expecting autosuccess. However, we're talking low level here. By the time Awakened Shrubs and flying swords become suitable encounter guards we're talking higher level and spellcasters have access to better spells. Dispel Magic becomes available. Banishment. Dimension Door. It costs a more valuable resource, but if getting past the uberguard undetected is so important spending the resource to do so is the cost of doing business.

[Man_Over_Game]

Only because the DM doesn't have to make anything up. The spell description says what happens and the DC is specified how it's determined. For skill use the DM makes it up. Your example earlier of a wizard using Spider Climb was bad form. It's not the simple case the guard autofails to see the wizard because he used the spell where the guard autosucceeds seeing the rogue because he climbed. The problem is wizard casts Spider Climb so autosucceeds on climbing the wall and out of sight having rolled no dice. The rogue has to roll to climb. If he succeeds on an Athletics check of DC whatever the DM feels likes he gets to climb to the top of the wall out of sight like the wizard did who didn't have to roll a die. His success chance is totally dependent on DM whim of how difficult the DM thinks it is to climb a wall. If he fails the check then it's DM whim again. Maybe the guard autosucceeds seeing the rogue, Maybe the DM will allow the rogue a chance not to be detected by rolling a stealth check. The DC for the stealth check is DM whim. It could be against the passive Perception of the guard or the DM may decide to give the guard a Perception roll to make this an opposed roll check.

There's this natural mentality that you need permission to succeed, but this becomes really problematic when you ask for multiple rolls.

On one hand, the DM wants you to be able to utilize all 3 of your skills that are relevant to this check, as you've invested into them all. But each die isn't increasing your chance to succeed, but your chance to fail. Even with a 70% chance to succeed on each of those 3 checks, that becomes a 34.3% chance of success when combined.

Heck, this is basically a mechanical, quantified version of the "Guy At The Gym" Fallacy, as the game becomes more difficult the more realism you try to pump into it.

The only solutions I can really think of is to assume there is only one means of failure for any given attempt (so you'd roll either an Athletics, Stealth, or Acrobatics, but assume success on the skills you didn't roll for), allow multiple rolls to only increase success (such as using the highest of the 3 rolls), lower the DC by 5 for every new skill roll you ask for (or something comparable to lower the DC when using multiple checks for failure), or come up with some major change to the skill system.

[Kane0]

The only solutions I can really think of is to assume there is only one means of failure for any given attempt (so you'd roll either an Athletics, Stealth, or Acrobatics, but assume success on the skills you didn't roll for), allow multiple rolls to only increase success (such as using the highest of the 3 rolls), lower the DC by 5 for every new skill roll you ask for (or something comparable to lower the DC when using multiple checks for failure), or come up with some major change to the skill system.

And if that’s just off the top of your head then there really is a lot of room for improvement.

[prabe]

The only solutions I can really think of is to assume there is only one means of failure for any given attempt (so you'd roll either an Athletics, Stealth, or Acrobatics, but assume success on the skills you didn't roll for), allow multiple rolls to only increase success (such as using the highest of the 3 rolls), lower the DC by 5 for every new skill roll you ask for (or something comparable to lower the DC when using multiple checks for failure), or come up with some major change to the skill system.

I'd be inclined to decide which Skills were relevant, then let the player pick a skill from those to be rolled (in case they have Expertise in something) and let them have an extra die (a la Advantage) for every other relevant skill they have Proficiency in. I kinda do something like that now, sometimes, but nothing in the lines of this specific scenario has arisen. It would at least give extra Skills a way to increase chances for success, rather than introducing more opportunities for failure.

But as you say, that's probably going well outside the RAW.

[OldTrees1]

3) Part of the problem is the independent problem caused by erroneously design ability checks using bounded accuracy's attack & damage model without providing the second half. This significantly shrinks the amount of progression available in that system. Combined with the anchoring effect of DC grades, WotC adventures' example DCs, and the tendency to start at 1st level, this readily results in DMs making ability checks risky maneuvers for non level appropriate results by Tier 2. But DMs can and do patch this issue since there are numerous ways to do so.


Can someone help me unpack this, I'm having trouble understanding it. And what would be these numerous solutions to that problem be?

Certainly.

I will first describe progression and windows by using encounters as an example:
An Nth level party has a range of encounters that challenge them. This spans from the ones just above guaranteed victory to those just under guaranteed defeat. This span is a window. At different levels the window has different mins and maxs. As the party levels from 1st to 20th, that window will continue to move towards harder encounters. I use the term "progression" to describe how much that window moves. Some games have 0 progression because the window never moves. Other games have the window move a bit but always have first and last level overlap a bit. D&D's default narrative has a rather large progression, the gap between the 1st level window and the 20th level window could be measured in multiple window lengths. This concept of windows and progression applies to more than just encounters. It applies to many subsystems in D&D. Attacks and Ability Checks are two examples.

How does bounded accuracy work:
Attacks are comprised of an accuracy or "attack" roll and a damage roll. Bounded Accuracy is the idea of minimizing the progression in the attack roll without compromising the progression of attacks themselves. They handle this by increasing the progression of the damage roll while shrinking the progression of the attack roll. Basically a 20th level attack action is still much stronger than a 1st level attack action, but both can hit the same target.

Why is this relevant to ability checks (including skill checks):
In 5E the progression of the accuracy of ability checks was also shrunk. However what got increased to maintain the overall progression of ability checks? Nothing. There was no second magnitude roll for WotC to increase. As a result the progression of ability checks shrunk massively out of a simple design error.

What do I mean by anchoring?

Games tend to start at low levels. This means DMs are more likely to assign DCs based on preventing a 1st level character from achieving it too soon rather than on when a character will master (get to 100%) that challenge. If we imagine a reasonable amount of progression, then this bias would have us set the DCs too high for mid-high level characters because the 5E progression was shrunk.
The DC grades (5, 10, 15, 20, 25, 30) have names that encourage you to start with DC 15 as your baseline. But DC 15 starts as a 45%/55% chance and barely increases to 55%/85% as you level up. This bias would have us set DCs a bit too high at low level and massively too high at high levels.
WotC adventures include example DCs. They seem to like DC 20 as a baseline. See previous point.

Now remember these DCs were might *cough* *cough* have been reasonable at low level. However at Tier 2-3 the PC is looking at a bad chance of doing something they could do in Tier 1. That means they expect to fail at doing something that isn't even level appropriate anymore.

How can DMs patch this?
Well simple, recognize bounded accuracy does not work on a 1 roll subsystem. So either make it a 2 roll subsystem or, more commonly, by discarding bounded accuracy for ability checks and increasing the progression in that roll.

Want to increase the distance the window moves? Increase (20th level modifier - 1st level modifier) / RNG size. RNG size being 21 in the case of a d20. Maybe change proficiency to be +1 to +7 instead of +2 to +6. Or maybe even more drastic +4 to +12.
Want to move only the minimum side of the window? Shrink the RNG size with level. Reliable Talent turns a d20 into something like 9+1d11. Imagine the minimum roll was level/3 (round up).
Want to move only the maximum side of the window? Grow the RNG size with level. Probably best to add a bell curve here. Imagine 3d6 at Tier 1 but 5d6 at Tier 3. At one point during development of 5E, proficiency was a die instead of a bonus.


When I wanted to make a skilled based character in 5E I dipped Cleric for Guidance, went Rogue for Expertise, stayed Rogue for Reliable Talent
So I went from 1d20+7 at 1st to 1d20(min 10)+1d4+12 at 11th

MinMax
1st827
20th2841
shift2014

In other words I overspecialized in breaking bounded accuracy, but that was required for the character concept to have growth in skills to match the growth in levels. This is not the normal. Unless you break bounded accuracy, your skills will be obsoleted, but math can fix that.

[Man_Over_Game]

Certainly.

I will first describe progression and windows by using encounters as an example:
An Nth level party has a range of encounters that challenge them. This spans from the ones just above guaranteed victory to those just under guaranteed defeat. This span is a window. At different levels the window has different mins and maxs. As the party levels from 1st to 20th, that window will continue to move towards harder encounters. I use the term "progression" to describe how much that window moves. Some games have 0 progression because the window never moves. Other games have the window move a bit but always have first and last level overlap a bit. D&D's default narrative has a rather large progression, the gap between the 1st level window and the 20th level window could be measured in multiple window lengths. This concept of windows and progression applies to more than just encounters. It applies to many subsystems in D&D. Attacks and Ability Checks are two examples.

How does bounded accuracy work:
Attacks are comprised of an accuracy or "attack" roll and a damage roll. Bounded Accuracy is the idea of minimizing the progression in the attack roll without compromising the progression of attacks themselves. They handle this by increasing the progression of the damage roll while shrinking the progression of the attack roll. Basically a 20th level attack action is still much stronger than a 1st level attack action, but both can hit the same target.

Why is this relevant to ability checks (including skill checks):
In 5E the progression of the accuracy of ability checks was also shrunk. However what got increased to maintain the overall progression of ability checks? Nothing. There was no second magnitude roll for WotC to increase. As a result the progression of ability checks shrunk massively out of a simple design error.

What do I mean by anchoring?

Games tend to start at low levels. This means DMs are more likely to assign DCs based on preventing a 1st level character from achieving it too soon rather than on when a character will master (get to 100%) that challenge. If we imagine a reasonable amount of progression, then this bias would have us set the DCs too high for mid-high level characters because the 5E progression was shrunk.
The DC grades (5, 10, 15, 20, 25, 30) have names that encourage you to start with DC 15 as your baseline. But DC 15 starts as a 45%/55% chance and barely increases to 55%/85% as you level up. This bias would have us set DCs a bit too high at low level and massively too high at high levels.
WotC adventures include example DCs. They seem to like DC 20 as a baseline. See previous point.

Now remember these DCs were might *cough* *cough* have been reasonable at low level. However at Tier 2-3 the PC is looking at a bad chance of doing something they could do in Tier 1. That means they expect to fail at doing something that isn't even level appropriate anymore.

How can DMs patch this?
Well simple, recognize bounded accuracy does not work on a 1 roll subsystem. So either make it a 2 roll subsystem or, more commonly, by discarding bounded accuracy for ability checks and increasing the progression in that roll.

Want to increase the distance the window moves? Increase (20th level modifier - 1st level modifier) / RNG size. RNG size being 21 in the case of a d20. Maybe change proficiency to be +1 to +7 instead of +2 to +6. Or maybe even more drastic +4 to +12.
Want to move only the minimum side of the window? Shrink the RNG size with level. Reliable Talent turns a d20 into something like 9+1d11. Imagine the minimum roll was level/3 (round up).
Want to move only the maximum side of the window? Grow the RNG size with level. Probably best to add a bell curve here. Imagine 3d6 at Tier 1 but 5d6 at Tier 3. At one point during development of 5E, proficiency was a die instead of a bonus.


When I wanted to make a skilled based character in 5E I dipped Cleric for Guidance, went Rogue for Expertise, stayed Rogue for Reliable Talent
So I went from 1d20+7 at 1st to 1d20(min 10)+1d4+12 at 11th

MinMax
1st827
20th2841
shift2014

In other words I overspecialized in breaking bounded accuracy, but that was required for the character concept to have growth in skills to match the growth in levels. This is not the normal. Unless you break bounded accuracy, your skills will be obsoleted, but math can fix that.

Erm, to unpack that even further:

Attack & Damage = P * A

Skills = P * 1

Where P is your percent chance of hitting the target with your D20 roll, and A is an amplifying effect that increases as you level.

We know there is an "A", as virtually every damage-dealing class in the game gains bonuses that allow them to deal more damage as you level. Fighters gets more attacks, Barbarians get bigger Rage bonuses, Rogues deal more Sneak Attack, spell selection expands to include higher damaging options, etc.

What doesn't work is not having an amplifier. Unless you want every creature in the game to have roughly the same amount of HP as a CR 1 creature, it doesn't work. In the same way, we cannot expect a DC 10 climbing check to be as relevant in a level 1 Adventure as it would be in a level 20 adventure. The AC/DCs scale up over time (because problems get harder), but while the Attack system scales because of the damage amplifiers, skills get no such thing.

As a result, skills basically stagnate as a Tier 1 feature, and don't really expand beyond that, unlike the attack system. It is effectively the same as a Glamour Bard using the Attack Action as his primary form of contribution (fine for a few levels, bad after 5).

[Pex]

Certainly.

I will first describe progression and windows by using encounters as an example:
An Nth level party has a range of encounters that challenge them. This spans from the ones just above guaranteed victory to those just under guaranteed defeat. This span is a window. At different levels the window has different mins and maxs. As the party levels from 1st to 20th, that window will continue to move towards harder encounters. I use the term "progression" to describe how much that window moves. Some games have 0 progression because the window never moves. Other games have the window move a bit but always have first and last level overlap a bit. D&D's default narrative has a rather large progression, the gap between the 1st level window and the 20th level window could be measured in multiple window lengths. This concept of windows and progression applies to more than just encounters. It applies to many subsystems in D&D. Attacks and Ability Checks are two examples.

How does bounded accuracy work:
Attacks are comprised of an accuracy or "attack" roll and a damage roll. Bounded Accuracy is the idea of minimizing the progression in the attack roll without compromising the progression of attacks themselves. They handle this by increasing the progression of the damage roll while shrinking the progression of the attack roll. Basically a 20th level attack action is still much stronger than a 1st level attack action, but both can hit the same target.

Why is this relevant to ability checks (including skill checks):
In 5E the progression of the accuracy of ability checks was also shrunk. However what got increased to maintain the overall progression of ability checks? Nothing. There was no second magnitude roll for WotC to increase. As a result the progression of ability checks shrunk massively out of a simple design error.

What do I mean by anchoring?

Games tend to start at low levels. This means DMs are more likely to assign DCs based on preventing a 1st level character from achieving it too soon rather than on when a character will master (get to 100%) that challenge. If we imagine a reasonable amount of progression, then this bias would have us set the DCs too high for mid-high level characters because the 5E progression was shrunk.
The DC grades (5, 10, 15, 20, 25, 30) have names that encourage you to start with DC 15 as your baseline. But DC 15 starts as a 45%/55% chance and barely increases to 55%/85% as you level up. This bias would have us set DCs a bit too high at low level and massively too high at high levels.
WotC adventures include example DCs. They seem to like DC 20 as a baseline. See previous point.

Now remember these DCs were might *cough* *cough* have been reasonable at low level. However at Tier 2-3 the PC is looking at a bad chance of doing something they could do in Tier 1. That means they expect to fail at doing something that isn't even level appropriate anymore.

How can DMs patch this?
Well simple, recognize bounded accuracy does not work on a 1 roll subsystem. So either make it a 2 roll subsystem or, more commonly, by discarding bounded accuracy for ability checks and increasing the progression in that roll.

Want to increase the distance the window moves? Increase (20th level modifier - 1st level modifier) / RNG size. RNG size being 21 in the case of a d20. Maybe change proficiency to be +1 to +7 instead of +2 to +6. Or maybe even more drastic +4 to +12.
Want to move only the minimum side of the window? Shrink the RNG size with level. Reliable Talent turns a d20 into something like 9+1d11. Imagine the minimum roll was level/3 (round up).
Want to move only the maximum side of the window? Grow the RNG size with level. Probably best to add a bell curve here. Imagine 3d6 at Tier 1 but 5d6 at Tier 3. At one point during development of 5E, proficiency was a die instead of a bonus.


When I wanted to make a skilled based character in 5E I dipped Cleric for Guidance, went Rogue for Expertise, stayed Rogue for Reliable Talent
So I went from 1d20+7 at 1st to 1d20(min 10)+1d4+12 at 11th

MinMax
1st827
20th2841
shift2014

In other words I overspecialized in breaking bounded accuracy, but that was required for the character concept to have growth in skills to match the growth in levels. This is not the normal. Unless you break bounded accuracy, your skills will be obsoleted, but math can fix that.

You're talking game math, which is fine, but I find your point more interesting in terms of what can be accomplished or rather can't.

For example, a DM would not think it reasonable for a 1st level character to bring a dead character back to life. It takes a 5th level character to do that with Revivify, which has its restrictions of an expensive material component and must be cast within a minute of death so the spell must be prepared if a player wants to use it at the moment it's needed. Therefore there's no way a 1st level character even proficient in Medicine can perform CPR to bring a dead character back to life. The DC is no. When characters are 5th level, it takes a 3rd level spell to bring a character back to life, so how is a 5th level PC to do it by a Medicine check? There's no DC guideline from the rules, so the DM has to make it up but does he want to make it up? A simple die roll should replace a 3rd level spell and a 300 gp diamond? DC remains No. Ninth level comes along and now there's Raise Dead for up to 10 days since death and 500 gp diamond. Maybe now Medicine CPR is fine to duplicate Revivify? Is it free of cost? Revivify is still available and more affordable to be prepared and used. What's the DC? 20? Why wasn't it DC 20 at 1st level then if you can put a DC on it? DC 25? 18 Wisdom and proficient, you have +8 to the roll. Only works if you roll a 17+. Not good odds, just use the spell. However, back at level 1 16 Wisdom and proficient you have +5. You can Medicine CPR on a Natural 20 and have Guidance maybe, Bardic Inspiration too for teamwork, but you as DM didn't want 1st level characters bring the dead back to life, so now the DC must be 30 which the 9th level character can't achieve anymore without a buff, but it's all DM made up. That won't help all the other players in all the other games where the DM won't even consider this. The rules say nothing. You can never bring the dead back to life with Medicine. You don't get greater magnitude of skill use at level 20 you didn't have at level 1.

[porchdog]

And most DMs don't have level 2 characters acting against fiends, constructs and fey, well, at level 2. Furthermore, a lot of DMs (not me, but I know quite a few) tend to populate their worlds mostly with humanoids, and make non-humanoid threats rarer - well, aside from the nicely generic low-level undead, I suppose. I can't call that rolling out the red carpet, since nothing is actually altered, the game works as expected when the spellcaster does their thing, and nobody catered to them specifically.

Like I said this is the big problem. Why DON"T most DMs use fiends, constructs and fey or anything other then the Easy Spell Target creatures? See, the problem? As soon as the DM says that they will only use creatures that are Easy Spell Targets, they are rolling out the Red Carpet for magic. It's not like there are not plenty of low powered fiends, constructs and so forth: just check the monster books.

I guess you might say the DM is doing it subconsciously....but they are still doing it.



Again, please shelve the distraction of GMs favoring magic. We are not discussing bad GMs here, we are discussing whether or not the system provides a good basis for Martials to interact with the world assuming a neutral GM guided by said system.

Well, lets not bother with if it's good or bad: how about you just acknowledge that it is a thing that really screws over martials. It's not even about the DM being neutral, because what the DM is doing is still creating the problem.

And this IS about the rule system. The rule system does provide a good basis for martials: if it is used. Replace even some of the Easy Magic Spell Targets and you will see a change in the game. Suddenly the spellcasters can't just Easy Spell there way past nearly everything, BUT the martials can still use their abilities against them. The martial can sneak past a dust mephit the same way they can a human.


Your example earlier of a wizard using Spider Climb was bad form. It's not the simple case the guard autofails to see the wizard because he used the spell where the guard autosucceeds seeing the rogue because he climbed.

In most cases it won't be the mundane roll vs the autosucceed spell. Most of the time the mundane DC roll will be quite easy. So it's not that the martial character will fail a check, it's that they don't get a fair shake. Like how by DM whim the guard sees the mundane climber every time, but amazingly misses the spellcaster.

Martials and spellcasters can both miss with and attack or have a foe make a save, but only the spellcaster gets the DM rolling out the Red Carpet so everything works out great for the character.




Right, so the spellcaster uses another spell if he prepared it.

Well, the spellcaster won't always have another appropriate spell read.



By the time Awakened Shrubs and flying swords become suitable encounter guards we're talking higher level and spellcasters have access to better spells.

What time? An awakened shrub has a CR of zero. The flying sword is all of 1/4 CR. Going by the rules, either would be just fine for an encounter for a 1 st level character.

[Man_Over_Game]

I'd be inclined to decide which Skills were relevant, then let the player pick a skill from those to be rolled (in case they have Expertise in something) and let them have an extra die (a la Advantage) for every other relevant skill they have Proficiency in. I kinda do something like that now, sometimes, but nothing in the lines of this specific scenario has arisen. It would at least give extra Skills a way to increase chances for success, rather than introducing more opportunities for failure.

But as you say, that's probably going well outside the RAW.

What's really ironic is that you're not really even houseruling all that much. We kinda have this weird mentality that, in order to check someone for poisons, we need a Wisdom (Medicine) Check, but really, the Ability Score used is the only part predefined by the DM:

"For every ability check, the DM decides which of the six abilities is relevant to the task at hand and the difficulty of the task, represented by a Difficulty Class.
[...]
Sometimes, the DM might ask for an ability check using a specific skill—for example, “Make a Wisdom(Perception) check.” At other times, a player might ask the DM if proficiency in a particular skill applies to a check. In either case, proficiency in a skill means an individual can add his or her proficiency bonus to ability checks that involve that skill. Without proficiency in the skill, the individual makes a normal ability check."

Xanathar's goes a step further, granting Advantage to an Ability Check if a second proficiency can be applied.


So I tell you that jumping the gap is a Strength Check, and then you tell me what skills you want to apply, Advantage if you have more than one. Kinda funny, when you consider things like playing with a grid is a variant rule, while creating a custom background isn't.

[prabe]

What's really ironic is that you're not really even houseruling all that much. We kinda have this weird mentality that, in order to check someone for poisons, we need a Wisdom (Medicine) Check, but really, the Ability Score used is the only part predefined by the DM:

"For every ability check, the DM decides which of the six abilities is relevant to the task at hand and the difficulty of the task, represented by a Difficulty Class.
[...]
Sometimes, the DM might ask for an ability check using a specific skill—for example, “Make a Wisdom(Perception) check.” At other times, a player might ask the DM if proficiency in a particular skill applies to a check. In either case, proficiency in a skill means an individual can add his or her proficiency bonus to ability checks that involve that skill. Without proficiency in the skill, the individual makes a normal ability check."

Xanathar's goes a step further, granting Advantage to an Ability Check if a second proficiency can be applied.


So I tell you that jumping the gap is a Strength Check, and then you tell me what skills you want to apply, Advantage if you have more than one. Kinda funny, when you consider things like playing with a grid is a variant rule, while creating a custom background isn't.

I'd either missed or forgotten the bit in Xanathar's about multiple Skills (potentially) granting Advantage; good to know I'm not wandering way far afield with that approach. I list skills (and occasionally Tool Proficiencies) because I don't know what Skills the PCs I'm GMing are proficient (or more) in, and it at least feels as though it speeds up play if they know what Skills apply instead of guessing and asking. Doesn't mean I'm unwilling to expand that if a suggestion makes sense, of course.

[Ignimortis]

Like I said this is the big problem. Why DON"T most DMs use fiends, constructs and fey or anything other then the Easy Spell Target creatures? See, the problem? As soon as the DM says that they will only use creatures that are Easy Spell Targets, they are rolling out the Red Carpet for magic. It's not like there are not plenty of low powered fiends, constructs and so forth: just check the monster books.

I guess you might say the DM is doing it subconsciously....but they are still doing it.


I don't think preferring to start the game in a relatively grounded setting is "rolling out the carpet" for casters. It's the default, it's the majority's expectation that at low levels, you'll be dealing with other normal humanoids, like humans and orcs and goblins and elves.

And even if your campaign starts in the court of the faerie king Oberon and his guards are constructs (who get blindsight, so it's 100% impossible to sneak past them without just going waaaay around - put two in a chokepoint, nobody is ever crossing it without making a lot of noise) and awakened plants, they still do nothing to compensate for the fact that Pass without Trace and/or Invisibility outclasses everything a sneaky character without Expertise will ever do - even at high levels. It's a problem of math, not GMing. There are quite a few spells which don't rely on the opponent being a humanoid to deal with them.

[Skylivedk]

Like I said this is the big problem. Why DON"T most DMs use fiends, constructs and fey or anything other then the Easy Spell Target creatures? See, the problem? As soon as the DM says that they will only use creatures that are Easy Spell Targets, they are rolling out the Red Carpet for magic. It's not like there are not plenty of low powered fiends, constructs and so forth: just check the monster books.
And later in the game the casters have tons of ways of dealing with all of the above. Both in terms of straight hp damage, but also stuff like Banishment which is almost an auto-win against what you mention here. The casters keep getting bigger and better tools in a way that Martials don't.

I'll repost my example from the earlier thread since it was never countered:
The Barbarian's rage tops out at 6 times per day. If he can fly, he cannot do the control (of Ancients Barb) etc. From 8 to 20 the Barbarian gets:
ASIs (normal speed)
Brutal Critical (less than 5% - less than 10% extra damage) x 3 (each time less relative damage)
Relentless Rage
Path Feature
Indomitable Might
Primal Champion

Brutal Critical is meh. Relentless Rage is cool - both in and out of combat applicability. The Path Feature is pre-locked since level 3. It can be cool without being mind-blowing. Indomitable Might is good, but too late and Primal Champion is good, but pretty boring.

At the same time the worst of the full-casters get:
7 more spell known (from level 5 to 9) (and 12 changes of previous class abilities, sorry, spells)
2 new metamagics
2 Sorcerous Origin (and ie while the Barbarian can get a Fly speed while raging tops 1 round at a time, the Sorcerer just gets a Fly speed)
1 Sorcerous Restoration

I cannot see how it is not painfully obvious that casters get more features, more powerful features and more flexibility in choosing said abilities.




I guess you might say the DM is doing it subconsciously....but they are still doing it.

My experience as a DM and as a player is wildly different. The published modules can't handle well-played spell casters and in sandbox games the fact that casters can change combat resources to utility resources stack the deck evermore in their favour.

Scrying, Dream, long-term minionmancy, resource creating spells etc are not matched by Martials.



Well, lets not bother with if it's good or bad: how about you just acknowledge that it is a thing that really screws over martials. It's not even about the DM being neutral, because what the DM is doing is still creating the problem.

And this IS about the rule system. The rule system does provide a good basis for martials: if it is used. Replace even some of the Easy Magic Spell Targets and you will see a change in the game. Suddenly the spellcasters can't just Easy Spell there way past nearly everything, BUT the martials can still use their abilities against them. The martial can sneak past a dust mephit the same way they can a human.
No, it doesn't. Also casters have equal access to skills so skills don't save Martials. It's highly presumptuous to believe that those stating that Martials suffer don't play by the rules.



In most cases it won't be the mundane roll vs the autosucceed spell. Most of the time the mundane DC roll will be quite easy. So it's not that the martial character will fail a check, it's that they don't get a fair shake. Like how by DM whim the guard sees the mundane climber every time, but amazingly misses the spellcaster.

Martials and spellcasters can both miss with and attack or have a foe make a save, but only the spellcaster gets the DM rolling out the Red Carpet so everything works out great for the character.

You're extrapolating from a few examples to include a very large amount of other posters in a way that's hardly fair to the points that were made across more than 50 pages. I don't think you do out of bad faith, but you are doing it nonetheless.



Well, the spellcaster won't always have another appropriate spell read.

As casters get increasingly more spells and slots this issue diminishes. Again, Martials get zilch in comparison.



I don't think preferring to start the game in a relatively grounded setting is "rolling out the carpet" for casters. It's the default, it's the majority's expectation that at low levels, you'll be dealing with other normal humanoids, like humans and orcs and goblins and elves.

And even if your campaign starts in the court of the faerie king Oberon and his guards are constructs (who get blindsight, so it's 100% impossible to sneak past them without just going waaaay around - put two in a chokepoint, nobody is ever crossing it without making a lot of noise) and awakened plants, they still do nothing to compensate for the fact that Pass without Trace and/or Invisibility outclasses everything a sneaky character without Expertise will ever do - even at high levels. It's a problem of math, not GMing. There are quite a few spells which don't rely on the opponent being a humanoid to deal with them.

Full co-sign.

About me suggesting some combat buffs as well: I do think gap closing (movement speed) becomes an issue at higher levels.

Besides that I agree that out of combat is where you find the biggest gap. I've already posted a lot of suggested modifications to 5e to address the issue, including removing certain spells, changing the skill system (to have more progression and unlocking heroic and epic abilities) and flexibility within and across subclasses when it comes to noon-combat martial abilities. I don't repost everything every time I post

[Morty]

4E's knock -- where IIRC the wizard effectively got to make a Int-based instead of Dex-based pick locks check and nothing more -- is a good model for how it could be done differently.

As weird as it is to give Pathfinder credit for anything, I think that's how it works in both its editions as well. But then, I consider Knock to be a poor example of spells outshining skills, because opening locks just isn't that important. If wizards having access to it overshadows rogues (which I don't think it does), that's more an indictment of the rogue class than anything else.

[Xervous]

Like I said this is the big problem. Why DON"T most DMs use fiends, constructs and fey or anything other then the Easy Spell Target creatures? See, the problem? As soon as the DM says that they will only use creatures that are Easy Spell Targets, they are rolling out the Red Carpet for magic. It's not like there are not plenty of low powered fiends, constructs and so forth: just check the monster books.

I guess you might say the DM is doing it subconsciously....but they are still doing it.



Well, lets not bother with if it's good or bad: how about you just acknowledge that it is a thing that really screws over martials. It's not even about the DM being neutral, because what the DM is doing is still creating the problem.

And this IS about the rule system. The rule system does provide a good basis for martials: if it is used. Replace even some of the Easy Magic Spell Targets and you will see a change in the game. Suddenly the spellcasters can't just Easy Spell there way past nearly everything, BUT the martials can still use their abilities against them. The martial can sneak past a dust mephit the same way they can a human.



In most cases it won't be the mundane roll vs the autosucceed spell. Most of the time the mundane DC roll will be quite easy. So it's not that the martial character will fail a check, it's that they don't get a fair shake. Like how by DM whim the guard sees the mundane climber every time, but amazingly misses the spellcaster.

Martials and spellcasters can both miss with and attack or have a foe make a save, but only the spellcaster gets the DM rolling out the Red Carpet so everything works out great for the character.




Well, the spellcaster won't always have another appropriate spell read.



What time? An awakened shrub has a CR of zero. The flying sword is all of 1/4 CR. Going by the rules, either would be just fine for an encounter for a 1 st level character.

1. Where is the explicit guidance stating DMs should bias their creature selection to include more creatures that can’t be targeted by such spells? There is no such guidance. Therefore the intentional act of putting in these monsters is an intentional act of limiting the casters, hence the putting up walls quote.

2. I don’t have to entertain your conjectures about GM caster favoritism. Such claims are both anecdotal and tangential to the topic at hand which is a discussion of whether or not the mechanical foundation of the system is tilted against Martials. You are free to put forth your personal experiences as was originally invited but a sample size of 1 that reeks of bias is not in any way a helpful reference for a discussion on mechanics.

3. DCs are assigned by GM fiat with insufficient guidance on both expected level appropriate benchmarks and what a given DC might actually map to. Again the GM has to actively choose to favor Martials here by letting their skills do stuff. There is minimal guidance on what most uncontested skill checks can accomplish. Players can’t assume anything beyond a handful of the more clearly defined options. They are dependent upon favorable GM rulings for even baseline functionality which goes beyond a neutral GM assumption for our analysis.

If you could kindly reference the DMG where it advises building walls in front of the caster in terms of taking their abilities into account and negating them through encounter design I will happily concede the point. Failing that your statements have no supporting evidence in this debate focusing on the system.

[Willie the Duck]

As weird as it is to give Pathfinder credit for anything, I think that's how it works in both its editions as well. But then, I consider Knock to be a poor example of spells outshining skills, because opening locks just isn't that important. If wizards having access to it overshadows rogues (which I don't think it does), that's more an indictment of the rogue class than anything else.

The example is just that -- an example (chosen because I remembered it). Opening locks certainly isn't that important anymore, but it was a key Thief ability back when the class was created. And it highlights well the issue that non-magical abilities have pretty much always had a roll involved, while magic doesn't necessarily.

Pertaining to indicting the rogue class -- well, despite my misgivings about how much the initial scenario actually shows, I think we're doing a bang up job of indicting the skill system in general, and the character concept of a lone-infiltrator. Previous editions may or may not have done better at letting a skulking character stealth, surprise, and take out random guards (I don't really think any other edition did it well), but the overall premise of 'I sneak on ahead and scout around' has been something that all the D&Ds have suggested was an iconic Thief/Rogue activity, yet providing for it woeful mechanical support.

[porchdog]

I don't think preferring to start the game in a relatively grounded setting is "rolling out the carpet" for casters. It's the default, it's the majority's expectation that at low levels, you'll be dealing with other normal humanoids, like humans and orcs and goblins and elves.

Well, you can call it a mistake, or oversight or unintended consequence. If the DM and players choose to have an Earth like setting, then magic will be over powered, problematic and game breaking.

And note this is not a problem with the rules, this is about the gamers choice. The rules are just fine as written: but when the game play ignores, skips or does not use all the rules, that is where problems start.

I wonder what majority your talking about? And, you might notice there are lots of creatures of every type other then humanoid in the rules, all CR one or less: So, if you don't think such creatures should be used a low level, when should they be used?

And look if you really want a "grounded" setting, it's not that you can't do it and have fun. The point is you can't change the game play and play a set way, and then say that there is something wrong.





And even if your campaign starts in the court of the faerie king Oberon and his guards are constructs (who get blindsight, so it's 100% impossible to sneak past them without just going waaaay around - put two in a chokepoint, nobody is ever crossing it without making a lot of noise) and awakened plants, they still do nothing to compensate for the fact that Pass without Trace and/or Invisibility outclasses everything a sneaky character without Expertise will ever do - even at high levels. It's a problem of math, not GMing. There are quite a few spells which don't rely on the opponent being a humanoid to deal with them.

This is my other point though. Are you saying Pass without Trace and/or Invisibility are so good that no foe can detect someone using them? Your not going on the side of when a character uses invisibility they are utterly undetectable by all, are you? And in 5E invisible doesn't mean hidden. If they are invisible, the creatures have disadvantage on attacking them, but they must use an action to hide to avoid the enemy knowing their location(For the purpose of Hiding, the creature is heavily obscured. The creature’s location can be detected by any noise it makes or any tracks it leaves.)

All of the above rules make invisibility a part of the game. Yet, many a DM will just toss their hands up and say "no one detects your invisible character".


And later in the game the casters have tons of ways of dealing with all of the above. Both in terms of straight hp damage, but also stuff like Banishment which is almost an auto-win against what you mention here. The casters keep getting bigger and better tools in a way that Martials don't.

Well, it's a good thing the game settings, features and creatures do in fact scale up too.



My experience as a DM and as a player is wildly different. The published modules can't handle well-played spell casters and in sandbox games the fact that casters can change combat resources to utility resources stack the deck evermore in their favour.

Published modules will always be Easy as they have to be as they are for the general public. It's the classic "everyone in the group must walk at the pace of the slowest walkers in the group". Though, even if they wanted too, a published module would need tons of more pages to handle powerful PCs, and that is not piratical.



Scrying, Dream, long-term minionmancy, resource creating spells etc are not matched by Martials.

Well, this goes back to if you just want to give martials mundane abilities that match the spells exactly. They can use the mundane ability Sight Beyond Sight, just like scrying; they can Attract Followers, just like minionmancy.




No, it doesn't. Also casters have equal access to skills so skills don't save Martials.

Martials get more skill points and skill related abilities.



You're extrapolating from a few examples to include a very large amount of other posters in a way that's hardly fair to the points that were made across more than 50 pages. I don't think you do out of bad faith, but you are doing it nonetheless.

I missed the last thread.... But I do see this is games all the time. Magic always works, and mundanes always get screwed is the default for many games.



Besides that I agree that out of combat is where you find the biggest gap.

I agree here. And if you want to homebrew a ton of fixes then you can.

My big point though, is you don't need to. If you use what is right there in the game rules, you can have a setting that is much less an easy target for magic.

And if you become aware of the free pass given to magic, and put a stop to it cold turkey....you will have a setting that is much less an easy target for magic.


1. Where is the explicit guidance stating DMs should bias their creature selection to include more creatures that can’t be targeted by such spells? There is no such guidance. Therefore the intentional act of putting in these monsters is an intentional act of limiting the casters, hence the putting up walls quote.

Well, D&D has always been lax with guidance. If you see the use of some creatures as an "intentional act of limiting the casters", I wonder: What is an intentional act of limiting the martials? I wonder if you are against such acts that limit martials?

There is no "guidance" to tell DMs to lock down maritals, and yet lots of DMs do just that.





2. I don’t have to entertain your conjectures about GM caster favoritism. Such claims are both anecdotal and tangential to the topic at hand which is a discussion of whether or not the mechanical foundation of the system is tilted against Martials. You are free to put forth your personal experiences as was originally invited but a sample size of 1 that reeks of bias is not in any way a helpful reference for a discussion on mechanics.

The use of everything in the game does fall under "mechanics" right? Like having that back castle door guarded by an awaked shrub, not a human guard.



3. DCs are assigned by GM fiat with insufficient guidance on both expected level appropriate benchmarks and what a given DC might actually map to. Again the GM has to actively choose to favor Martials here by letting their skills do stuff. There is minimal guidance on what most uncontested skill checks can accomplish. Players can’t assume anything beyond a handful of the more clearly defined options. They are dependent upon favorable GM rulings for even baseline functionality which goes beyond a neutral GM assumption for our analysis.

Ok, this is the basic foundation of D&D, so this is all good here.



If you could kindly reference the DMG where it advises building walls in front of the caster in terms of taking their abilities into account and negating them through encounter design I will happily concede the point. Failing that your statements have no supporting evidence in this debate focusing on the system.

But it's not about that? Your taking it to personal.

Right now in many D&D games casters get a sweet free ride pass as the DM and players all give them a free pass to do whatever they want to make a fun game.....while at the same time locking down marital characters to be ineffective or useless.

It's a choice. It's not game design or rule design.

And once you take that away, no more free pass for magic, you will suddenly see the game work much differently.

[Xervous]

Well, you can call it a mistake, or oversight or unintended consequence. If the DM and players choose to have an Earth like setting, then magic will be over powered, problematic and game breaking.

And note this is not a problem with the rules, this is about the gamers choice. The rules are just fine as written: but when the game play ignores, skips or does not use all the rules, that is where problems start.

I wonder what majority your talking about? And, you might notice there are lots of creatures of every type other then humanoid in the rules, all CR one or less: So, if you don't think such creatures should be used a low level, when should they be used?

And look if you really want a "grounded" setting, it's not that you can't do it and have fun. The point is you can't change the game play and play a set way, and then say that there is something wrong.





This is my other point though. Are you saying Pass without Trace and/or Invisibility are so good that no foe can detect someone using them? Your not going on the side of when a character uses invisibility they are utterly undetectable by all, are you? And in 5E invisible doesn't mean hidden. If they are invisible, the creatures have disadvantage on attacking them, but they must use an action to hide to avoid the enemy knowing their location(For the purpose of Hiding, the creature is heavily obscured. The creature’s location can be detected by any noise it makes or any tracks it leaves.)

All of the above rules make invisibility a part of the game. Yet, many a DM will just toss their hands up and say "no one detects your invisible character".



Well, it's a good thing the game settings, features and creatures do in fact scale up too.



Published modules will always be Easy as they have to be as they are for the general public. It's the classic "everyone in the group must walk at the pace of the slowest walkers in the group". Though, even if they wanted too, a published module would need tons of more pages to handle powerful PCs, and that is not piratical.



Well, this goes back to if you just want to give martials mundane abilities that match the spells exactly. They can use the mundane ability Sight Beyond Sight, just like scrying; they can Attract Followers, just like minionmancy.




Martials get more skill points and skill related abilities.



I missed the last thread.... But I do see this is games all the time. Magic always works, and mundanes always get screwed is the default for many games.



I agree here. And if you want to homebrew a ton of fixes then you can.

My big point though, is you don't need to. If you use what is right there in the game rules, you can have a setting that is much less an easy target for magic.

And if you become aware of the free pass given to magic, and put a stop to it cold turkey....you will have a setting that is much less an easy target for magic.



Well, D&D has always been lax with guidance. If you see the use of some creatures as an "intentional act of limiting the casters", I wonder: What is an intentional act of limiting the martials? I wonder if you are against such acts that limit martials?

There is no "guidance" to tell DMs to lock down maritals, and yet lots of DMs do just that.





The use of everything in the game does fall under "mechanics" right? Like having that back castle door guarded by an awaked shrub, not a human guard.



Ok, this is the basic foundation of D&D, so this is all good here.



But it's not about that? Your taking it to personal.

Right now in many D&D games casters get a sweet free ride pass as the DM and players all give them a free pass to do whatever they want to make a fun game.....while at the same time locking down marital characters to be ineffective or useless.

It's a choice. It's not game design or rule design.

And once you take that away, no more free pass for magic, you will suddenly see the game work much differently.

Yes, changing something about a broken system can yield a new state where the new system is not broken. That’s standard Oberoni.

We are clearly in agreement that walls need to be put in front of casters to prevent them from trampling campaigns into the ground and that GMs need to handle skills favorably for Martials in order to allow them to contribute, but the fresh face to 5e is not aware of this. They are given a book of rules and empty holes for them to fill in, a book that presents casters and martials at implicit equal worth and validity. Trial and error should not be necessary to see past the falsehood of balance that is put forth. If the system works by walling wizards and carpeting fighters it should state that outright rather than leave it up to a new GM to puzzle out over a span of months or years if they even manage to arrive at the realization at all.

[Morty]

The example is just that -- an example (chosen because I remembered it). Opening locks certainly isn't that important anymore, but it was a key Thief ability back when the class was created. And it highlights well the issue that non-magical abilities have pretty much always had a roll involved, while magic doesn't necessarily.

Yes, the legacy of being "the lock and trap class" is hanging over the rogue to this way, much to its detriment. Either way, Knock does illustrate the issue, I just question its weight in actual play.


Pertaining to indicting the rogue class -- well, despite my misgivings about how much the initial scenario actually shows, I think we're doing a bang up job of indicting the skill system in general, and the character concept of a lone-infiltrator. Previous editions may or may not have done better at letting a skulking character stealth, surprise, and take out random guards (I don't really think any other edition did it well), but the overall premise of 'I sneak on ahead and scout around' has been something that all the D&Ds have suggested was an iconic Thief/Rogue activity, yet providing for it woeful mechanical support.

In all fairness, stealth being troublesome isn't restricted to D&D. It's easy for it to end up in a situation where the sneak has to make multiple rolls and failing any one exposes them, forcing them to run or fight. But in general yes, D&D 5E's skill system is troublesome on its own and also stacks up poorly to spells. It wouldn't be as problematic in a system without a staggering variety of convenient and powerful spell at casters' easy disposal.

[Willie the Duck]

In all fairness, stealth being troublesome isn't restricted to D&D. It's easy for it to end up in a situation where the sneak has to make multiple rolls and failing any one exposes them, forcing them to run or fight. But in general yes, D&D 5E's skill system is troublesome on its own and also stacks up poorly to spells. It wouldn't be as problematic in a system without a staggering variety of convenient and powerful spell at casters' easy disposal.

Too true, and honestly a bunch of the systems people bring up as better for skills are actually worse for stealth -- GURPS and Hero System, for instance, have charts of penalties for stealth checks, and super-disfavor the poor discovered infiltrator when it comes to escaping alive (Cunning Action is one really nice bone that 5e threw the infiltrating rogue, in that now they can at least disengage and book it back to their friends better).

The 'but the other guy/other class I could play has spells' issue is always going to be D&D's albatross. Made worse by how the spells just work. They are relatively sure things/known qualities/quantities -- the infiltrator wizard knows exactly how many times they can _____ (skulking spell) and how likely they are to succeed when using it.

[Skylivedk]

This is my other point though. Are you saying Pass without Trace and/or Invisibility are so good that no foe can detect someone using them? Your not going on the side of when a character uses invisibility they are utterly undetectable by all, are you? And in 5E invisible doesn't mean hidden. If they are invisible, the creatures have disadvantage on attacking them, but they must use an action to hide to avoid the enemy knowing their location(For the purpose of Hiding, the creature is heavily obscured. The creature’s location can be detected by any noise it makes or any tracks it leaves.)

Yup, Pass without Trace is way better for the team trying to infiltrate than a Rogue playing bodybag. The bonus is 3-5 times bigger in tier 1 and 2 (than expertise) and it includes the entire team. No contest.



All of the above rules make invisibility a part of the game. Yet, many a DM will just toss their hands up and say "no one detects your invisible character".

And let's just disregard these examples since they're not essential for the contentious points. There's of course an issue of how a DM has no guidance on when sound stops being a factor, but we can leave that as a niche case


Well, it's a good thing the game settings, features and creatures do in fact scale up too.
For Martials, not really compared to casters, ie. my example with the Sorcerer which you conveniently neglected.



Published modules will always be Easy as they have to be as they are for the general public. It's the classic "everyone in the group must walk at the pace of the slowest walkers in the group". Though, even if they wanted too, a published module would need tons of more pages to handle powerful PCs, and that is not piratical.
Tomb of Annihilation is best cleared with casters and half-casters. So is Rise of Tiamat.



Well, this goes back to if you just want to give martials mundane abilities that match the spells exactly. They can use the mundane ability Sight Beyond Sight, just like scrying; they can Attract Followers, just like minionmancy.

No. I have specifically said I don't want that multiple times. I don't assume bad faith. I do dislike you saying what I want. Stop that. Quote me if you want.




Martials get more skill points and skill related abilities.
Unsubstantiated claim. Counter: bards and knowledge clerics. What do you even mean with those two terms? Skill points and skill related abilities? I hope you don't mean the useless fodder such as Hide in Plain Sight (Ranger). Fighters and Barbarians both get significantly less. There's spells on basically ALL spell levels that either enhance or surpass skill checks.



I missed the last thread.... But I do see this is games all the time. Magic always works, and mundanes always get screwed is the default for many games.

Go read some of it maybe :)



I agree here. And if you want to homebrew a ton of fixes then you can.

My big point though, is you don't need to. If you use what is right there in the game rules, you can have a setting that is much less an easy target for magic.

And if you become aware of the free pass given to magic, and put a stop to it cold turkey....you will have a setting that is much less an easy target for magic.

Thank you. I already do homebrew.

It doesn't change that tier 3 and 4 have nothing or next to nothing for most Martials.



But it's not about that? Your taking it to personal.
Why so you claim that? Have you read their mind?




It's a choice. It's not game design or rule design.
Skill system and class features ARE hammer design. They are both lacking in flexibility and in qualitative gains through level progressions.


And once you take that away, no more free pass for magic, you will suddenly see the game work much differently.

Nope, and so far you've provided no evidence to your point. I am not attacking you as a person, just the weakest links in your chain of arguments which is currently evidence that isn't conjecture.

[Necrosnoop110]

Certainly.

I will first describe progression and windows by using encounters as an example:
An Nth level party has a range of encounters that challenge them. This spans from the ones just above guaranteed victory to those just under guaranteed defeat. This span is a window. At different levels the window has different mins and maxs. As the party levels from 1st to 20th, that window will continue to move towards harder encounters. I use the term "progression" to describe how much that window moves. Some games have 0 progression because the window never moves. Other games have the window move a bit but always have first and last level overlap a bit. D&D's default narrative has a rather large progression, the gap between the 1st level window and the 20th level window could be measured in multiple window lengths. This concept of windows and progression applies to more than just encounters. It applies to many subsystems in D&D. Attacks and Ability Checks are two examples.

How does bounded accuracy work:
Attacks are comprised of an accuracy or "attack" roll and a damage roll. Bounded Accuracy is the idea of minimizing the progression in the attack roll without compromising the progression of attacks themselves. They handle this by increasing the progression of the damage roll while shrinking the progression of the attack roll. Basically a 20th level attack action is still much stronger than a 1st level attack action, but both can hit the same target.

Why is this relevant to ability checks (including skill checks):
In 5E the progression of the accuracy of ability checks was also shrunk. However what got increased to maintain the overall progression of ability checks? Nothing. There was no second magnitude roll for WotC to increase. As a result the progression of ability checks shrunk massively out of a simple design error.

What do I mean by anchoring?

Games tend to start at low levels. This means DMs are more likely to assign DCs based on preventing a 1st level character from achieving it too soon rather than on when a character will master (get to 100%) that challenge. If we imagine a reasonable amount of progression, then this bias would have us set the DCs too high for mid-high level characters because the 5E progression was shrunk.
The DC grades (5, 10, 15, 20, 25, 30) have names that encourage you to start with DC 15 as your baseline. But DC 15 starts as a 45%/55% chance and barely increases to 55%/85% as you level up. This bias would have us set DCs a bit too high at low level and massively too high at high levels.
WotC adventures include example DCs. They seem to like DC 20 as a baseline. See previous point.

Now remember these DCs were might *cough* *cough* have been reasonable at low level. However at Tier 2-3 the PC is looking at a bad chance of doing something they could do in Tier 1. That means they expect to fail at doing something that isn't even level appropriate anymore.

How can DMs patch this?
Well simple, recognize bounded accuracy does not work on a 1 roll subsystem. So either make it a 2 roll subsystem or, more commonly, by discarding bounded accuracy for ability checks and increasing the progression in that roll.

Want to increase the distance the window moves? Increase (20th level modifier - 1st level modifier) / RNG size. RNG size being 21 in the case of a d20. Maybe change proficiency to be +1 to +7 instead of +2 to +6. Or maybe even more drastic +4 to +12.
Want to move only the minimum side of the window? Shrink the RNG size with level. Reliable Talent turns a d20 into something like 9+1d11. Imagine the minimum roll was level/3 (round up).
Want to move only the maximum side of the window? Grow the RNG size with level. Probably best to add a bell curve here. Imagine 3d6 at Tier 1 but 5d6 at Tier 3. At one point during development of 5E, proficiency was a die instead of a bonus.


When I wanted to make a skilled based character in 5E I dipped Cleric for Guidance, went Rogue for Expertise, stayed Rogue for Reliable Talent
So I went from 1d20+7 at 1st to 1d20(min 10)+1d4+12 at 11th

MinMax
1st827
20th2841
shift2014

In other words I overspecialized in breaking bounded accuracy, but that was required for the character concept to have growth in skills to match the growth in levels. This is not the normal. Unless you break bounded accuracy, your skills will be obsoleted, but math can fix that.
Wow. Thanks. Relatively new to 5E, at least with play experience, learned a lot here.

So let me ask you this, what do you think of a houserule that has two "proficiency tracks" one for combat +2 to +6 and one for skills that would range from +4 to +12?

[OldTrees1]

You're talking game math, which is fine, but I find your point more interesting in terms of what can be accomplished or rather can't.

For example, a DM would not think it reasonable for a 1st level character to bring a dead character back to life. It takes a 5th level character to do that with Revivify, which has its restrictions of an expensive material component and must be cast within a minute of death so the spell must be prepared if a player wants to use it at the moment it's needed. Therefore there's no way a 1st level character even proficient in Medicine can perform CPR to bring a dead character back to life. The DC is no. When characters are 5th level, it takes a 3rd level spell to bring a character back to life, so how is a 5th level PC to do it by a Medicine check? There's no DC guideline from the rules, so the DM has to make it up but does he want to make it up? A simple die roll should replace a 3rd level spell and a 300 gp diamond? DC remains No. Ninth level comes along and now there's Raise Dead for up to 10 days since death and 500 gp diamond. Maybe now Medicine CPR is fine to duplicate Revivify? Is it free of cost? Revivify is still available and more affordable to be prepared and used. What's the DC? 20? Why wasn't it DC 20 at 1st level then if you can put a DC on it? DC 25? 18 Wisdom and proficient, you have +8 to the roll. Only works if you roll a 17+. Not good odds, just use the spell. However, back at level 1 16 Wisdom and proficient you have +5. You can Medicine CPR on a Natural 20 and have Guidance maybe, Bardic Inspiration too for teamwork, but you as DM didn't want 1st level characters bring the dead back to life, so now the DC must be 30 which the 9th level character can't achieve anymore without a buff, but it's all DM made up. That won't help all the other players in all the other games where the DM won't even consider this. The rules say nothing. You can never bring the dead back to life with Medicine. You don't get greater magnitude of skill use at level 20 you didn't have at level 1.

Great concrete example Pex. This ^ is why the math matters.


Wow. Thanks. Relatively new to 5E, at least with play experience, learned a lot here.

So let me ask you this, what do you think of a houserule that has two "proficiency tracks" one for combat +2 to +6 and one for skills that would range from +4 to +12?

Giving everyone Expertise for skills is a common idea, has merits, but it has 2 complications
1) What about Rogue and Bard? Both have class features intended to break skill expectations and perform higher than their level normally does.
2) Some skills are used in combat. 5E had a reason for shrinking accuracy and increasing magnitude. Do we want Shove to behave similarly?

But a good improvement is still an improvement.

But since it is a houserule, I should also suggest you ask your players. What part of progression matters the most to them? Mastering old possibilities (get to 100% so you don't roll) or achieving new possibilities (being able to hit higher and higher DCs). Giving everyone Expertise applies to both goals equally. If your group values one more than the other, you might bias progress in that direction. Personally I have something like a 75%/25% preference so I would like some mix of Expertise and Reliable Talent baked into the base system.

[Pex]

I'm convinced by experience. Playing my barbarian/fighter/rogue at level 16 once again the DM just creating a party level appropriate encounter the only thing I can do is Dodge. We're fighting Bel and devils from a boat on the River Styx in Gehenna. Too many details to explain, but this is not the first time I do nothing in combat. Lack of flying is a factor. Terrain is a factor despite speed not reduced due to Ring of Free Action. Lack of any effective range attack is a factor, though that's my build specifically. Fighters, rangers, and rogues are not lacking effective range attacks. I'm too mundane for the campaign at this point and/or I'm too specialized for melee tanking. I think it was 14th level when it was starting to get to me. It got real bad in one battle about two months ago where I did absolutely nothing for the entire game session. I love the character, but I want this campaign over with.

The druid is Captain Marvel.

Maybe I'm just emotionally biased ranting right now, but I am really ticked off. The DM didn't do anything wrong.

Out of combat the DM is accommodating with skill use.

[Morty]

Too true, and honestly a bunch of the systems people bring up as better for skills are actually worse for stealth -- GURPS and Hero System, for instance, have charts of penalties for stealth checks, and super-disfavor the poor discovered infiltrator when it comes to escaping alive (Cunning Action is one really nice bone that 5e threw the infiltrating rogue, in that now they can at least disengage and book it back to their friends better).

The proper approach to stealth is to roll once and only call for more rolls when the situation changes. When you fail a roll, it means you draw someone's attention in some way and you're at risk. A lot of this can be handled by the GM, but the system can and should encourage such measures. Which 5E... well, doesn't.


The 'but the other guy/other class I could play has spells' issue is always going to be D&D's albatross. Made worse by how the spells just work. They are relatively sure things/known qualities/quantities -- the infiltrator wizard knows exactly how many times they can _____ (skulking spell) and how likely they are to succeed when using it.

There's a reason 5E gives out spellcasting so widely - there's only four classes that don't have it by default, two of them can get it through a subclass and anyone can get them through feats and racial abilities. It was decided that spells are the only way in which you can do anything cool and interesting. Because that seems to be what D&D means to most players.

[Aimeryan]

Why are you rolling initiative if nobody initiated it?
The Defender only gets to use their Reaction if their Initiative is higher than the Offender. They still don't get to act any sooner, they just get to use a Reaction (if one is relevant to the situation). Someone who is faster should be able to use their Reaction better than someone else who was slower and surprised, no?
That's more of a difference between the AC rules, not really a direct reflection of Surprise rules. Unless you want to start doing Touch AC again. They took that out because simplicity was more valuable than realism.


Although, small correction on bullet 1, the order goes:


Determine surprise.
Establish positions.
Roll initiative.


As a result, Initiative should never be rolled until after the Defenders notice a threat, or when the Offender becomes one.



Exactly - it makes no sense.
Yup, which means the surprised defender was not surprised - eh?
Yeah, don't get me started on AC rules - probably my least favourite part of D&D. 3.5e was better for it than 5e, but still...


Unfortunately, 5e makes you roll initiative even for surprised individuals. This can result in reactions occurring on your turn - including, say, to shout out an alarm. As losing initiative is a very common occurrence, even against the weakest of enemies, it means you cannot reliably take out an enemy with stealth.

In 3.5e, a surprised enemy got no turn in any way until after they had been attacked, which means no shouting out an alarm.

[Man_Over_Game]

Great concrete example Pex. This ^ is why the math matters.



Giving everyone Expertise for skills is a common idea, has merits, but it has 2 complications
1) What about Rogue and Bard? Both have class features intended to break skill expectations and perform higher than their level normally does.
2) Some skills are used in combat. 5E had a reason for shrinking accuracy and increasing magnitude. Do we want Shove to behave similarly?

But a good improvement is still an improvement.

But since it is a houserule, I should also suggest you ask your players. What part of progression matters the most to them? Mastering old possibilities (get to 100% so you don't roll) or achieving new possibilities (being able to hit higher and higher DCs). Giving everyone Expertise applies to both goals equally. If your group values one more than the other, you might bias progress in that direction. Personally I have something like a 75%/25% preference so I would like some mix of Expertise and Reliable Talent baked into the base system.

One interesting piece of game design philosophy is Random Chance vs. Stats.

The more RNG you add, the less your decisions matter, and the better the odds are for the underdog.

Too much RNG, and you're playing Candy Land and you have no decisions to make. Not enough RNG, and now the results are decided by variables that existed before the game even started (such as player experience).

So if you want your stats to matter more, all you have to do is weaken the RNG. Make the game more predictable. This is often done just by using more dice with an appropriate total (which is how Dis/Advantage works).

Might I suggest just using a 3d6 when rolling an an unopposed check? That'd fix almost all of the concerns.

[OldTrees1]

One interesting piece of game design philosophy is Random Chance vs. Stats.

The more RNG you add, the less your decisions matter, and the better the odds are for the underdog.

Too much RNG, and you're playing Candy Land and you have no decisions to make. Not enough RNG, and now the results are decided by variables that existed before the game even started (such as player experience).

So if you want your stats to matter more, all you have to do is weaken the RNG. Make the game more predictable. This is often done just by using more dice with an appropriate total (which is how Dis/Advantage works).

Might I suggest just using a 3d6 when rolling an an unopposed check? That'd fix almost all of the concerns.

I suggest you clarify how a 3d6 addresses my concerns. When I do the math it looks like 3d6 is just an inefficient, indirect, and incomplete approach. My concern is not the candyland effect, although it does criticise that too. My concern is about progression and thus I am expecting solutions that cause the windows to move further across the RNG as the PC progresses from min to max level. This is not to say using 3d6 is not a good idea, just that I do not see it as being substantially related to my concern.

Increasing curve of the bell curve does not move the min and max* so it does not move the window. Since it does not move the window, it will struggle to increase progression. When the goal is progression, there are solutions that directly have the desired goal.
1) Increase the difference in the static bonus from level 20 - level 1. This increases both min and max.
2) Increase the RNG, without rebalancing, as you level. 3d6 -> 4d6 -> 5d6 -> 6d6 or have a bonus die that grows from 1d4 -> 1d12. This puts most of the progression on the max.
3) Introduce a minimum result on the die. 1d20 -> 1d20 (min 5) -> (min 10) -> (min 15). This puts most of the progression on the min.

*although 3d6 is closer to 1d16**, so you are shrinking the RNG size which does increase the ratio of the current +2 vs the RNG. So 3d6 does improve it from +2/20 to +2/16.

**If you don't look at the strict min and max but at the 5% points (2-19 on 1d20) then using a bell curve can shrink the RNG size further. 3d6 ranges from 6-15 for the same probabilities as 1d20 ranges from 2-19. So you could think of 3d6 as only a 1d12 if you discount the bottom at top 5%. I don't think I share this perspective (at the 5% mark), but I mentioned it for completeness because at enough standard deviations it would apply.

[Man_Over_Game]

I suggest you clarify how a 3d6 addresses my concerns. When I do the math it looks like 3d6 is just an inefficient, indirect, and incomplete approach. My concern is not the candyland effect, although it does criticise that too. My concern is about progression and thus I am expecting solutions that cause the windows to move further across the RNG as the PC progresses from min to max level. This is not to say using 3d6 is not a good idea, just that I do not see it as being substantially related to my concern.

Increasing curve of the bell curve does not move the min and max* so it does not move the window. Since it does not move the window, it will struggle to increase progression. When the goal is progression, there are solutions that directly have the desired goal.
1) Increase the difference in the static bonus from level 20 - level 1. This increases both min and max.
2) Increase the RNG, without rebalancing, as you level. 3d6 -> 4d6 -> 5d6 -> 6d6 or have a bonus die that grows from 1d4 -> 1d12. This puts most of the progression on the max.
3) Introduce a minimum result on the die. 1d20 -> 1d20 (min 5) -> (min 10) -> (min 15). This puts most of the progression on the min.

*although 3d6 is closer to 1d16**, so you are shrinking the RNG size which does increase the ratio of the current +2 vs the RNG. So 3d6 does improve it from +2/20 to +2/16.

**If you don't look at the strict min and max but at the 5% points (2-19 on 1d20) then using a bell curve can shrink the RNG size further. 3d6 ranges from 6-15 for the same probabilities as 1d20 ranges from 2-19. So you could think of 3d6 as only a 1d12 if you discount the bottom at top 5%.

It removes a lot of weight on the outliers. Makes your success rates based more on your bonuses. Then increase the value of those DCs, so that DC 15 is more heroic, and DC 30 is god-levels of power.

The 3d6 binds you to only fail on the things that are truly out of your scope, and to succeed on everything else.

In order to hit that DC 30 5% of the time, a 1d20 needs a +10 bonus, while the 3d6 would need a +14.
In order to miss a DC 10 5% of the time, a 1d20 needs a +8 bonus, while the 3d6 would need a +4.

So from my perspective, a 1d20 as the same difference between the minimum and maximum DCs across 2 points as the 3d6 would have across 10. Each bonus point you have carries distinct weight.



I guess the question I need to ask is, what do you consider your definition of progression? Succeeding a DC 20 50% of the time when it used to be 25%? Rolling twice as higher of a number than I did a few levels ago? Never having to roll for most checks with my best skill?

[OldTrees1]

I guess the question I need to ask is, what do you consider your definition of progression? Succeeding a DC 20 50% of the time when it used to be 25%? Rolling twice as higher of a number than I did a few levels ago? Never having to roll for most checks with my best skill?

3d6 does increase consistency. I will readily recognize it as being a useful lever to toggle based on preferences. I was talking about progression instead (explained below)


My definition of progression:
You have a range of DCs from the one you just barely can't fail to the one you just barely can't pass*. This is your window.
Leveling will move that window. It will bring new abilities into the possible and let you master (100%) old abilities.
That is progression.

How much progression do I expect? Depends on the level range the sub system is used for (1-20 in this case) and how much progression is expected in level appropriate abilities in that system. In 5e it depends on which subsystem you use as your metric. If I use encounters, then I expect the window to move twice its length. However I understand if people only wanted it to move its length and I would begrudgingly compromise for 75% of its length.

Your specific questions
1) A DC 20 going from 25% to 50% on a 1d20 would be the window moving 25% of its length because 5/20=0.25 (and 12.5% on 3d6 because 2/16=0.125).
2) Not sure what exactly you mean here. Hopefully my detailed answer covered it.
3) Never having to roll for most checks in your specialty? Depends on playstyle? A Tier 4 character doing lots of Tier 2 stuff would not have to roll. A Tier 4 character doing Tier 4 stuff generally would roll. Which end of the window does your character play at the most?

*Some will round after a certain point. A 1 in 1e100 chance (picked for being hyperbolic and uncontroversially small) is close enough to 0 for most. Where that point is I am not sure. I don't start rounding at 5%. However for completeness I will say I would start rounding eventually.

[Telok]

Well you can work on quantifying things using arcana and the rules to identify what spell is being cast. At what point should a wizard be able to automatically identify a spell that the wizard has been casting since level 1? Should it be the same as the wizard automatically identifying a 1st level druid spell? Should it be dependent on intelligence, proficiency, level, total modifier?

[5eNeedsDarksun]

I'm convinced by experience. Playing my barbarian/fighter/rogue at level 16 once again the DM just creating a party level appropriate encounter the only thing I can do is Dodge. We're fighting Bel and devils from a boat on the River Styx in Gehenna. Too many details to explain, but this is not the first time I do nothing in combat. Lack of flying is a factor. Terrain is a factor despite speed not reduced due to Ring of Free Action. Lack of any effective range attack is a factor, though that's my build specifically. Fighters, rangers, and rogues are not lacking effective range attacks. I'm too mundane for the campaign at this point and/or I'm too specialized for melee tanking. I think it was 14th level when it was starting to get to me. It got real bad in one battle about two months ago where I did absolutely nothing for the entire game session. I love the character, but I want this campaign over with.

The druid is Captain Marvel.

Maybe I'm just emotionally biased ranting right now, but I am really ticked off. The DM didn't do anything wrong.

Out of combat the DM is accommodating with skill use.
The more I think about this, the more I keep coming back to two ideas I threw on the table a while ago. 1: adjust HP: drop full casters down 1 category, 1/2 and 1/3 casters stay the same, and add 1 bonus HP to non-casters. 2: Allow a 4th magic item attunement which prevents concentrating on a spell.
Thinking back to my experience in 2e, a d4 wizard would average 3hp per level +2hp Con bonus max for 5 hp/ level. My current sorcerer gets 4hp + 1hp (Dragon origin) + 3hp from a 16 Con for 8 hp/ level. 50% more hp while the Fighters and Barbarians are stuck with the same makes a huge difference. Carrying around nearly the same hp as the melee characters makes a huge difference with how I'm able to play my character. I can use the vast majority of my resources aggressively without worrying too much about getting smashed and the DM has to balance encounters accordingly.
On the 2nd point, 1 more attunement is substantial; in the example it sounds like Winged Boots would be a gamechanger.
I'm sure these ideas aren't the magic bullet, but they'd be a pretty simple step in the right direction.

[Pex]

The more I think about this, the more I keep coming back to two ideas I threw on the table a while ago. 1: adjust HP: drop full casters down 1 category, 1/2 and 1/3 casters stay the same, and add 1 bonus HP to non-casters. 2: Allow a 4th magic item attunement which prevents concentrating on a spell.
Thinking back to my experience in 2e, a d4 wizard would average 3hp per level +2hp Con bonus max for 5 hp/ level. My current sorcerer gets 4hp + 1hp (Dragon origin) + 3hp from a 16 Con for 8 hp/ level. 50% more hp while the Fighters and Barbarians are stuck with the same makes a huge difference. Carrying around nearly the same hp as the melee characters makes a huge difference with how I'm able to play my character. I can use the vast majority of my resources aggressively without worrying too much about getting smashed and the DM has to balance encounters accordingly.
On the 2nd point, 1 more attunement is substantial; in the example it sounds like Winged Boots would be a gamechanger.
I'm sure these ideas aren't the magic bullet, but they'd be a pretty simple step in the right direction.

Spellcasters having hit points is not a problem. That was never the problem. The problem in my game is I can only do one thing. I do it really, really well, but when I can't do it because the encounter doesn't make it available I can't do anything else. I don't resent the druid being Captain Marvel. I resent I can't do anything more than Hulk Smash. The other martials aren't as bad off as me. The Shadow Monk has maneuverability and spells being Night Crawler. The Horizon Walker Ranger has his arrows to attack anywhere and spells being Hawkeye. The Swashbuckler Rogue doesn't get hit and sneak attacks in melee or range but is really the party's face being Black Widow. The Wild Magic Sorcerer has her spells being Iron Man. The Forge Cleric has his spells and attacks being Thor. Maybe my problem is being more one trick pony than a martial, but I'm not built for range even if I tried and it's too late at this level to fix it. I lack versatility which means the only thing I get to do sometimes is Dodge.

[5eNeedsDarksun]

Spellcasters having hit points is not a problem. That was never the problem. The problem in my game is I can only do one thing. I do it really, really well, but when I can't do it because the encounter doesn't make it available I can't do anything else. I don't resent the druid being Captain Marvel. I resent I can't do anything more than Hulk Smash. The other martials aren't as bad off as me. The Shadow Monk has maneuverability and spells being Night Crawler. The Horizon Walker Ranger has his arrows to attack anywhere and spells being Hawkeye. The Swashbuckler Rogue doesn't get hit and sneak attacks in melee or range but is really the party's face being Black Widow. The Wild Magic Sorcerer has her spells being Iron Man. The Forge Cleric has his spells and attacks being Thor. Maybe my problem is being more one trick pony than a martial, but I'm not built for range even if I tried and it's too late at this level to fix it. I lack versatility which means the only thing I get to do sometimes is Dodge.
Captain Marvel, Iron Man, and Thor can take a beating and keep going. What if they couldn't? And what if Hulk had a magic item that let him get where he needed to go?
I mean, if you're unhappy with your build, I guess you're unhappy with your build and maybe could have devoted some resources to range attacks. I will say we houserule that you can draw and throw any weapon as many times as you have attacks, so our strength based characters tend to have a quiver of javelins hanging around, making their useful (but not optimal) range 120'. For the life of me I have no idea why RAW archers get this benefit, but not others. Anyway it sounds like the fix to you being effective, if not interesting, isn't that complicated since, as you say, you can do your one thing really well.

[Xervous]

I agree there could have, should have been more guidance on what each tier of play entails and more focus on making sure each class has options that allow them to engage with expected challenges at that tier. Leave players free to make a fireball only wizard or a barbarian who just hits really hard, but don’t make them the only options.

[Eldariel]

I'll say this. Trying to fix caster HP by lowering HD is just silly. If anything, PC HP is too low for almost all characters: a single crit from high dice attack can generally drop a level appropriate character not using shapeshifting.

If you want to do something meaningful, just give warriors more HD for each level, multiply their Con modifier or something. Currently temporary HP, feats, and Con modifier are the primary sources of HP. For warriors to gain meaningful advantages, they need more of those (or more resistances).

[Xervous]

I'll say this. Trying to fix caster HP by lowering HD is just silly. If anything, PC HP is too low for almost all characters: a single crit from high dice attack can generally drop a level appropriate character not using shapeshifting.

If you want to do something meaningful, just give warriors more HD for each level, multiply their Con modifier or something. Currently temporary HP, feats, and Con modifier are the primary sources of HP. For warriors to gain meaningful advantages, they need more of those (or more resistances).

I’ve hardly ever seen hp pointed to as a problem outside of low levels where everyone is a hot sneeze away from a funeral and even then 5e has nice kiddie glove buffers warding off death. The issue with Martials is not a numbers game, it’s a lack of buttons they can push. 3.5e Martials could win at the numbers game (effectively making the initiative roll a save or die opposed skill check) but that was only a fraction of the bigger picture still. Hp doesn’t let the fighter fly, it doesn’t let him recruit twelve orphans to run surveillance on the corrupt baron, nor does it let his reputation precede him at a banquet.

[Eldariel]

I’ve hardly ever seen hp pointed to as a problem outside of low levels where everyone is a hot sneeze away from a funeral and even then 5e has nice kiddie glove buffers warding off death. The issue with Martials is not a numbers game, it’s a lack of buttons they can push. 3.5e Martials could win at the numbers game (effectively making the initiative roll a save or die opposed skill check) but that was only a fraction of the bigger picture still. Hp doesn’t let the fighter fly, it doesn’t let him recruit twelve orphans to run surveillance on the corrupt baron, nor does it let his reputation precede him at a banquet.

I think the fact that classes have about equivalent HP is an issue, if not a big one. Yeah, it's not a qualitative issue, but it feels a bit silly that supposedly squishy caster can easily have more HP and EHP than a guy who is made to take hits (Fighter/Paladin for example). I definitely think they could use a bit more though far as the issues in this comparison, this is maybe 10th in importance.

Mostly just pointing out that I feel the suggestion few pages back about dropping caster HD is just pretty pointless and accomplishes approximately nothing.

EDIT: This is a subset of the "warriors not [sufficiently] better than casters in combat to make up for their overall deficiencies". Warriors feel like they SHOULD be the best at taking hits but fact is, there's hardly any difference