D&D General Explain Bounded Accuracy to Me (As if I Was Five)

Because the full variance is used over time. And you also, at least relative to my experience, seem to be downplaying the variance in play at any given level. A 5th level PC still encounter 3rd level creatures. An 8th level PC encounters both 6th and 12th level creatures. Etc.

The scaling is what generates the working fiction of all this. An 8-level gap between standard and minion roughly corresponds to the level range of creatures in play at any given level. (From around L-2 up to say L+4.)
None of this makes sense to me. Nothing about this requires the numbers to scale so rapidly. Nothing.

All it does, is necessitate changing the attack bonuses and defences of the monsters, to keep up with the treadmill, in the process obscuring what the numbers actually represent.
 

log in or register to remove this ad

EzekielRaiden

Follower of the Way
This is really not even whether you have things such as minions or not. You could have them with bounded accuracy too, if you wanted. The point is about the uselessness of the rapidly escalating numbers in 4e. You don't actually use them! When they would matter, the enemy stats are changed so that they don't! That is just silly. If you don't want the impact of the escalation, just don't have it. Whether you have easily killable mook versions of the monsters on top of that is another matter entirely.
Yes, you do, if you follow the guidelines for how encounters are supposed to be built in 4e.

Here, let me get the quotes for you:
When you’re building an adventure, try to vary the encounters you include, including combat and non-combat challenges, easy and difficult encounters, a variety of settings and monsters, and situations that appeal to your players’ different personalities and motivations. This variation creates an exciting rhythm. Adventures that lack this sort of variety can become a tiresome grind.
(excised portion regarding encounter complexity, e.g. terrain, plot importance, conflicting interests, etc.)
If every encounter gives the players a perfectly balanced challenge, the game can get stale. Once in a while, characters need an encounter that doesn’t significantly tax their resources, or an encounter that makes them seriously scared for their characters’ survival—or even makes them flee.
The majority of the encounters in an adventure should be moderate difficulty—challenging but not overwhelming, falling right about the party’s level or one higher. Monsters in a standard encounter might range from three levels below the characters to about four levels above them. These encounters should make up the bulk of your adventure.
Easy encounters are two to three levels below the party, and might include monsters as many as four levels lower than the party. These encounters let the characters feel powerful. If you build an encounter using monsters that were a serious threat to the characters six or seven levels ago, you’ll remind them of how much they’ve grown in power and capabilities since the last time they fought those monsters. You might include an easy encounter about once per character level—don’t overdo it.
Hard encounters are two to three levels above the party, and can include monsters that are five to seven levels above the characters. These encounters really test the characters’ resources, and might force them to take an extended rest at the end. They also bring a greater feeling of accomplishment, though, so make sure to include about one such encounter per character level. However, be careful of using high-level soldiers and brutes in these encounters. Soldier monsters get really hard to hit when they’re five levels above the party, and brutes can do too much damage at that level.
Monsters that are more than eight levels higher than the characters can pretty easily kill a character, and in a group they have a chance of taking out the whole party. Use such overpowering encounters with great care. Players should enter the encounter with a clear sense of the danger they’re facing, and have at least one good option for escaping with their lives, whether that’s headlong flight or clever negotiation.
It explicitly says you should use some encounters that are really easy, and some that are pretty hard; it even pointedly does not say you shouldn't use very high-level monsters, only that doing so should be done with great care and, generally, preparation for giving the players some other kind of solution besides fight-to-the-death (such as fleeing or parley.)

They do, in fact, want the impact of the escalation. They want there to be SOME fights which really are a cakewalk. And some fights which are brutally hard. While also having actual, observable progression, where what you could barely fend off X levels ago is chump change to you now. That's even called out as a technique to use to give players context and meaning for their advancement over time, to genuinely use the exact same monster from "six levels ago"

Your thesis is simply wrong. Not only does the text explicitly say you shouldn't have every fight be in lockstep with the PCs, it says that fights should vary a fair amount over time (even though the central tendency should still be somewhere in the "medium" to "hard" range). Further, it gives useful advice for how to do things well outside those ranges, beyond simply saying that you should use minions or the like. Though it does also recommend their use (in a different section, not the one quoted here), alongside various other approaches and tools to help create more interesting, varied, memorable combats and challenges.

And this does, in fact, relate back to "bounded accuracy." Because a system designed the way 5e is struggles to do what I've just described. A CR 4 monster is something that should be a nasty threat for a group of level 2 characters (over the line for Deadly, though not at a hypothetical "Deadly+" difficulty). But a full eighteen levels later, despite the rules claiming that 10 CR 4 monsters would be an "easy" fight (a bit below Medium, actually), that fight would be way, way more deadly.

There just isn't enough room in 5e to allow that real, obvious, unequivocal feeling of growth. 4e offers not one but (at least) two distinct ways to achieve that feeling: just straight-up using a monster 5-8 levels below the party, or turning it into a Minion. Both clearly demonstrate that the players have massively outclassed such monsters. There is no such thing as "massively outclassed such monsters" for anything past CR 2-3 in 5e. Only a massive damage spell--e.g. dropping literally a 9th level spell like meteor swarm--has such potential, and even then it isn't guaranteed because that's assuming every target fails its save, and at least a handful should pass.
 

Edgar Ironpelt

Adventurer
I have found that, much like the oh-so-wonderful days of 3e, 5e DMs usually call for rolls for everything under the sun. Sometimes even in places where the text explicitly says you shouldn't. E.g. the return of "you must roll stealth every single time you take an action" expectations. Coupled with the "unless it's so easy a fever-addled five-year-old could do it while sleeping, yes, I'm going to have you roll" pattern and the "if I'm having you roll, it's going to be DC 15 or higher because DC 15 is defined to be a medium check" pattern, this makes the swinginess anywhere from "incredibly irritating" to "deeply unpleasant."

"Let it ride," only asking for rolls where failure is both interesting and not unlikely, and using any DC below 10 (and that only if the DM is feeling generous) is, in my experience, quite rare. Much to my consternation. (Thankfully, I seem to currently be in a game that actually does do at least the first two things, which is a refreshing change of pace.)
I'm a GM who is willing to call for rolls for everything under the sun, and I'm also willing to extend DCs down to 5, 0, or even negative values. I want mechanics that give guidance to questions like "Is this task really easy enough that a fever-addled five-year-old could do it while sleeping?" It's why I whole-heartedly embraced 'Take 10' when I first saw it in 3e, and why I hate auto-fail and fumble rules with a bitter blazing burning passion.

And I love it when difficulties and modifiers create a smooth progression between "only the most expert of experts can be expected to succeed" and "even that feverish five-year-old can be expected to succeed." By contrast, when narrative or fluff text says "This character is an Expert with this skill, and should automatically succeed with this routine (at least for him) task," but the crunchy mechanics say "This character's stats give him an annoying large chance of failing if the task is gamed out," then one of the two is lying to me - and I do not like that.
 

EzekielRaiden

Follower of the Way
Because you don't need the escalating numbers to produce that experience! They do not contribute to that.
Yes, you do.

Because otherwise you get the "this level 0 commoner has a 25% chance to deceive the Prince of Lies himself, while the Prince of Lies himself has a 25% chance to fail to deceive this level 0 commoner." A problem, I will note, that was actually raised and discussed during the D&D Next playtest, more than a decade ago. And, of course, WotC did nothing whatsoever about it.

The scaling, contra to what so many claim, actually does serve a very important function. It is what makes demigods terrifying to random farmboys, and random farmboys irrelevant to demigods. The rules themselves instruct the 4e DM to make use of this, by including combats sometimes much further outside the "typical" range than even @pemerton described--perhaps as far as -8 on the low end and +8 on the high end, as I explicitly said earlier, which is anywhere between a quarter and half of all possible creatures.
 

Yes, you do.

Because otherwise you get the "this level 0 commoner has a 25% chance to deceive the Prince of Lies himself, while the Prince of Lies himself has a 25% chance to fail to deceive this level 0 commoner." A problem, I will note, that was actually raised and discussed during the D&D Next playtest, more than a decade ago. And, of course, WotC did nothing whatsoever about it.

The scaling, contra to what so many claim, actually does serve a very important function. It is what makes demigods terrifying to random farmboys, and random farmboys irrelevant to demigods. The rules themselves instruct the 4e DM to make use of this, by including combats sometimes much further outside the "typical" range than even @pemerton described--perhaps as far as -8 on the low end and +8 on the high end, as I explicitly said earlier, which is anywhere between a quarter and half of all possible creatures.

This would make sense if one actually used stats objectively, which 4e doesn't. So comparison of commoner vs demon prince is irrelevant, as that's not gonna happen nor is something the game cares about. NPC stats get changed to create desired narrative experience in relation to the PCs, so any comparison between NPCs is meaningless. And of course if we start to compare NPCs, 4e approach produces absurd results like commoners being able to one shot high level minions.
 

tetrasodium

Legend
Supporter
Epic
Because you don't need the escalating numbers to produce that experience! They do not contribute to that.
The thread is about bounded accuracy... so..... On top of the reason stated in 344,you very much do need them in order to provide room for players to gain cool new toys for their PCs through adventuring & for the GM to use those cool new toys for motivation. The importance and function of this dopamine loop was nicely described in the 2e dmg here. Now in 5e the "treadmill" PCs were on has been replaced with a stationary bike equipped with an electric motor that guarantees a full workout as long as the rider doesn't pass out from boredom.
 

Horwath

Legend
Precisely. Keeping mechanical representation rigidly fixed, and simply hoping that changes in player stats will by happenstance result in the monster transitioning from "terrifying threat" to "minor annoyance"....just doesn't really get the job done. It's not that there isn't a theoretical way for it to happen, I'm sure someone could come up with some (dare I say it) white room arguments to the contrary. But in practice, it's just...not how 5e ends up playing.

I suspect this is, at least in part, because HP has to bear effectively the whole burden of representing monster scaling, given AC is damn-near static once you get to ~20, but damage output scales in a pretty chunky, uneven fashion and much, much more slowly than HP. Creatures at the absolute bottom end of the progression (e.g. CR 1 or less) may be usable aplenty, but once you get to CR 4 or so, things plateau so much that it's just too much work to use lots of those monsters. Typical HP for a CR 4 creature is well over 100 (reportedly in the 116-130 range), and even a well-equipped high-level Fighter simply can't cleave through that in one attack.
There is ONE monster of CR4 with 120 HP, weretiger, but paired with horrible 12 AC, rest are from 97 HP to 36 HP, averages around 65 HP


Even if we assume a Fighter has a +3 weapon, +11 attack bonus, 4 attacks per attack, and hits with every one of those four attacks (which will be much higher than actual expected damage output, even accounting for crits), that's at best 8d6+32 = 60. It takes pulling out all the stops to potentially take out a CR4 creature in a single hit as a level 20 Fighter, which everyone always tells me is the king of damage and should be absolutely destroying everything. Heck, even throwing in GWM without any accuracy penalty (which would definitely be an exaggeration of damage), that's still only 100 total damage--a Fighter who lands every single GWM attack cannot consistently kill a CR 4 creature in one round.

So we're stuck having to work with pretty much CR1 and CR2 creatures exclusively if we want anything like "minions," and that roundly interferes with the "this enemy, who was once a deadly danger, is now barely chump change." Further, typical damage is in the 27-32 range, meaning ~29.5 damage on a hit. Even if we assume a low but moderate hit rate (typical attack bonus is +5, player AC is rarely higher than about 18 without a shield, so 40% is on the low end of reasonable), a swarm of, say, six of these creatures is going to get around 71 damage in on the first round, and probably another 30-40 on the second....when a 20th-level Fighter, we'll assume maxed Con, only has about 15+11x19 = 224 HP. So, despite this being only a quite "small" group of "weak" creatures, going up against one of the allegedly strongest characters around...you're still looking at taking a major beating (albeit possibly distributed across the party) and not actually cleaving through these things quickly at all.
so yeah, if you take tankiest CR4, you will need on average 5 attacks with +3 greatsword and GWM, but for most you will need 3 attacks.

and if you are facing several of them, it's time for action surge, 8 attacks with 15% crit chance and almost certain 1 killed per round will give you 9th attack as bonus action.
2nd round again action surge, that is 18 attacks in two rounds. That certainly looks impressive.
 
Last edited:

DEFCON 1

Legend
Supporter
I'll just throw in here that I find it funny that some people think it's some big issue when the ogres you find in an area have 88 hit points the first time you arrive, but then they drop down to 1 hit point (and become minions) when you come back... and yet don't seem at all concerned when the PCs had like 8-12 hit points when they first arrived in the area, but now have 50-- 75-- over 100 hit points as they've returned.

Seems to me that if we are willing to accept such a wide disparity in player character hit point changes over the intervening weeks/months/years before returning to the area without batting an eye... there's absolutely no reason why the monsters they face couldn't/wouldn't/shouldn't have the same thing.

But these discussions also are another indicator to me that I was not wrong in my thinking these past twenty years that combat in 3E worked well... the combat in 4E worked well... and the combat in 5E worked well. Because everyone is making good points at what the strengths of each individual game had, and why their respective games worked. Sure... some people prefer one of these games over the others... but they all work in their own ways to produce good gaming. That's what I always felt when I played 3E, 4E & 5E (and heck, AD&D and 2E too), and the fact that others are making similar cases for these respective games bears this out.
 

Edgar Ironpelt

Adventurer
Where do you think MMORPGs got the idea from? The old story of beginning adventurers being hired to clear out some rats from a basement is old. They even make fun of it in Baldur's Gate 3.
The concept is not unreasonable. The problem people (including me) see is when experienced adventurers then go on to hunt Rodents of Unusual Size, rather than trolls or winter wolves or displacer beasts - opponents that just have bigger numbers, rather than opponents that are dangerously different.
 

I'll just throw in here that I find it funny that some people think it's some big issue when the ogres you find in an area have 88 hit points the first time you arrive, but then they drop down to 1 hit point (and become minions) when you come back... and yet don't seem at all concerned when the PCs had like 8-12 hit points when they first arrived in the area, but now have 50-- 75-- over 100 hit points as they've returned.

Seems to me that if we are willing to accept such a wide disparity in player character hit point changes over the intervening weeks/months/years before returning to the area without batting an eye... there's absolutely no reason why the monsters they face couldn't/wouldn't/shouldn't have the same thing.
I don't think most people have an issue with there being orc runts that are weaker and orc war chiefs that are much more powerful. They probably don't even have a problem with over time the exact same orc runt becoming a war chief later. Here the difference in power is diegetic: the orc actually got more powerful in the fiction too.

It feels a bit different when the change in stats is not actually representing a change about the creature, rather than just a change in their narrative role. This is basically how minionisation used to work. Now neither of these is objectively wrong, but it is understandable if some people are fine with the former but not the latter.

But these discussions also are another indicator to me that I was not wrong in my thinking these past twenty years that combat in 3E worked well... the combat in 4E worked well... and the combat in 5E worked well. Because everyone is making good points at what the strengths of each individual game had, and why their respective games worked. Sure... some people prefer one of these games over the others... but they all work in their own ways to produce good gaming. That's what I always felt when I played 3E, 4E & 5E (and heck, AD&D and 2E too), and the fact that others are making similar cases for these respective games bears this out.
Yeah, that's quite fair.
 

Remove ads

Top