So... is 5th better or worse than the '3's, or merely different...

[calebrus]

(I do not have the 5e experience to say exactly where those numbers break down at,

That's all you had to say.

[Battlebooze]

Absolutely. I am a kingdom-saving dragon-slaying knight, after all.

Also, how many olympic-level humans do you think exist in the world? Because I get the impression that you're vastly underestimating this amount.

Spoiler: Answer

Show

It's about 13,000, in case you were wondering, plus several thousand more if you count similar world-class events like the Wimbledon, Tour de France, or the Ironman.

Also include most soldiers in worldwide special forces and of course, professional athletes.

[ChubbyRain]

Also include most soldiers in worldwide special forces and of course, professional athletes.

Yeah even the D-League NBA athletes are super fricken in shape compared to the common person.

A lot of the time it isn't even skill that keeps a player in the minor leagues or whatever, it can be mental or attitude or even contract control.

[Battlebooze]

I've played tons of 3.X and I still enjoy it. I would have no problem going back to playing a 3.x game.

So far though, I think I'm enjoying 5.0 a little more. Of course I haven't played 5.0 for nearly as long, so my opinion could change over time.

I like 5.0's simplified skills and the improved scaling of Armor class. Spells casters are still very powerful, but martial types are doing better so far. I don't see as many traps to fall into that would lead you to make a mostly useless character. Over all, in my opinion 5th is better.

[calebrus]

Absolutely. I am a kingdom-saving dragon-slaying knight, after all.

Also, how many olympic-level humans do you think exist in the world? Because I get the impression that you're vastly underestimating this amount.

Spoiler: Answer

Show

It's about 13,000, in case you were wondering, plus several thousand more if you count similar world-class events like the Wimbledon, Tour de France, or the Ironman.

The world's population is 7.125 billion.
Your 13,000 equates to 0.00018% of the population.
So yes, that's what I'd consider a handful, comparatively.
Less than a handful, to be honest.

If you are better than 99.999% of the population, then you are one of the best in the world. And you think that's appropriate at level 8.

[Kurald Galain]

The world's population is 7.125 billion.
Your 13,000 equates to less than 2% of the population.

So what? The amount of kingdom-saving dragon-slaying knights also equates to (substantially) less than 2% of the population.

By the time I'm level 15, my character is literally saving the world, because that's what happens in heroic fantasy settings. He is literally one of the ten most powerful characters in the world, and most of the other nine are either his party members, his mentor, or the BBEG. How on earth do you figure that my character still needs to be weaker than an Olympic athlete, or Special Forces agent, or NBA star? That doesn't even make sense to me.

[obryn]

The world's population is 7.125 billion.
Your 13,000 equates to 0.00018% of the population.
So yes, that's what I'd consider a handful, comparatively.
Less than a handful, to be honest.

Is this relevant? The PCs are, by mid-high levels, among the wealthiest, most influential, and most powerful people in the fantasy world.

They also get more screen time than everyone else in the setting combined, by virtue of being actually played by the players.

[calebrus]

So what? The amount of kingdom-saving dragon-slaying knights also equates to (substantially) less than 2% of the population.

I missed a button. Check it again.

[Kurald Galain]

I missed a button. Check it again.

You also missed the point. Check it again.

[mephnick]

By the time I'm level 15, my character is literally saving the world, because that's what happens in heroic fantasy settings.He is literally one of the ten most powerful characters in the world, and most of the other nine are either his party members, his mentor, or the BBEG.

To be fair this is setting dependent. In Forgotten Realms this is completely false. Also a flat skill system doesn't prevent you from being one of the 10 best in the universe. Being able to do something great 20% more often is a pretty huge deal in our world. The strongest guy in the world isn't 10x as strong as your average 25 year old male.

[calebrus]

So why even have skill checks? You just tell the DM what happens, and then you drink your Mountain Dew and grin.
While we're at it, why even have a DM. You obviously don't need him, because you should never fail at a task.
While we're at that, why even play? You obviously aren't playing for any sense of excitement, because you can never lose.
Just write a novel proclaiming the deeds that your character achieved. That would probably be more satisfying for you than playing a game where you might fail a climb check and not get to the top of that pole.

[Battlebooze]

Look at the basic stat distribution!
Player characters really are the equivalent of professional athletes/rock stars/super scientists/religious icons. By level 8, they can have one stat at 20! That is as strong, as smart, as wise, as charismatic as any human will ever get.

They certainly aren't regular citizens.

[Xetheral]

The lazy noble isn't in the adventuring party. The rules are for adventurers. What the lazy noble can or cannot do is entirely up to the DM.

5e has two equally-valid systems for designing NPCs. You can build them as monsters or you can build them as PCs. In both cases, the bonus on a skill check is determined by (Proficiency, if applicable) + (Attribute Modifier). In either method, the lazy, untrained noble has the same chance to succeed at a given task as an untrained adventurer with the same attribute score.

Yes, the DM can always go outside the system and add houserules to give the NPC miscellaneous bonuses or penalties in order to achieve a preferred result. Or the DM could houserule that, despite having the same bonus, the noble needs to roll while the adventurer autosucceeds, or that the noble automatically fails while the adventurer gets to roll. But the DM can always do that, in any edition. What's relevant to the comparison of 5e and 3.5 is the extent to which a DM needs to adapt the system to achieve that preferred result.

IMO, the important number is not the absolute adjustment, but the reduction in failure rate. It's the failure rate which the adventurer is concerned with.

if you're trying a DC 5 task, the untrained +0 has a 20% chance of failure. The +3 has a 5% chance of failure, and the +5 has a 0% chance of failure. In terms of decision making, the task is daunting but doable for the untrained, mostly safe for the proficient person (75% reduction in failure rate) and totally routine for the experienced (100% reduction in failure rate).

It's perfectly acceptable to look at the failure rate instead of success rate (in a binary system, after all, they are inverses). However, you can't just look at the proportional change in failure rates... you also have to consider the absolute change to get the full picture. Going from +0 to +3 does indeed drop the chance of failure vs DC 5 proportionately by 75%. But it's just as important that the failure rate only decreased by 15 percentage points. Sure, you might only fail only 1/4 as frequently with the higher bonus, but that higher bonus still doesn't help you at all 85% of the time (i.e. unless you roll a natural 2, 3, or 4 on the d20, the +3 didn't help).

If your DM is properly using DCs of 5-15 on a regular basis then the skill system works just fine. IMO the main problem is that 3.5 experienced DMs who consider such DCs to be a total joke. If anything worth doing is DC20 or higher than skills are a joke because they'll never be worth it...

IMO, you guys really are just going in with preconceptions of what DCs skill checks ought to be. I honestly believe that you are looking at the skill bonuses and comparing to DC20 and thinking "What's the point? This is useless."

DC 5 is not laughable baby-task in 5e - it's a legitimate difficulty threshold. DC5 of falling into a deadly pit means that a skilled or high-attribute character can complete it routinely, but those lacking such bonuses face sufficient risk that a safer means of completing the task should be sought.

Furthermore, setting DCs is by far the _easiest_ way to scale the world to your expectations. It can be done totally on the fly to reflect your expectations of what characters can do.

You can certainly scale the absolute chance of success across the board by choosing to use lower (or higher) DCs. But it doesn't change the fact that the difference between a trained and untrained character is still far overshadowed by the random element of the d20.

[obryn]

So why even have skill checks? You just tell the DM what happens, and then you drink your Mountain Dew and grin.
While we're at it, why even have a DM. You obviously don't need him, because you should never fail at a task.
While we're at that, why even play? You obviously aren't playing for any sense of excitement, because you can never lose.
Just write a novel proclaiming the deeds that your character achieved. That would probably be more satisfying for you than playing a game where you might fail a climb check and not get to the top of that pole.

Oh, well this is definitely a sterling contribution to the discussion. A+++.

To be fair this is setting dependent. In Forgotten Realms this is completely false. Also a flat skill system doesn't prevent you from being one of the 10 best in the universe. Being able to do something great 20% more often is a pretty huge deal in our world. The strongest guy in the world isn't 10x as strong as your average 25 year old male.

It's not just (or even mostly) the success rate. It's the failure rate. Reliability. How reliably is a high-level adventurer able to perform a task that is DC 25? Or even DC 20? And how much better are they at it now than they were at 1st level?

[Kurald Galain]

To be fair this is setting dependent. In Forgotten Realms this is completely false. Also a flat skill system doesn't prevent you from being one of the 10 best in the universe.

Yes it does, because a flat skill system means that the so-called best in the universe will lose about 14% of the time to a random average nobody (such as the untrained couch potato wizard in the PC party). Like I said before, you can't fix the issue by changing the labels on the numbers.

Being able to do something great 20% more often is a pretty huge deal in our world. The strongest guy in the world isn't 10x as strong as your average 25 year old male.

Actually, yes, the strongest guy in the world is pretty close to ten times as strong as your average 25 year old male.

You probably weren't expecting that but yes, those are the numbers

(edit) source: clean & jerk record in the Guinness book.

[Fwiffo86]

Meh, both skill systems (3.x and 5) are flawed for various reasons. Some people like one or the other. Me personally, I feel that the 3.x skill system was completely ridiculous. When you are assigning DCs to do acrobatics checks to walk on air, you have jumped the shark. Assigning DCs to gauge every conceivable action is also likewise silly (eyes GURPS angrily).

It seems that many people feel that this should be (or an approximation) the rule in 5e. I am glad its not. A skill check is nothing more than a way to resolve a completely mundane action. Even the epic heroes in fantasy had problems with things outside of their actual fame. Hercules for example was fantastically strong, but was a terrible person. (represented by supreme STR checks, terrible CHA checks). In a more modern comparison Goku is a combat genius (superior combat stats across the board) but is a complete moron (social skills of none other than he's a nice guy).

The expectation that ALL skills should exceed a target situation by a specific level is silly. I get this impression because no one has said anything about having a select number of skills exceed these levels, but not all of them.

Traditionally, the skill system of previous editions (before 3.x) have been workable but bad. Bad because they were added as a second thought, in an attempt to make the game more indepth. Quibbling over how your skill system drags down the rest of your character is also silly, since the skills are not the defining characteristics of that character.

5e's greatest merit is that everything stays relative at all levels, despite people slinging around reality warping magic and fighting dragons. That's why I have dumped 3.x in favor of 5e, and I am never looking back.

[Vogonjeltz]

...and? We're talking of similarly-leveled characters having problems with climbing a rope. And how things that used to be Hard are now just ... Moderate.

Yes, and taking my comment in context reveals that I'm saying that wish is not the appropriate example nor a reasonable baseline to use for why tasks that are actually hard should suddenly not be hard.

So my point is that pointing to the perception of wish (which actually reflects a 3.5 edition wish, not the 5th edition wish) and then asking why those who don't cast wish can't do thing X is misguided.

I have more but first let's bring in Kurald's post that relates to it:

For almost everyone in almost every campaign, the number is not +23 but something like +6 (four points from training, two from a non-primary attribute). And that's just not very impressive.

No, it's not meant to be either. The intention of the skill system is that it reflects those tasks that are difficult enough as to require a roll in the estimation of the DM, yet still remain within the realm of the possible.

A two prong test:
1) Is this action possible? (Yes/No)
2) Is this action difficult enough that auto-success is not a thing? (Yes/No)

Answer: Pick a DC

Very Easy
Easy
Medium
Hard
Very Hard
Nearly Impossible

I think where this might break down is the perception of what these descriptive terms actually equates to. My understanding is that this is meant to be an absolute scale, in other words, the Nearly Impossible is really just that, only achievable by your average character with a great roll and a very large passive bonus (i.e. a 19-20 with +6 proficiency and +5 stat bonus). It's also intended as a term for an activity where literally the best of the best are expected to fail almost every time.

Instead I think there is error introduced when reading the DCs and expectations of some subjective scale come into play. It's not Subjectively easy, it's Objectively easy. a DC 10 is objectively easy, not subjectively easy. Even an easy task (hit the center of the target) can be flubbed, unless of course the person is proficient and has a good ability score (in which case they never fail at it).

...You're joking, right? Right?

Okay, fine, I'll repeat myself one last time.

Attack rolls don't increase much over 20 levels. Skills also don't increase much over 20 levels. The lack of increase is fine for attack rolls and not skills because of the context each is in, in-game. Because hit points and damage scale, attack rolls don't need to scale nearly as much.

The rolls scale, they get objectively better with level.

It's not just that you don't auto-succeed. You never even get into success chances which are all that great for anything that's particularly difficult. The difference between a novice and a master is pretty tiny.

The difference in success between someone trained and someone untrained is up to 30%, 60% for those who have expertise, assuming both have the same innate ability and skill levels.

The difference between someone who is untrained and has no real ability and someone who does is up to 60%, 90% for those with expertise.

I'd also hasten to mention, it must not be a really difficult task if literally anyone could do it all the time. That would be a very easy or easy task.

With a 16-17 starting stat and proficiency in a skill, a natural 20 results in the character being able to, with a great deal of luck, make a dc 25 check numerically (technically you rolled a 20 so you win anyways,

This is a relic of the 3.5 era, in 5th edition a 20 isn't necessarily a success.

In the "jump a chasm" example, the DM might let you grab onto the ledge if you miss the athletics check by 5 or less.

Also a 3.5 relic, there's no check required to jump. Jumping distance is just a function of your strength score (and if you moved before the jump or not). So if you are strength 20, you can just automatically succeed at jumping 20 feet. The skill description certainly implies that someone can try to make an athletic check to jump a longer distance or perform a stunt while jumping, but the standard jump is checkless and an auto success.

Just from this table, we can see that with skills the easiest task is downright constant, with damage there is a 4.8 fold increase across 20 levels. On the hard end, skills get a 7 fold increase across all levels, attacks get a 33.6 fold. There's effectively a constant which has to do with how much attacks grow relative to skills, which increases by level. This is also a conservative estimate, in practice I'd expect the constant to be higher.

The damage increases proportionally at rates identical to that of the skills in terms of success rate.

(i.e. @20, 4080 is 85% of 4800; 2880 is 60%; 1680 is 35%, the mix 3696 is 77%, identical to that of the skill check success rate).

Damage is no more successful than skills at any point against the same task DC.

Right, but they want checks and DCs to scale. They can't understand that the checks and DCs don't scale.... AT ALL.... which is why bounded accuracy works.

From the Article on Bounded Accuracy. Relevant portions have been bolded:

Excellent catch, that answers everything.

Now any of us that have an IRL job that's skill-related are being paid precisely for performing difficult tasks successfully. So that means that in 5E terms, we're all 18th-level rogues...

And in D&D 5e terms those tasks don't rate as even Hard, but perhaps just Average or even Easy. These are objective descriptions, not subjective. If the task was really a Hard one, then even the best of the best have a chance of failure, and although they regularly succeed, there's the possibility of failure. Someone who's not the best would have a chance at success, but regularly fail.

You're severely underestimating how many people are actually good at something. Someone with a masters degree and five to ten years of relevant job experience? You'd better believe that someone like that is capable of performing difficult tasks successfully.

Nobody is saying they aren't capable of performing Hard tasks successfully, but they aren't going to do it every time without error. For every Cuchulainn there's a Cuchulainn who failed to identify and kills his own son. Oops, an easy task no less.

[Seerow]

That's all you had to say.

And yet you believe that the Fighter should only be facing a single level 1 creature at level 8, because bounded accuracy. I would hazzard a guess you have even less experience.

Dismissing a long well reasoned argument simply because I have not played through a half dozen 1-20 campaigns so I can eyeball without doing a lot of math exactly how many creatures a character at a given level can take on is ridiculous.

[MadBear]

See, the problem with this idea is that now every first level character is a top-tier athlete, and the untrained couch potato wizard has a 30% chance of succeeding at an Olympic task.

You can't fix this issue by putting different labels on the same numbers.

Depends entirely on what you want to accomplish.

What I'm describing in my idea is that even the str 8 wizard is more athletic because of his adventuring days then the average commoner. In this situation there are no couch potatoes.

What you seem to not like is that there isn't a huge degree of seperation between classes. This is true, and isn't modeled by the game. The game assumes that every hero should have at least a shot at accomplishing a easy-hard skill, with those better trained accomplishing it more routinely, and by the end, an almost impossible task has the smallest chance of success by a highly trained hero.

Now, my idea fixes the issue in that a character can accomplish Olympic level feats. In that goal it works perfectly and fits with the DM decided level of skill power in the game. My fix has the side effect of making all heroes are now inherently better then common people. But at the end of the day, a DC 15 task at level 20 is something that a wizard will only make 30% of the time while the fighter succeeds 85% of the time (higher if their a champion).

Now if you want to define the problem in such a way that my idea no longer fixes the problem, that's fine. Just remember that's not the problem I was addressing, and your concern (if it is in fact about the degree of seperation between characters) isn't something that I think is important to me. I'm ok with Gandalf having a small shot at accomplishing a fairly remarkable atheletic feat, that Aarogorn could trivialize routinely.

[Seerow]

Depends entirely on what you want to accomplish.

What I'm describing in my idea is that even the str 8 wizard is more athletic because of his adventuring days then the average commoner. In this situation there are no couch potatoes.

This completely contradicts the majority of people who insist they like 5e because at level 1 they are normal people, and grow into being heroes.

You are straight up saying the majority of people on your own side of the argument are wrong, and heroes are awesome olympic tier athletes straight from level 1, and that is why they don't need to improve meaningfully.



This is ignoring the fact that we don't actually have a different set of skill rules for NPCs who aren't adventurers, or even anything in any book stating what you suggest might be the case.

[calebrus]

And yet you believe that the Fighter should only be facing a single level 1 creature at level 8, because bounded accuracy.

So me a favor and link to the post where I stated this.
I'll wait.

[Kurald Galain]

What you seem to not like is that there isn't a huge degree of seperation between classes. This is true, and isn't modeled by the game.

What I mean is, let me quote Xetheral again,

• Training in a skill doesn't provide a particularly noteworthy mechanical advantage over untrained characters. At low levels the mechanical difference is negligible. At high levels the difference becomes mechanically noticeable, but remains quite small compared to many players' expectations.
• Expertise, which helps address the previous problem, is only available to two classes.
• Bounded accuracy means that mid- and high-level characters are sharply limited in what can reliably be accomplished with skills in comparison to other editions.
• Extreme cases can be better modeled under the 5e rules by not requiring a roll and simply determining success and failure. And yet there is no guidance on what qualifies as an extreme-enough case. This is even more important than individual DM DC preferences, because the choice of when not to require a roll creates a discontinuity in the probability table. Characters whose bonuses relative to the DC fall near this discontinuity may have vastly different levels of mechanical effectiveness, depending solely on when a given DM decides a bonus is big enough to not require a roll.

Now, my idea fixes the issue in that a character can accomplish Olympic level feats.

No, it doesn't. The only thing you've done is redefined "Olympic level feat" as "something that most people can do". Because if you say "DC 15 is Olympic level" then that precisely means that every character in the world can do it 30% of the time, and that's almost the exact opposite of what "Olympic" means. Once more, you really can't fix this issue by changing the labels around.

[Seerow]

So me a favor and link to the post where I stated this.
I'll wait.

I'm sorry, you said that dealing with 6 level 1 mooks was a deadly threat for a level 8 character. A slightly different argument, but downplaying character capability significantly.

But apparently all you have to offer to back up your claim is dismissive one line posts. That's pretty telling.

[Battlebooze]

So why even have skill checks? You just tell the DM what happens, and then you drink your Mountain Dew and grin.
While we're at it, why even have a DM. You obviously don't need him, because you should never fail at a task.
While we're at that, why even play? You obviously aren't playing for any sense of excitement, because you can never lose.
Just write a novel proclaiming the deeds that your character achieved. That would probably be more satisfying for you than playing a game where you might fail a climb check and not get to the top of that pole.

I think I can explain why players succeeding most of the time at "normal" uncontested skill checks isn't a big problem at all.

What is more interesting, a player failing to climb that mythical pole, or a player succeeding and then having to discover what's up at the top?

Secret doors for example, are they more interesting to the story when found, or not?

Most of the time, successes leads to more options for interesting play.

Real failure often comes from not trying. Don't check for traps because we think the door is safe. Don't try to climb the pole because we don't think anything interesting is up there. Don't check for secret doors, this is just a boring room.

Even a character who succeeds 100% can still make bad choices and make mistakes. They can chose the wrong friends, they can be in the wrong place at the wrong time, they can pick fights against enemies just as good as they are.

[MadBear]

This completely contradicts the majority of people who insist they like 5e because at level 1 they are normal people, and grow into being heroes.

You are straight up saying the majority of people on your own side of the argument are wrong, and heroes are awesome olympic tier athletes straight from level 1, and that is why they don't need to improve meaningfully.



This is ignoring the fact that we don't actually have a different set of skill rules for NPCs who aren't adventurers, or even anything in any book stating what you suggest might be the case.

Whoa there. I was suggesting that by modifying the DC's you can easily represent heroes capable of performing olympic level feats of skill.

I didn't say or mean to suggest that this is how the game is "meant" to be played. The skill system is in the hands of the DM and they can tweak it to represent heroes going from Commoner-->Heroic or Heroic--> super heroic the way that some have said they wanted.

You're also right that NPC's don't have a set of skill rules. That's because what they do is up to the DM and not the skill check system. You can see it in the design philosophy that has NPC's not getting PC class levels to build, and completely misses my point.

[Fwiffo86]

What I mean is, let me quote Xetheral again,

• Training in a skill doesn't provide a particularly noteworthy mechanical advantage over untrained characters. At low levels the mechanical difference is negligible. At high levels the difference becomes mechanically noticeable, but remains quite small compared to many players' expectations.
• Expertise, which helps address the previous problem, is only available to two classes.
• Bounded accuracy means that mid- and high-level characters are sharply limited in what can reliably be accomplished with skills in comparison to other editions.
• Extreme cases can be better modeled under the 5e rules by not requiring a roll and simply determining success and failure. And yet there is no guidance on what qualifies as an extreme-enough case. This is even more important than individual DM DC preferences, because the choice of when not to require a roll creates a discontinuity in the probability table. Characters whose bonuses relative to the DC fall near this discontinuity may have vastly different levels of mechanical effectiveness, depending solely on when a given DM decides a bonus is big enough to not require a roll.

No, it doesn't. The only thing you've done is redefined "Olympic level feat" as "something that most people can do". Because if you say "DC 15 is Olympic level" then that precisely means that every character in the world can do it 30% of the time, and that's almost the exact opposite of what "Olympic" means. Once more, you really can't fix this issue by changing the labels around.

Perhaps this.....

Characters (and NPCs) without the actual skill always roll with disadvantage.

[Seerow]

Whoa there. I was suggesting that by modifying the DC's you can easily represent heroes capable of performing olympic level feats of skill.

I didn't say or mean to suggest that this is how the game is "meant" to be played. The skill system is in the hands of the DM and they can tweak it to represent heroes going from Commoner-->Heroic or Heroic--> super heroic the way that some have said they wanted.

What about people who want to go from Commoner->Super Heroic? Because that is what the game should handle. It was able to handle it in past editions. So why is it now in 5e that we only get a third of the expected progression? And why is it that players are lauding this much narrower focus? I can only assume it is because some players are selfish, and the narrower range the system now covers happens to overlap with what they preferred.

You're also right that NPC's don't have a set of skill rules. That's because what they do is up to the DM and not the skill check system. You can see it in the design philosophy that has NPC's not getting PC class levels to build, and completely misses my point.

Woah there. So NPCs now run entirely on fiat for anything they attempt to do that is not combat related?

In what way is that better than 4e's system, where disassociation was decried as such an awful thing?

No. I don't believe this. NPCs run on the same skill system as PCs, because in absence of an alternate mechanic (for example NPC stat blocks work as an alternative mechanic to class levels for NPCs) they have to follow the same rules, or nothing functions.

[calebrus]

I'm sorry, you said that dealing with 6 level 1 mooks was a deadly threat for a level 8 character. A slightly different argument, but downplaying character capability significantly.

But apparently all you have to offer to back up your claim is dismissive one line posts. That's pretty telling.

You're the one that said a half dozen mooks was a cake walk, and that even a dozen was reasonable.
But you also said that you don't have experience, so you really didn't know what you were talking about.
Well, I do have 5e experience, and I'm telling you that half a dozen mobs is a threat at level 8, and that a dozen will likely kill you.
It's all about the number of incoming attacks.
Your AC doesn't scale. 18 is amazing at level 1, and it's still good at level 20.
But you know what? The mooks can hit it.
So those 6 or 12 attacks every round are going to add up, and eventually overcome you. It becomes an arms race. They are still a threat, and the more of them there are, the bigger the threat.
That's bounded accuracy. It makes the old adage of safety in numbers actually true in 5e when it never was previously.

edit:
Someone did the math and found that if you populate an entire encounter with mobs 3 levels lower than the party, it becomes a TPK threat simply because of the number of incoming attacks.
That's not a threat for a solo character. That's a real threat for an entire party to wipe.

[Kurald Galain]

Perhaps this.....

Characters (and NPCs) without the actual skill always roll with disadvantage.

That would certainly help a lot, yes.

[Fwiffo86]

So those 6 or 12 attacks every round are going to add up, and eventually overcome you. It becomes an arms race. They are still a threat, and the more of them there are, the bigger the threat.
That's bounded accuracy. It makes the old adage of safety in numbers actually true in 5e when it never was previously.

edit:
Someone did the math and found that if you populate an entire encounter with mobs 3 levels lower than the party, it becomes a TPK threat simply because of the number of incoming attacks.
That's not a threat for a solo character. That's a real threat for an entire party to wipe.

to add to this....see threads about the Skelly bros.