I’ve been thinking about WotC’s new “bounded accuracy” idea a lot lately. The long and short of it is that I don’t like it. On face value it solves a problem (scaling bonuses and DCs don’t mean anything) that we created ourselves when we stopped letting 5th Level Adventurers encounter a 10th Level Roper. We developed a fetish for ‘balanced encounters’ and, yes, when you scale monsters and obstacles to the party’s level, monsters and obstacles will scale to the party’s level. The answer is to stop scaling to the party’s level; then the whole thing goes away. Let the players experience things they can’t overcome, and then show them the same thing when they can overcome it and the sense that level advancement is pointless goes away. But it means showing players Really Hard encounters and Really Easy encounters all the time. It means setting DCs based on actual properties of the obstacle, not on how big the character’s bonus is (or should be).
Building appropriate DCs is actually pretty easy. Once you have the right notion of what the D&D system is supposed to model, you can get an objective sense of how hard things are. DC 20 isn’t “the DC that’s hard for 3rd Level Adventurers,” it’s “the difficulty of performing master-quality work.” And you can do this because you can break down what a character’s bonuses mean.
The catch is combat. At least, that’s the hook I’ve been stuck on since i started chewing on this issue. Deconstructing to-hit bonuses is still pretty straightforward. If you’re stronger you can swing your sword better, faster, more accurately, so Strength plays a factor. There’s a practical limit to strength (there’s an old Roles Rules and Rolls post that equates STR scores with “strength of n men”), and it’s based off a measurable quality of a creature. There’s also equipment to consider (since masterwork or Magic weapons can help score a telling blow), and lastly there’s training — which is represented as Base Attack Bonus and goes up based on level and class. if you have a complaint about the rate that BAB increases that might be a valid argument to make, but the system models several (fairly distinct, I think) tiers of adventuring, and there’s a hard limit on BAB within a tier (the best you can do is be a Fighting Man and get +Level).
The crux of the problem seems to be Armor Class, whichgenerally follows this model of ability+training+gear, except that there are a few extra things to consider and (here’s the problem) some of these factors seem to be abused in a lot of cases.
There are a bunch of factors that contribute to Armor Class — it starts with a base of 10 and then you add things for Dexterity modifier, armor or shield (gear) used, the size of your target (smaller targets are harder to hit), Dodge, Deflection, Luck, and Natural Armor. There doesn’t seem to be a good gauge for how these components should be scored, though. There’s a rough measure for Dexterity (compare it to human norms) and Armor and Shield (compare it to existing equipment; or just use existing equipment), and Luck is possible the most straightforward (each +1 is a 5% better chance then it should be otherwise). But there’s no notion of any kind of measure or scale for Natural Armor, and this is the one that seems to be abused the most.
Take for example the Purple Worm from the Pahfinder SRD. This is a CR12 creature, meant to challenge characters of, say, 8th to 12th level. It has a Dexd mod of -2 and a Size mod of -4, so we’re at a base AC of 4. If we add the equivalent of Plate Mail (+9) and a Tower Shield (+4) to the worm we get an AC17, which isn’t too bad — an untrained, untalented assailant only has a 20% chance of doing damage to the creature, which kind of makes sense for a huge, clumsy worm. But it’s expected to go up against characters with a +8 or +10 Base Attack Bonus, and likely have better gear and abilities, so we quickly ramp up to a 75% or better chance to score a blow. so what solution to the designers choose? They give the creature a +22 Natural Armor Bonus, close to double the effectiveness of full plate and a tower shield. Now the AC is 26 and the worm is a threat to it’s target audience. This happens again and again with large, clumsy, high-level monsters where the inflated Natural Armor (+25, +28, +30) “makes up” for the penalties of size. And I’d be OK with that, if there were some reasonable measure to hold creatures to when making the determination. A better solution, I think, is to add health instead. I’m not sure what the right formula is, but I think “add 5% health for each point of AC you remove” is probably a good place to start.
(It’s worth noting that this isn’t really a problem until higher level creatures, where they have to contend with size-based penalties and high expected BABs from the PCs.)
As a final note, something that’s occasionally brought up is that bounded accuracy means that (1) low level characters can be effective against high level monsters, even if the monster’s HP prohibits a ‘fair fight’, and (2) low level monsters can continue to pose a real threat to high level characters and never need to be ‘retired.’
My fist response is that this makes no sense right on the face of it based on the scope of adventures that the D&D system can account for. We’re dealing with a system that can handle work-a-day muck farmers (1st Level Commoner) all the way up to persons that rival the gods in power (20th Level Cleric), and everything in between. Something that threatens a farmer (say, a pack of goblins) isn’t going to be a blip on the radar for a god. And likewise, no farmer is going to have a chance of success if they engage a creature that was designed to kill gods. When bounded accuracy promises to let farmers challenge gods what it’s really promising to do is shrink the range of adventure that the system can model. It is reducing the scope and ability of the system, and I think that’s a negative thing. if you don’t like high level play, don’t play high level adventures. I’m personally a huge fan of E6 and generally get bored above Level 5, but that doesn’t mean reducing the system to only be capable of addressing gritty fantasy adventures is a good thing.
Secondly, the system already allows this. Because of the rules of Natural 20 (auto-hit) and Natural 1 (auto-miss), there is an upper and lower bound on the chance to deal damage. A muck farmer already has a 5% chance to hurt the Tarrasque (and a 1-in-400 chance to crit, depending on how you do threat rolls), and despite being made to kill gods the Tarrasque can still miss hitting the muck farmer (and will miss 5% of the time). Does this mean that an army of farmers can turn away the Tarrasque? It would have to be a huge army, but yes, that’s a possibility and has been since at least 2000 (and I’d guess even before then). It’s not likely to happen, but I don’t think it should be likely that an army of goblins can take out mythic heroes either, whether we’re talking about 3.5 or 5E.