This is a critique of the 5e designers' goals with Bounded Accuracy and their execution of it in 5e. It's not a criticism of anyone's gaming preferences or of people who enjoy 5e. If anything, players and DMs would be better served if the designers were clearer and more honest about their goals with D&D and provide gamers the tools they need to accomplish them, whatever the goals are. ADDITIONS/ERRATA: -One more effect of Bounded Accuracy I realized after making this. (And hell the video was already too damn long): ... The decision to lessen the importance of level and rely more on "abilities, not numbers" is one reason why it's hard to balance encounters in 5e. Making level matter less for monster statistics means that what's more important are (1) the number of creatures (e.g., action economy, the reason why you must add a multiplier when there are multiple creatures when calculating an encounter) and (2) hard-to-quantify abilities (e.g., a shadow's Strength Drain ability). Some may prefer this style of monster design and don't care much for ease of balancing encounters, but this IS one effect of BA that many would find undesirable. This is not INHERENT to BA: PF2e's encounter balancing still works with its Proficiency Without Level variant. But 5e's designers, when going with BA, decided to approximate difficulty with abilities instead of numbers (with limited success), and also prompted them to "let go" of concerning themselves with balance and leave balance up to Dungeon Masters, as I say in the final part of the video. -Looks like no one caught the joke at 0:49 yet. Bonus points if you do! (Requires knowledge from past videos) -28:16 I should've mentioned that the scaling of monster save DCs is too aggressive in the monster chart. It assumes that EVERYONE increases their bonus with their ASI, but they can only do that with one of their 6 saving throws! -30:15 The Banished fighter still returns if the party disrupts the caster's concentration. Still suxx tho (and the Fighter loses their agency) -Regarding "slogs," in fairness to 5e, I should have mentioned that 5e has a tool of foregoing rolling a d20 and simply estimating how many times a large number of creatures will hit a single foe ("Handling Mobs," DMG 250). It's a rough tool (if they need to roll a 5 to hit it says EVERY creature hits, instead of 80% of them) and it has the downside of eliminating randomness entirely for to-hit chances, however. -43:10 High-level fighters can only do 9 attacks once per short rest; however, I would still argue that 5 attacks per round shows the same problem of being a "slog" with little excitement. -49:25 the "master thief" of at least 11th level would have Reliable Talent so would not have a chance of failing. However, this highlights another issue: only Rogues of Level 11+ can do this. One goal of the D&D Skills system starting with 3e was to provide abilities that ANY class could be good at. Also, in 5e only the Rogue can reliably do Athletics checks, which seems strange and unintended... -@scottdouglass2 mentions that 5e's approach requires controlling DAMAGE scaling as you level, but 5e doesn't do this: "A fighter scales with level by getting more attacks, but a barbarian only gets 2 attacks, 3 with a bonus action. You can also scale with better magic items, but the core system doesn't address the topic well. Compare that to 13th age where your damage is baseline a number of dice equal to your level." -I didn't mention it in this video because it was already too long, but I think PF2e having more degrees of proficiency (Untrained - Trained - Expert - Master - legendary) instead of an on/off binary makes it easier for the GM to say "you the Wizard auto-succeed because you're proficient enough" and "you the Barbarian can't possibly succeed because, while you picked up a few books (you're Trained) you're not proficient enough (you're not a Master)." By RAW the rules say you can require a minimum proficiency level to even attempt something. Which the DM can do in 5e, but there's less rules support for making such calls. -I've seen a couple of comments saying "Well if the DM does X then..." Not everything should be on the DM. Why not put some responsibility on the designers?

This video does a great job breaking down one of my biggest problems with 5e. I tell people “bonuses are too small, except when they aren’t” but it doesn’t usually make sense with a lot of rambling haha

That's why I liked the "take ten" rule from 3rd edition (I think). When you're good enough at something, you don't have much chance to fail at it unless you're really in a crunch to do it, like when you're in a big hurry or improvising tools.

I remember reading an article a while ago about why campaigns in D&D fizzle out after the 12th level. I hadn't played a campaign that went beyond that at that point in time, so I didn't have any personal experience to inform my opinion, and there didn't seem to be any real consensus in the comments. Watching this video, though, really opened my eyes. I'm currently playing a campaign where we've just barely reached level 12, and so we're getting into the territory of monsters with +13 to their hit dice. In a recent battle against a dragon, the dragon could only miss our monk character if it rolled a 1 - which wasn't the best feeling for the monk! My paladin got plate mail at level 4, so with her shield, she's had a 20 AC, and as the monsters have been hitting more often, I've been increasingly wondering when we're going to find some better armor... Finding out that's not how the game works - that I've already got the best I should expect - is *really* disappointing! The idea that my character was better at avoiding enemy attacks when she was level four than she'll be at level 15 is a huge demotivator. It makes the rarity of campaigns that go beyond level 12 make a lot more sense.

it is extremely weird to hear that the game with extradimensional travel and wish also wants to be a game where kicking down a door remains a hard task, even for a guy who"s slain a dragon lmao

I'd be really curious what the results would be like if we surveyed players with questions like How likely is it that... a sneaky thief should be able to sneak by an unskilled guard? or a level 5 fighter win a fight against two level 1 fighters. I'm curious how people feel the probabilities should work instead of how they do.

The bit about the door is telling. They cite providing a moderate challenge for high-level characters as the reason to have a door be solid adamantine warded with arcane sigils. This misses the point entirely! The reason that the door is so strong is that it explains why _low-level_ characters haven't opened it. If busting the door down only requires a DC 17 check, then we're presented with a plot hole: Why hasn't a first-level character busted down this door already and absconded with the treasure? Verisimilitude failure right in the article that introduced bounded accuracy as a boon to verisimilitude.

"After a short period of GMing…" or they could print a table for me to read. 5e really has some strange stuff.

Bounded accuracy also just buffs casters. That level 20 "master" locksmith can't get that hard mundane lock? Well this level 3 wizard got you.

I think power creep is probably an important thing worth mentioning. I would imagine if you were running 5e with just the core rules and core rulebooks--no Tasha's, Xanathar's, no variant rules like multiclassing or feats, etc.--I imagine that the whole thing probably holds together a lot better. But as the newer books added more variants and abilities and powerful subclasses to the game without changing the underlying math or assumptions, it gets harder and harder to keep the system from collapsing. I also think that the fact that D&D seems to work best at levels 3-8 is probably tied to this... the system isn't actually robust enough to cover the entire level range, and both players and DMs kind of intuitively feel there's a sweet spot in this range where the balance mostly works, and outside this range the system is much more unstable.

Hot take: it's a good thing to trivialize the d20 roll. It was one of the strong points of 4e. At some point, your bonus got so high, that the d20 roll got less and less important. Which is a good thing, as your character got more and more competent. You don't want your demigod to be constantly at the whims of RNGsus.

I think the Numenera/Cypher System Rules implementation of Bounded Accuracy is a great example of achieving all of D&D 5e's goals -- less magic items, but offering a more clear sense of progression as you level up with their Effort and Edge mechanics.

it seems a hilarious amount of 5e's issues are solved by doing what 4e did

This explains so many problems I have with my current 5e PC. I'm playing a Artificer/Rogue multiclass. I quickly learned that this is a terrible build, but I loved the character, and stuck with him through the rough combats, and inability to contribute to a lot of rolls. Eventually, I shifted from a poorly-optimized dual-dagger wielder, to a switch-hitting ranged AC tank thanks to the Armorer subclass of Artificer. It's cool that I can do that, but I'm still miles behind our Wizard, Bard, and Barbarian who are all monsters in combat. A huge part of this problem is because our DM throws a lot of monsters at us that force saving throws rather than targeting AC. Few of these were Dex or Int saves, which were my proficient saves from my first level in Artificer. Instead they mainly required Con, Wis, and Cha saves. My Cha has been 16 since the start of the campaign, since the character is a pretty good diplomat, but because I am not proficient in Cha saves, the save has _never_ grown past +3. Making matters worse, my Wis and Con scores are 11 and 10 respectively. Oh dear. The Barbarian hasn't hardly cared about these saving throws because she has resistance to basically everything, and the Bard and Wizard are either not in the way because they don't want to be in melee, or are proficient in the saves. My character's martial incompetence was even crap at low levels because a core Rogue ability I chose was rendered irrelevant by the party. Both casters continually selected save-or-suck spells, which target enemy DCs, and the Barbarian uses Reckless Attack to gain advantage on everything, which negates the Master of Tactics feature from, the Mastermind Rogue subclass, which lets me help allies on attacks from 30 feet away. That doesn't work on spells that force saving throws, and it advantage doesn't stack with itself, so the Barbarian doesn't need it. Now, our DM is incredible. They are a master storyteller, and have even worked at an indie TRPG dev studio. They get the math of the system. This isn't the DM's fault, these are core weaknesses baked into 5e. All of this crystalized for me in a "greatest hits" fight our DM set up for us. A crazy Fey Lord created illusory versions of our past set-piece fights to keep us distracted while he did nefarious things. Our DM expressly told us that the fights were _exactly the same_ as what we encountered at lower levels. While we trivialized the replay of our very first major combat encounter of the campaign (from when we were level 3), our next big set-piece fight (from level 5) was incredibly challenging for us (now at level 12)! We actually did _worse_ in the illusory version of that fight, despite having two extra DMPCs with us the second time around! This is, frankly, ridiculous. D&D is a heroic fantasy system where your characters are supposed to grow dramatically in their combat prowess. At level 12 we should not have been greatly challenged by a major boss fight from level 5. This isn't a system like GURPS which is built in a way where starting threats can remain incredibly threatening for an entire campaign. This is D&D, and even my poorly built Artificer/Rogue should be able to look at a squad of enemy soldiers, at level 12, and not think to himself it might be time to run.

What I really never understand is this strange yearning for the goblins you fight at level one to be a challenge the entire game. Like, even in Monster of the Week shows, the mooks our heroes fight are eventually more of a distraction to the main show of the villian. I would personally be annoyed if I kept fighting the exact same enemies throughout a campaign. I'd rather be fighting clearly improved goblins, and the ones we were fighting flee at the sight of the party.

Something I've learned about the math of Pf2e, which you somewhat addressed, is that level added to progression and non-stacking bonuses means that "on level" challenges basically cancel out those increasing bonuses, meaning success or failure on those "on level" challenges primarily comes down to the actions you take in the encounter, whereas the significantly higher or lower level encounters become commensurately impossible or trivial, providing (IMO) a great balance of palpable power growth AND action/die rolls mattering. TL/DR: Every +1 matters means that "on level" encounters rely on tactics and decisions, but adding level to proficiency means that low or high level enemies FEEL like they are lower or higher level than the players. On the topic of being "rules light", it *could* be considered "rules light" in the sense that it has fewer codified rules than games like Pf2e and 3.5, but it substitutes rules for rulings in a variety of situations. The thing is, "rulings" are simply "rules" that players didn't know until they try something - so as you play and need rulings to supplement the lack of rules, you then create those rules, which then become subjective to the DM and tabke rather than objective to the game itself. So, to be glib, its "rules light" until you play the game.

I really felt this when getting into BG3: bounded accuracy in 5e made the dice more important than my skill/ability than the bonus at virtually every level.

I love the idea of certain things remaining difficult like the door. For example, I don't imagine that Gandalf could kick down an iron door, despite being a powerful wizard. But I would also be very disappointed to watch Gimli or Aragorn, two seasoned warriors, having to spend 10 minutes kicking at a door before they could get through. So I get what they were going for. But by giving the dice roll so much weight compared to your stats/skills, you've added an absurd amount of RNG into the game. It kills immersion and story-telling on a regular basis. I mean here I am giving a detailed description like, "My grizzled fighter, his muscles like iron and his gaze like steel, approaches the small dilapidated shack. After weeks of tracking, questioning, bribing, and more than a few threats, he has finally tracked down his mark to this address. Unconcerned, he takes a long swig from his flask, sniffs, spits onto the dusty road, then raises a heavily booted foot to kick down the door- an action he has performed more times than he could remember over his many years bounty hunting... Unfortunately, he isn't able to kick in the door because of the 25% chance of failure, so after a couple more tries, he eventually ends up just breaking a window and half-crawling, half-falling inside so he can do the combat." Every party basically becomes the robbers in Home Alone, comedically incapable of performing routine tasks.

Wow. I had forgotten how 4e did the saving throw thing. My groups really didn't enjoy the "daily powers" for martial classes, but they really seem to have done EVERYTHING else really well. Edit: Also, parrying with a weapon or a sheild is a skill, it makes sense that more experienced fighters are harder to hit. Part of the bounded accuracy problem is the D20. 5% chance is not really "nearly impossible." 2d10 would have 5x the ganularity.

As an example of the failing of 5e's bounded accuracy: In a recent game, DM has the party(level 13) go through a stealth segment to avoid detection by a cult. The shadow monk casts Pass without trace on us, and so even my half-plate artificer gets over a 20 stealth despite rolling a 4(+7 Stealth from proficiency, +10 pwt) Then because more than half of us can fly now, and the ones who can't have dimension door, we teleported/flew into the final room to confront the boss directly.

What you say about D&D 5e being "rules light" near the end really strikes at the core of why I went back to 3.5e after playing only the starter box adventure in 5e. The game is heavily streamlined and cut down, and in doing this it dumps most of the decisions on the DM that were clearly laid out in older editions. It's so much harder to DM as a result, not even counting the horrible imbalance between player power and monster power.

I love the idea of bounded accuracy in 5e, but as demonstrated by a character I built to be a "dodge tank" for a campaign I was unhittable and therefore unkillable, but only for the first 4ish lvls until monster scaling caught up to my start point and due to bounded accuracy I wasnt able to keep getting harder to hit, so by ~lvl6 I was just getting knocked out in the first or second round of combat every encounter due to taking CON as a dump stat thinking that high AC an bounded accuracy would be sufficient to keep the character alive unless the enemy got lucky.

There's an interesting phenomenon in PF2 that doesn't occur in 5E as well that you didn't mention under slog. In 5E, those "bag of hitpoints" monsters lead to very long combats. Meanwhile, most PF2 encounters only last about three rounds even when a big bad in involved unless the party is just absolutely not equipped correctly for the monsters involved. Occasionally combat will go longer, but it's not frequent and every round past the third gets more and more unlikely.

One issue I have with D20 systems that can impact this that I don't believe you mentioned is the dice itself. Players expect that in general you'll get a middling roll but the extremes are just as likely as the averages. Those extreme rolls can make play more exciting for some but it does also contribute to the issue of low level characters having a decent chance to succeed and high level characters having a decent chance to fail which is a problem that kept coming up in the video. In quite a few other systems I've seen the use of two dice instead of one for rolls (with the result being the sum of the two rolls) which naturally results in more middling rolls and less extremes. In playing those systems I found that this helps a lot to mitigate that issue of low level characters achieving what they shouldn't too much and high level players failing what should be an easier task too often.

The more this video went on the less this seemed like an issue with "bounded accuracy" itself and more a bunch of smaller issues surrounding it such as monster design, saving throw design, bonuses. All which have been implemented better in other systems.

I honestly really like the idea of bounded accuracy. 5e however just failed at implementing it.

At the end of the day, my biggest problem with this system is the assumption that more HP is the best way to make higher level fights interesting. It isn't. It just makes them take longer; sometimes, a lot, lot longer.

I am that player who played every edition from B/X up to 4E and got pushed away during the 4E days by the game just, not feeling like an RPG anymore. 5E promised a return to simplicity and a return to traditional games modes. It could not possibly have delivered more poorly. It doesn't even get an honorable mention on my "Top 5 favorite editions of D&D" list. I've been playing a PF2E campaign for the last few months and it, despite also being complex, delivers so much better. My character concept was kludgish as can be and it *feels* complete at level 2 after getting one archetype feat. 5E can't do that, not even with a pretty straightforward character concept. The way forward, for me, is not through WotC.

51:12 alternative title 'adventurers not needed, the townsfolk can kill the dragon themselves.'

Excellent video. Magic items do indeed shift the balance a lot once you start distributing them. One last thing (maybe a future video topic?) : some spells and abilities are overly powerful at low levels in D&D.

This explains a lot of the complaints I've had about the system for years. It also explains why multiclassing is so strong in the system. That's usually the only way to get proficiencies you didn't already have. Including saves, and the ability to acquire expertise that some classes just simply don't ever get access to.

First, I think bounded accuracy was an admirable goal. 3.5's modifiers completely outstripped the range of a D20 die roll, making the roll irrelevant. It also encouraged ridiculous specialization from PCs. For example, I ran a 3.5 campaign through to 20th level and the druid hyper focused on Spot and Listen so that, with magic, he had around a +65 to +70 to those skills. Add that to speadsheets needed to track bonuses and the ability of PCs to extend durations enough to cast buffs, sleep, regain spell slots, and cast more buffs before combat even started and we ended up with players with 50-70 spell effects on them at one time. Now, throw enemies with Greater Dispel Magic at will and you had the means of 2-3 hour combat rounds. Not fun! Also, I had PCs with such high ACs that any creature that had a chance to hit them auto-hit any other PCs who did not max out AC. 5e fixed a lot of this. Unfortunately, it could not get around the fact that massive hit points are not fun. I get why they changed from AC as the prime defense determinant to hit points as the prime defense determinant. It's more fun for a player, especially one who waits 30 minutes for his or her turn, to hit something and at least do a bit of damage as opposed to swinging and missing. And the idea raised by The Rules Lawyer that this increase in hp at the expense of AC wrecks the narrative idea of swinging and missing or parrying fails to recognize that hit points are an abstraction. A 20th level fighter with 150 hit points is not getting chunks of flesh ripped out of him every time he takes 20 points of damage. That PC is literally parrying or dodging successfully when he takes 20 damage. But more hp also turns every single combat into a slog. They also rip the stakes and urgency out of every combat as well as a great many other situations. They wreck verisimilitude (oh...you just fell down a 100-foot pit onto the hard stone floor and took 35 damage and have 115 hit points left...how do I explain that narratively?). The answer is to move to a system where hit points are static and reasonable, like BRP/Runequest. In that system, you can make yourself harder to hit. You can wear armor to reduce damage if you are hit. But, because you have so few hit points, every swing matters. Every swing can bring you down. Combat, therefore, must be well considered, engaged tactically, sometimes run away from, and is often short and quick. In BRP/RQ, a goblin still matters to a "high level" PC because that goblin coming at you with a 2-foot-long sharp blade might (not likely, but might) get past your parry, find a seam in your armor, and pierce your jugular vein. That's the proper way to keep low-powered monsters relevant at all play levels and keep every combat important and exciting. That's the best way to impose bounded accuracy, by emphasizing the accuracy part. The hit point phenomenon also has implications outside of combat, as it affects roleplaying. Is the local farmer protective of his daughter's virtue? In D&D, a 10th level fighter doesn't give a crap. If that farmer, angry at the PC, comes up behind the PC in town and swings a two-handed axe at the fighter, there are no real consequences. Even if the fighter is not wearing any armor. Even if the farmer has complete surprise. Even if the farmer gets advantage. Even if the DM house rules that the farmer automatically criticals without having to roll to hit and does maximum damage....it still doesn't matter. Even if the farmer is beefy as hell. A 20 STR farmer who max criticals with a battle axe does 21 damage. The 10th level fighter with a 14 CON (which is substandard) will have 58.5 hit points remaining on average. He will turn around and waste the farmer. Because of this, the PC has no need to fear the farmer's threats about respecting his daughter's virtue. Actions have few immediately grave consequences in D&D. Finally, as an aside, the advantage/disadvantage system is broken in 5e. It works fine at the middle of the dice roll bell curve but gets destroyed at the ends. A person who needs to roll an 11 on the D20 to hit someone has a 1 in 2 chance to hit. With disadvantage, he or she has a 1 in 4% chance to hit. That means disadvantage cuts the chance of success in half. That feels right. But now assume the person needs to roll a 20 and has a 1 in 20 chance to hit. With disadvantage, that drops to a 1 in 400 chance to hit. That's ridiculous. And if they have the dreaded Luck Feat, that drops to 1 in 8,000 chance to hit. I DM'd for a Bladesinger whose entire build was made to max AC and impose disadvantage. Yeah, I could get him in other ways, like forcing a CON or CHA saving throw on him, but there was basically no chance for him to be hit (he had the Shield spell as well, of course). Bounded accuracy is supposed to stop BS like this, but given the many ways to break AC limits by PCs and the weird math surrounding disadvantage, even the auto-hit on a 20 rule became meaningless for him.

PF2e with proficiency without level and automatic bonus progression kind of accomplishes a lot of what 5e set out to do but much better.

Funny how all of these problems (and more) were pointed out by the mathematically inclined fans of 3.5.

I’ve felt for so long that DnD (above maybe 5th or 6th level) felt way too much like everything was a damage sponge. It always broke my immersion for the “evasive rogue” to never seem to get better at avoiding attacks, only better at having a big hp number. I never considered that bounded accuracy could be to blame!

42:06 5e actually ahs rules on handling fights with lots of enemies. Simplifying it a lot consists on asuming that the average number of enemies that will hit the PC hit.so for a gorup of 8 enemies, you may determine that each round the character will be hit 3 times according to his AC and the attack bonus of the creatures.roll that damage and keep going rather than rolling 8 attack rolls. 49:45 I think it's ok that the circus performer, can still slip and make a mistake. I see that as a feature, not a problem.

Still blows my mind that wotc not only abandoned game balance as a core design goal but that they admitted it openly.

I see what you did there at 0:49 @TheRulesLawyerRPG. A tip of my glasses to your resilience and this community.

I remember running 5e and trying to figure out what was going on with the monsters in the manual. It didn't make sense, there seemed to be no system behind it. Then I realized that you were supposed to choose a CR for the monster, and derive everything from that, rather than make the monster and derive a CR from its stats. I gave up, made monsters the 3e way, TPK'd parties, and haven't looked back. I gave 5 a fair shake, but B/X is the way god intended us to play D&D.

Say what you might about the level scaling issues in D&D 3, 3.5, and Pathfinder 1 (and there were many), the Take 10 / Take 20 rules helped avoid a lot of the skill challenge problems you outlined in the latter part of this video.

Instant like. I’ve been complaining (in quiet) about 5E’s attempt at bounded accuracy for ages, and I’m happy to see others spotlighting its issues.

Absolutely fantastic video. I just wanted to add that the unreasonable potential of level 1 players is also why many DMs essentially run the skill check system as being entirely limited to tier 1 feats of skill. A level 20 barbarian can't be allowed to rip the gates off a castle or pull an entire pirate ship ashore because that allows a level 1 barbarian with the right team composition to consistently create an effect equivalent to a 5th+ level spell (fabricate, in the first case). It's puts DMs in the uncomfortable position of effectively banning an entire portion of the character sheet after a certain point because the range from 1 to 20 allows anyone to potentially do anything at any point. Because of this I know many DMs that fervently believe simply making up DCs as you go, essentially scaling the "feat of skill" separately from the DC (so a DC 15 check at level 1 is an iron banded door and at level 20 it's that adamantine door with rune wards). When I told them that the game seems to indicate that DCs are intended to remain mostly stationary, I got several reactions as though I were a crazy person and even one who told me not to spread around ideas that were so clearly against the intended functionality of the system (humorous, as they clearly understood the system well enough to know what I was suggesting destroyed skill expert characters and consider it painfully obvious to the point they assumed the system compensated for that with the power of DM fiat). Edit: To add one last point, a regular large village or small town, one with enough people to have a small box of gold, can easily kill an Adult Red Dragon with bows. In their example of rallying the villagers, the heroes aren't needed. That town should have at least 40 combatants between guard and hunters who are proficient with bows. I mathed it out once and the dragon dies before its breath attack comes off cooldown. Even hit and run doesn't help it with the range of longbows.

For me the best part of DMing pathfinder has been knowing that level is absurdly good at telling how difficult a fight will be. My 3rd level party just faced a 6th-level ghost, and even alone it was a really tough boss encounter. The new GM Core even has a few combos that make encounter building take seconds. "If you want an easy framework for building an encounter, you can use one of the following basic structures and slot in monsters and NPCs. • Boss and Lackeys (120 XP): One creature of party level + 2, four creatures of party level - 4 • Boss and Lieutenant (120 XP): One creature of party level + 2, one creature of party level • Elite Enemies (120 XP): Three creatures of party level • Lieutenant and Lackeys (80 XP): One creature of party level, four creatures of party level - 4 • Mated Pair (80 XP): Two creatures of party level • Troop (80 XP): One creature of party level, two creatures of party level - 2 • Mook Squad (60 XP): Six creatures of party level - 4" Because of Unbounded Accuracy, all of this works suprisingly well. Enemies who used to be tough are now just standard creatures, and will become lackeys over time. Just like how Dark Souls bosses eventually become standard enemies.

On troops - Ironsworn, while being a solo-first design and based on PbtA (so, focusing on horizontal rather than vertical growth) strongly encourages running... I think it's 5, might be 4... of the same enemy in an encounter as a group, clustering them together and upping the difficulty rating of them by one. And then again when you get five of clusters. (I think it has appropriate names for the clusters to norse fantasy). As a side note, I'm not sure 'bounded accuracy' and 'zero to hero' power progression are really compatible design goals. At least not without going into the realms of hp bloat comparable to a Final Fantasy game. There's nothing wrong with the idea that a monster that's a threat to a party in session 1 should still be a threat to a party in session 30 - but the times I've seen it work tend to be games that focus on horizontal growth, so you gain more options but those options aren't intended to be more powerful than your existing options, and your stat growth - if present at all - is going to be in the order of a couple of +1s while keeping all the stats within a tight -1 to +3, range over the entire campaign to distribute among your five stats. A different framework for character progression across a campaign (And one I'd say I prefer except I'm currently prepping to run Animon Story which is a game that takes inspiration from Digimon. And as such... Yeah it has a very vertical form of character growth...)

As much as I'm enjoying this video in general, I have to just say I love the fact that there's a Pathfinder 2e monster that can _honest-to-God Naruto ninja-log you._

I've thrown CR based encounters out of the window after party reached level 7, it became inconsistent and monsters were usually too weak offensively and defensively. If I used that monster stat by cr table, monsters had too much hp and combat was a slog. I now use "Heroic Homebrew Monster Maker" by AHero and finally combats have proper difficulty and length.

So deeeeep. I love it! Thanks for putting words to what I never could have and helping me understand the art and science of game design at a whole other level.

Actually in 5e you can end a TPK encounter at level 5. I put 2 potentially deadly or TPK level encounters infront of my party last week. The Bard used hypnotic pattern and not a single NPC made the save. The only NPC still standing was the one with legendary resistance and she was dealt with easily despite my plan for her to leave when the party was mostly dead. So yeah. It wasn't a great session on my end, but the players had a lot of fun. Eventually it will probably get boring though.

So I do think that 5e’s more “Bounded Accuracy” is, as was said, easier for DMs/GMs to get started with. And for a brand new DM, anything to make things easier for them, I approve of. But I do think that both Players and GMs will, over time, want an experience similar to 2e’s system unbounded. Also, I do think 6 different saving throws is ridiculous, 3 is much more reasonable

One of the things that bugged me the most about 5E is the level 1 characters being able to do "nearly impossible" things. The math basically allowed a village of non-leveled NPCs take down a ancient dragon with archery. Yeah there will be a lot of losses, but even a level 0 character can get a nat 20, and enough of those will whittle the dragon down. PF2 has no such problem. A level 1 character rolling a nat 20 against an ancient dragon is still likely to be missing the dragon.

Great video. I think there's a mix of problems here. Some are stylistic to how bounded accuracy works and some are the implementation of bounded accuracy in D&D. I'll describe two of the latter: 1) Times the designers forgot they were doing bounded accuracy. These are usually (but not always) from spells that are too powerful. I think two of the clearest examples are guidance (which has no business being a cantrip in a game with bounded accuracy as a goal) and pass without trace (which makes everyone in the party a high level sneaky character). You have numerous more examples, but it really creates a problem when you are deviating from your baseline immediately. It also makes features that tweak bounded accuracy very powerful (e.g., paladin aura). 2) They forgot to include the class features (other than spells) that are supposed to advance your character. I don't think a level 20 wizard who's never been interested in physical fitness should be able to jump a 15 ft gap, which they can in all numbers go up systems. However, they have multiple class features they could have opted into to deal with this problem in the form of spells (jump, misty step, fly, etc.). A fighter as you note does not automatically succeed this jump. This to me is not a problem with bounded accuracy, but with the design of the fighter. In general the non-spell casting classes do not have sufficient class features that let them achieve goals outside of bounded accuracy and thus achieve the desired fantasy.

Bro's singlehandedly lawyering his way into Pathfinder 2e sucess and D&D 5e's demise. I really like both games, but I must stop and applaud you for the way you studied carefully each of your arguments to point exactly weak spots that would shine what you had to say. Bro, you're a genius.

I've long clung to bounded accuracy as a primary redeeming quality of 5e, and the source of my hesitancy to embrace PF2. Listening to you break it down... well, I'm convinced beyond a reasonable doubt.

It's worth noting, I believe the 5E DMG does have simplified rules for running groups of monsters, but it's a somewhat obscure variant rule. Zee bashew has a video on it, 5e mob combat I think

5e's approach to items goes against almost every lesson I have learned from both BG3 and the DND movie. Outside of roleplay, the most enticing aspect of BG3 was hunting down rare and powerful items to complete my character builds, and felt like it had far more depth and flexibility than class customization. Meanwhile, the D&D movie basically managed to make all of the MCs' problem solving far more interesting by just adding a magic item that was, honestly, above the party's level, but gave them a unique and powerful way to solve their problems. Making players deal with weak gear doesn't make the small bonuses more important, it just makes the boring-as-hell class system even more dominant, and robs beginner DM's of the opportunity to give their characters interesting or exciting loot.

Bounded accuracy would be perfectly fine in 5e if it also included conditional auto successes and powerful features that more or less read "you can do X" without a check needed. It doesn't make sense for a god-slaying PC to get bogged down in knee deep mud so they gain a feature that simply states they're immune to it along with other similar effects. The monk's wall run is a good example of the kind of thematic feature every class should get in spades. Barbarians should eventually be able to lift and throw tremendous weight, jump around like the hulk, be immune to non-magical damage, and basically become Hercules by the time they're level 16. Being able to use a giant stone pillar as an improvised weapon and leaping up to smack a dragon out of the sky is the kind of thing needed to really sell the idea of being powerful. And in the case for attacking and saves there should be cut off points based on the HD of creatures. What I mean is as a PC levels up their attacks and spells automatically succeed against creatures of a certain HD or lower with max damage. Should a level 10 wizard really fail to cast hold person on a 1 HD goblin? No, they're competent enough at this point that it's trivial.

* if you want a very flat game, then why do a class and level system?There are lots of games that are built around flat systems where advancement is not so general, like Savage Worlds. One of the points I like about class and level systems is that you can go from a guy on the street to somebody shifting around the planes and dealing with cosmic threats. * It is so fun in PF2 to fight a monster that used to be a boss monster, but is now just one of the new boss's minions! * the idea of just increasing the hit points is practically saying "We want higher level play to be a slog where you just keep doing the same thing for longer." * While 5e is built on being stingy with magic items and not having a "magic item shop", one of their game worlds is Eberron, which specifically has a dragonmarked house that makes and sells magic items.

Part of the problem with 5e's "bounded accuracy", from what I can see, is having items and abilities that just stack numerical bonuses onto those rolls. Everyone's got a 20 in their ability score and a +4 proficiency bonus? Great, so everyone's got a +9 to hit... oh, except the fighter with the archery style and a +2 longbow, he's got +13 to hit... The bounded accuracy was... unbounded from the start...

I appreciate this entire post and the nod with the "big damn heroes!" reference. Very nice!

as far as your point at 41:19. You're absolutely right. However intellect devourers can be added to that small list as well.

Bounded systems: One set of numbers goes _vroom!,_ turning your characters into nearly unstoppable demigods of mayhem. Unbounded systems: Another set of numbers goes _vroom!,_ turning your characters into nearly unstoppable demigods of mayhem. It's really just a question of where on your character sheet you want the frustratingly-large numbers to be, and which of the many math calculations you will need a calculator for.

30:20 Good thing fighters have Indomitable to cover this gap somewhat, although it would be nice if it was basically legendary resistance. 43:10 Fighters do not get to do up to 9 attacks "per round". Maybe on one turn once per rest. Realistically it's 4-5 attacks. 49:00 Again, ignoring a very pertinent class feature: Reliable Talent. Your 10th level circus peformer turned master thief isn't failing that DC 15 Acrobatics check.

Great analysis. Boy I wish PF2e remaster had adopted the 4e "Highest of 2" for saves. The tyranny of Con, Wis, Dex triumvirate must be stopped!

I feel like I've actually listened to a professional lecture on game design, well done

Excellent job on packing this into a 1-hour video. I'm surprised it's so short, but you covered everything important. I feel like I've discussed parts of this so many times but never encompassing all salient points, and it's good to now have a single video that we can point to in the future where it's explained in detail. Cheers to that. I also work in game design, so analyses of games to this depth of detail is very helpful in helping people make better games.

I like the idea off keeping stuff small so that numbers don't balloon out of control** ... but Bounded Accuracy had to compensate by ballooning HP. **: Eventually, everything in 3.5 had "Natural Armor" to keep ACs competitive.

Most of the issues you point out with bounded accuracy come from the core mechanic of D&D: the use of the D20. A D20 has a flat probability curve: all numbers are equally likely. If you use 3D6 instead, you will be rolling numbers in the 8 to 13 range 80% of the time. A +11 with 3D6 means you will fail a DC 15 check only if you roll a 3 (about 0.5% chance), making a specialist actually good at stuff (for comparison, DC 15 would be 9.25% chance with +0).

The DM is supposed to ask for the rolls, and can choose not to. DMs don't use the "take ten" rule nearly enough. It's meant to get around the tactic of "I try this over and over for an hour", but it can and should be used to handwave trivial rolls. If the stakes are high and it's interesting, dramatic, or funny, make people roll even if it's small. But ultimately you never _have_ to make a player roll.

I'm proficiency without level kind of guy so I actually like bounded accuracy idea. However that's my personal taste, and I can see why other people have other preferences. It's much like taste in food, some like ti sweet, some like it salty and some like it spacey.

My impression is that the designers were not really trying to solve the problems they thought they were trying to solve. The real issue is a much older schism, between High Fantasy and Low Fantasy. D&D had its roots much more firmly in the High Fantasy genre, which is why the rules can seem so outlandish at times. In practical terms, you're taking characters who are well trained in their given fields starting out, becoming heroes of the caliber of The Fellowship of LoTR in the mid levels, and in the high levels they're super heroes, demi-gods, and the like, who would be the subject of epic poems in ancient times. This completely flies in the face of Low Fantasy where you start out as an incompetent pretender, and might be able to fake it until you make it at the mid levels. If you survive to the high levels then you're John Wick or Liam Neeson in Taken, which is to say you're a total bad-ass with a scary reputation but death is always just one mistake away. The problems arise when these two genres start to conflict with each-other, because the genres are mutually exclusive. It sets up conflicting expectations between all of the players at the table, and even within each and every player's understanding of the game. As I have said before, if you try to play Aragorn (a high fantasy character) in Game of Thrones (a low fantasy setting), you'll end up with Ned Stark. (a man who met a very undesirable end because he didn't really understand the rules of the game he was playing) I suspect that the bounded accuracy was meant to drag D&D 5e toward the Low Fantasy end of the spectrum, because that is much more reflective of modern culture and storytelling in the mainstream now. However, D&D still has it's roots mired in its high fantasy origins, and the mechanics still reflect that origin. The designers failed to grasp what they were trying to accomplish so they ended up with a patchwork mess of rules. They were trying to fit a square peg in a round hole. If I'm being honest, the schism between the two genres is such that they would be better served coming up with an entirely different set of rules for low fantasy games, say a handbook for it that revises all of the classes and gives means of making monsters and challenges more grounded toward that style of play. Just my opinion there, though.

I firmly believe that dice rolls should always matter. The whole point of having dice at all is so that there can be genuine tension in the game. Otherwise it's just like playing pretend or reading a book. There may be stakes, but the outcomes are predetermined by either the story or the math. It sucks that sometimes you might fail at a task you should be really good at, but if there wasn't some chance of failure, the dice shouldn't have even been rolled to begin with.

Excellent video. I'm actually thinking of making a video on my house rules that attempt to 'unbound' 5e math. This is a great reference point.

Loved the video and thoroughly enjoy you content in general, even if I am not a particularly strong adherent of Pathfinder 2e. From a couple of comments I noticed you seem to agree (please, correct me if I am wrong) with the idea that some of the choices made by D&D 5e design (such as the bounded accuracy) feel dissonant because they do not fit with the traditional power fantasy that the system allegedly aims to represent. Were that the case, did you ever consider making videos examining other RPG systems that better fit other styles and genres of roleplay? I think a very interesting set of videos could emerge from analyzing how other systems accurately decided to represent a certain genre, or make a certain narrative statement, by virtue of well designed mechanics.

I quite like the AC 40 dragon it means "hard to damage" it is quite straightforward, directly threathening. Hitpoint based math is just wrong: 5000 hitpoins dragon is about cheesing it with an army of cheap peasants, you will need more than 12 goblins if AOE fireball is available, or you need "smarter" goblins, to challenge abilities of lvl5 party... and smarter goblins can just be different creature than the stupid lvl1 ones.... The bigger numbers exist there to simplify the calculation. AC is just amalgamation of chance to evade, chance to block chance to resist the damage. But its "universality" breaks back to variants what players want to do anyway and what makes sense: Flat footed AC, spell AC (touch attack AC), damage reduction... It is balancing between too much to calculate for the attack and between the boringness of always doing the exactly same chance roll.

I love this so much!! I'm a long time TTRPG player, but I don't have the head for maths, so a lot of the issues I have with 5E's proficiency progression is hard to put into words.

Hi! I'm a game designer. I wrote the Rags to Riches RPG system, and have acted as a consultant for a couple of indie companies (though usually on the side of historical content rather than quality control). Watching this video has spelled out some of my own specific issues with D&D's numbers in way that I hadn't quite put into words before. It's insightful and I agree with it completely. It's also given me inspiration to get back to work on my own d20 system, which is something I put aside many years ago due to my own disillusionment with the RPG industry. I don't know if you'd be interested in partnering up at some point in future, I feel like you'd make a wonderful QA guy even if you have limited experience with other systems. But I definitely wanted to say because of you it's given new life to an old project which I'm quite happy with, so thank you.

The idea that the same monsters can present a consistent threat to the party just sounds so bizarre. Imagine this: You're John Fighterman, and you recently left you hometown with your grandpa's longsword to join up with a group of adventurers. You're level 1, and currently getting rolled by goblins in a cave. Now, fast forward to the end of your adventure. The evil dragon lays slain at John's feet, its caustic blood coating John's sword. John Fighterman has become a legend, a warrior without peer, the very peak of what mortal kind are capable. His songs are sung far and wide, his deeds echoing into the future, inspiring a new generation of brave heroes. John decides to visit his old hometown, adorned in all the splendors of Faerun, intent to return his grandpa's longsword. You're level 20, and currently getting rolled by goblins on a roadside ambush. Sure, you have so many hitpoints you can't really be killed by them, but you're also supposed to be the best in the world. Despite being level 20, the goblins can still block your blade, the blade that felled a dragon. Despite being level 20, the goblins can still strike through your armor, the armor that blocked dragon claws. The threats your supposed to face are battles for the fate of the world, foes that set the gods atremble, and yet you can still be hurt, and eventually killed if you're really unlucky, by goblins.

I think you have overstated the issues here, probably due to liking Pathfinder’s style. Dnd does have some balance issues - especially around monsters. But for the most part, the system does work as you are stating it should. Characters at level 20 are exponentially stronger than at 1, not marginally. While your AC only goes up 2 or 3 and your “to hit” by 6, other things go up. Your HP will be many times higher, and you gain extra attacks as well as damaging effects for those attacks. And spell power rises at a similar rate. Good video though. 👍

This video helped me to put words to some frustrations that I've felt about D&D 5e, as a DM, since my party reached level 8-10. I've had some difficulty challenging my party in combat without creating fights that are only slogs, while at the same time certain skill tasks feel really "swingy" - things that shouldn't be hard for an expert sometimes fail inexplicably for that character, and sometimes things that should be almost impossible are overcome by a lucky die roll. To a certain degree, that can be fun, but it's happened enough that it starts to stretch verisimilitude. I know there are some things I can do to help mitigate that (allowing experts to 'take 10' automatically for easier skill checks, borrowing homebrew/3rd party monsters such as those from MCDM's book), but until now I'd thought that maybe the issue lied more with me as a relatively inexperienced DM. It was helpful to understand that a lot of the problems actually come from the math of the system itself, which explains why things that help fix it are typically homebrew rules and third party content.

Static DC's create benchmarks for players to overcome and get better. 3.x did this to great effect. The numbers eventually got so high that you couldn't fail at a certain task anymore. However, there is a narrative advantage to this that some people overlook. For example: At level 1 in 3.x a rogue cannot consistently balance on a surface that is less than 2 inches wide (DC 20) and is only comfortable making the attempt out of combat when he isn't distracted and may use a balancing pole to gain a circumstance bonus. At level 5, that same rogue can now confidently balance on that same surface and can do so in combat with a reasonably small amount of risk. In addition, he is no longer considered flat-footed while balancing on said surface (5 ranks in balance negates the penalty). At level 10, that same rogue can balance on that same surface even when it's slippery or foggy out. In 3.x you can take a -5 penalties to balance to move at full speed instead of at half-speed. At this level, he can succeed at that in normal or mildly hazardous conditions regularly. At level 15, that same rogue cannot fail to balance on that same surface at all in any normal or mildly hazardous condition. This is the point where people often say, "Why have numbers so large if you can't fail?", but imagine he's taken some DEX damage, and he's sickened. So he has two stacking penalties. It's during a blizzard, and he can barely see (severely obstructed +5 to DC) and the surface is made of ice (extremely slippery +5 to DC). The story being told here is that he's got to find shelter, and he needs to balance on the narrow icy surface over a cliff in order to do so. He's dizzy from his cold, but he is so good at balancing normally, that even under extreme conditions, there is still a chance he can pull it off. Having breakpoints where you no longer have to roll is only a problem if the game doesn't give you tools to modify the challenge. This admittedly comes with increased complexity and a steeper learning curve.

2:52 A lawyer being thorough, I’m Jack's complete lack of surprise. Of course you’d be thorough. If is kind of helpful to those of us that are not so inclined. Thank you!

It's funny how I disagree on a fundamental level with literally every single argument in that article

Very thourough, I like it. A thing I've been doing in my campaign is let the party have higher ACs and as they hit Tier 3 of play I realised that the monsters "appropriate" for their level couldn't hit them most of the time and that the monsters that could via abilities targeting saving throws would be able to take out half their hit points in one go. Once this happened for the first time one of the players got freaked out thinking I was trying to kill their character outright. I wasn't (I had calculated it so that even full damage would leave them with a sliver), but it taught me that I needed to readjust by homebrewing monsters and enemies. And thanks to this video it's clear to me how: Allowing the beasties to have better To Hit bonuses and ACs while tuning down the damage by about 1-3 dice depending on the ability... And coming up with a way to improve the party's saving throws. I will now commence the deep dive into magic items that improve saves, which will of course also help the baddies in the fight against the party.

Savage Worlds doesn't call what it does "bounded accuracy" but it does a passable imitation of what D&D 5 sets out to do, IMO. As you Advance, you're likely to pretty quickly reach the upper levels of "raw numbers" in your core stuff. (Most Traits max out at d12, though you can get higher with certain Advancement options.) And there's a chance for any character (or enemy) to get a stupidly high result on a Trait test or damage roll because (almost) all dice explode. You can absolutely reach a high rank, and wade through masses of goblins who would have massacred you and your friends 10 or 20 sessions ago. BUT those goblins, too, might roll specially high on their damage. It isn't likely, but it CAN happen. The part of the system where characters (PC and otherwise) are distinguished, is in Edges: this is how one character is distinguished from another with a similar function. The term I usually hear is "lateral" advancement (vs. D&D's "vertical" advancement).

Half the complaints here are "I don't like that style," and "I think most groups agree with me." Not to suggest that 5e is fine as is. It has problems. Vomiting every possible interpretation of what is wrong with the system clouds the issues because many of the different complaints are about the same underlying problem. I understand that this is fundamentally a PF > D&D video. My complaint is that's all it ends up being. And it doesn't really do a good job being that. Had it spent more time focused on contrasting how PF2e handles what 5e does poorly, rather than reiterating by calling out different forms of the same root problem, it would have been a better video. If you want to promote a "better" system you have to do more than just trash talk. Make a case for what's better. (I also understand that the real money is in stirring up as much conflict as possible in the comments, but an actual contrast between the systems would have the same effect on the partisans and result in similar levels of engagement.)

4e was alot better and more fun system. More choices and more abilities other than swing your sword 2 times. Never played a campaign where at high lvl you didn't have +3 weapons and armor. Most fighter/paladins had high 20's ac. Mainly paladins cause they outshine fighters easy.

The only issue I had with this video was using an AD&D module and the treasure it gave to reference the "low vs high" amount of magic items in a game. In AD&D the exp system was very different, where only a fraction of your exp came from killing monsters. Most of your exp came from the treasure you found (1/2 the value if you kept it and the value you get if you sold it was converted 1:1) while also requiring you to pay a fee (and time) to level up based on the level you are advancing to. For that reason, a lot of adventure paths (and DMs in general) will actually give out a fair amount of treasure and simple magic items knowing that a fair bit of it will be sold off for exp or used to pay for leveling, two things that 5e doesn't have in its system. Great video and good comprehensive numbers, though! I love how 'simple' 5e was made, its a great starting point for new players. Just wish the RUNNING of the game was as simple and accessible, haha.

Excellent video. IME it's almost impossible to explain to a player who wasn't heavy into the math of 3.x/pf -and- sees the trainwreck in 5e's bounded accuracy setup. It's great having a single source video that just goes down the list kicking over failure after failure before a player can say "but but bounded accuracy was made to do $NextThing not that"

A lot of the points you make are incredibly weak, pretending certain things are problematic when they are really not at all. Chuckling through most of your sentences doesn't really help you come off as objective either. To give some examples: o) The "Creating Quick Monster Stats" chart you showed is just that - it's meant to help the DM create monster stats quickly. It's not meant to be a clearcut structure that all creatures in the game are built upon like you pretend it to be. An umber hulk and a troll are both CR5. Would the game be more fun if the stats of both matched that chart exactly? Or would it be more fun if the desingers took the liberty to tune the troll to feel as much like a troll as possible - and do the same for the umber hulk? Imho, it's very obviously the latter. At the end of the day, CR is just meant to help DMs to know *roughly* how hard an encounter is going to be. There's not need for it to stick to a restrictive math formula - and the game is better for it. o) Building your case of "you can have crazy bonuses at level 1!" on outlier subclasses (like the peace cleric) and exploits (like help-action spamming with familiars) is not a great plan. The DM can simply play without well known things like this, that can negatively impact the game. Don't like how twilight cleric, silvery barbs, etc. mess with your game? Simply don't include them. Don't like exploits like help-action spamming familiars or specific overpowered multiclass combinations? Just play without them. Not only are those issues *very* easy to avoid, but invariably you'll have balance-outliers in pathfinder and older editions as well. Despite its flaws, 5e is arguably doing very well on that front in comparison. o) Your example of the level 3 fighter getting +20 to the attack roll backfires in several ways: 1 - You downplay the fact that this does indeed consume several party resources and requires multiple players to pull off. 2 - You pretend like reaching a high attack bonus is a problem in the first place. That level 3 fighter doesn't need a +20 to hit the troll they'll be up against. In fact their excessively high attack bonus will only marginally affect their damage output. o) The level 1 rogue breaking the difficult lock is another case where you pretend like the designers fucked up, when really everything is working as intended. So the rogue tries to pick a difficult lock - they use the skills they specialized in; the bard inspires them; the cleric grants them guidance - and they still end up with a roll that's tense and might easily fail? Sounds like the players are working as a team to achieve a common goal - and the roll of the die is still meaningfully affects the outcome. How exactly is this failing game design? (Also, getting advantage on picking a lock isn't nearly as easy as you insinuate either - probably requires yet another PC to help out with a "cat's grace" spell.) It's great that you love pathfinder, and I'm sure you'll enjoy it far more than 5e, but none of the things you mentioned make 5e any less *fun*. The whole point of it is too keep things simple and moving quickly. If you are looking for a slower, more mathy experience, then it's just not the right game for you. EDIT: The part where you pretend like a low level barbarian can outperform a high level wizard in an arcana knowledge check is also complete nonsense. Why would the DM ask the barbarian to roll a check for something they can't possibly succeed at? It's just another example for a situation where 5e promotes the DM to decide how to resolve a situation, rather than trying to create more and more rules to cover every possible circumstance.

When I ran 5e, I really struggled with balancing encounters. It seemed more an art than something that I could calculated, and ended up relying on buffing or nerfing enemies on the fly depending on how the combat happened. Since we switched to Pathfinder 2e its been much easier to actually math encounter balance.

You fully explained what I always felt! Thanks

about the grimoire of arcane and amulet of the devout requiring a level dip... 1- for spell caster a level dip can hurt a lot. a wizard 10/cleric1, sure he grab the armor proficiency from the cleric and have access to healing spell that wizard cannot learn and still has 6th level spell slot. But here what everyone forget: he is now a level 11 character. a level 11 wizard would be able to cast level 6 spell (which are more powerful then upcast level 5 spell) but now he only have access to 5th level wizard spell at best. 2- Then also remember you are limited to 3 attuned item, so if you have a grimoire of arcane and an amulet of the devout, you only have 1 attunement slot remaining: what will it be? a staff? a wand? a defensive ring? you can only have 1. 3- The DM decide what item are given to the party. The DM control which item can be found. If the DM give a amulet of the devout AND an grimoire, it's the DM fault (or the cleric of the party gave his item to the wizard, so yeah you got a super powerful wizard but a weaker cleric, if your party want to play League of Legend and feed all the power to a single character, as the DM you can easily show them the flaw of that tactic by sending a monster who resist the wizard (imagine a Rakshaza who is immune to spell lower then 6th level, against our wizard10/cleric1 who cannot learn 6th level spell and the party gave him all the magic item)

Rok the level 36 (minus one experience point) monk born without arms, in a setting where metals are extremely rare, with weapons nor armor by build and even ethics, who has NEVER SEEN AN ANVIL BEFORE, tried making a horseshoe for a quest, it was suitable for greater DIVINE enchentment and arcane enchantment! He became a GOD OF SMITHING. Then he tripped over a branch.

Bounded Accuracy makes Dragon's hide not so tough, Will-o-Wisps not so fast, complex locks less confusing and tricky, and massive leaps shorter. By making the numbers stay the same, you make the impossible feats less impossible, and thereby making the moment when the heroes become strong enough to do the impossible, significantly less impactful.

When you mentioned troops at 42:12 , I immediately thought of minion monsters from 4e. Not only do they have only a single hit-point, but the damage they deal is precalculated and static, so they are easy to run in large numbers. If anything, the main complaint people had with 4e was that, unlike other versions, there was strict mathematical progression and bounds, and it made it feel game-y for some.

As a 5e dm, i noticed this issue right away, and i reworked my game math to include the magic item bonuses as part of the expectation of what monsters should have to face reasonably head to head. And this actually gave me the space to do some much more interesting stuff. As an example, i let every player figure out their own custom bonus that makes them heros. Some forged a weapon as they adventured, making it more and more powerful as they leveled (and it eventually led to me revamping magic items too into stackable tiers of abilities, really went well.) While others wanted to hone some special technique. This allowed me to have a barbarian with this epic great axe that could even cut through magic defenses, a roguelock that had a custom pact, and a monk that literally had an ability to pump his ki points into his base stats for short term boosts. And with the revamp of my basic expectations of what my dm side numbers should be, it didn't break the game, it made it more fun. If you want to know my math, i noted the spell level increase points and divided those by 3. At each of these levels, you should expect the whole party to have gained +1 offensive and +1 defensive factors due to equipment. Some will gain that bonus sooner than others. This means i expect my high end pc to hit to reach +14, and pc high end defenses to reach ac comparable to that. And how i let specialists do what they wanted to specialize in, was to Not factor in their class and feat related bonuses into the math. While this may sound dumb, its actually playing to the wants of the player's when making their character progression choices, and when something beats a specialist's numbers, its meant to show this is important or there is some other factor to the action. That it's a puzzle, not an obstacle. If they figure out a way to beat that puzzle without using the puzzle parts? Great. Let them. Its what heros do. On the dm side of the math, one very easy way to balance things the stuff already in the book is to take those hit dice, and just max the number once you hit about level 9. 10d6 is 60 hp. Yeah sure you have more hit points, but that increase in the fighter's to hit with magical bonuses just going to make it a for sure 3 hit combo every turn. And this lets the monster take that hit more than once, but not more two or three times. Players ive had enjoy hitting every attack against basic enemies, and racking up sizable damage in a turn, and watching that monster staggering but still getting back up for a little more, rush in and slam them back as best they can in turn. Bosses should be harder to hit. Its actually an easy way to show. "Oh this guys a bad ass i cant just wail on them.". my game went from 1 to 20. This showed me when those pitfalls in the design were an issue, and this was my solution.

In defense of the level 5 power spike, Fireball is an outlier. In all respects it's statted like a 5th level spell while you get it as 3rd level. Compare Shatter at 2nd level. Fireball deals twice the damage at almost triple the range with quadruple the AoE.

As a pretty big fan of 5e, you are right: monster scaling in the monster manual basically doesn't work right. I don't think this is the fault of bounded accuracy necessarily, I just think the existing monster manual sucks (at the task of of properly assigning CR and scaling enemies appropriately for levels). You probably could make a monster manual that scaled properly, but the officially printed one does not.

The various team members who invested in specific abilities to increase accuracy or checks of players is exactly what they mean when they say simply leveling doesnt solve the problem, but rather the choices the charachters made. They expect that you cant open the level 30 lock unless you have a rogue, or bard or a bunch of other features working together. They also didnt make a perfect system because 5e also wants to introduce things that surpass the normal. most enemies might be a level 17 AC guy with 150 hp, but they wanted there to sometimes be a level 17 enemy with 200 hp in a bespoke fashion. This makes certain monsters more interesting because they are different. Is the bonded accuracy of 5e flawless? not at all, but its a valid design choice versus the other options, and definitely mitigates some of the issues they were trying to solve.

The 5e DM’s guide does specify that the table in the book and the information they include in the book is not exhaustive of everything they use for making statements blocks. The table is specifically called a table for quick monster stats. Additionally, a lot of monsters have resistances or immunities to some damage which explains some lower hit points in the monster manual. There are definitely a lot of problems with 5e monster stats but it would be wrong to draw conclusions purely from the table.