A weird thing about monsters

[Witty Username]

So, I have no work in the morning an have mild insomnia, let's throw something out.

I was watching a video, which I am not going to link because it is more than an hour and a number of you people don't do that stuff, it was mostly about how bounded accuracy doesn't work as advertised and the flaws in its implementation in 5e,
In Short,
-Saves are busted because DCs are expected to increase over time but saves are not, removing one of the goals of increased player agency without being inhibited by needing to significantly plan out a character
- The math of the game is fundamentally off because attack bonuses and damage values of monsters increase by CR, but not PC AC values, leading to a feeling of characters getting weaker as they level rather than stronger
- the lack of need for mechanical bonuses from other sources leading to looser game design with multiple areas that are easy to exploit


But the thing I am more interested in, Is this nugget. Its a blog from 2018 (so definitely not new information), but the idea it posits is interesting. It is essentially a list of observations comparing the DMG monster guidelines and Monster Manuel statblocks and noting some weird stuff, in short:
- Monsters tend to have lower HP and expected damage across the board than the DMG would have us believe
- Monsters tend to have higher attack values then what the DMG would have us believe
- Monster HP, AC, Damage and Attack bonuses have mild correlation. The DMG would have us believe high attack bonuses would be balanced by lower damage and vice versa. but in practice there is little to suggest that in the MM. High AC monsters have comparable HP to low AC monsters and High Attack bonus monsters have comparable damage to Low Attack Bonus monsters.
- Of particular personal validation, Resistances and Immunity have no correlation to a monsters HP to a given CR, which makes sense as vulnerability does not according to the DMG, despite being the same thing from a game perspective presented differently

Does this match up with what people have observed using these systems? What are your opinions on any of this?

[Unoriginal]

So, I have no work in the morning an have mild insomnia, let's throw something out.

I was watching a video, which I am not going to link because it is more than an hour and a number of you people don't do that stuff, it was mostly about how bounded accuracy doesn't work as advertised and the flaws in its implementation in 5e,
In Short,
-Saves are busted because DCs are expected to increase over time but saves are not, removing one of the goals of increased player agency without being inhibited by needing to significantly plan out a character
- The math of the game is fundamentally off because attack bonuses and damage values of monsters increase by CR, but not PC AC values, leading to a feeling of characters getting weaker as they level rather than stronger
- the lack of need for mechanical bonuses from other sources leading to looser game design with multiple areas that are easy to exploit


But the thing I am more interested in, Is this nugget. Its a blog from 2018 (so definitely not new information), but the idea it posits is interesting. It is essentially a list of observations comparing the DMG monster guidelines and Monster Manuel statblocks and noting some weird stuff, in short:
- Monsters tend to have lower HP and expected damage across the board than the DMG would have us believe
- Monsters tend to have higher attack values then what the DMG would have us believe
- Monster HP, AC, Damage and Attack bonuses have mild correlation. The DMG would have us believe high attack bonuses would be balanced by lower damage and vice versa. but in practice there is little to suggest that in the MM. High AC monsters have comparable HP to low AC monsters and High Attack bonus monsters have comparable damage to Low Attack Bonus monsters.
- Of particular personal validation, Resistances and Immunity have no correlation to a monsters HP to a given CR, which makes sense as vulnerability does not according to the DMG, despite being the same thing from a game perspective presented differently

Does this match up with what people have observed using these systems? What are your opinions on any of this?

PCs don't feel like they become weaker over time, because their HPs keep getting higher even if the AC does not.

The devs decided during the early days that getting hit by a big attack and surviving when the same big attack would have killed you five levels ago was more gratifying than not getting hit by a opponent when the same opponent would have hit you five levels ago.

Personally, I agree, because opponents who can't hit just get phased away.

I don't understand what you mean by "saves are busted". Saves do increase, if you invest into them. That's what agency is.

[Theodoxus]

I don't understand what you mean by "saves are busted". Saves do increase, if you invest into them. That's what agency is.

Outside of Monk, you'll only ever get 3 good saves, a max of 2 in the 'big 3'. So, not sure how much agency a specific character has in that regard. You're always going to be 'bad' in Con, Dex, or Wis. And some classes/builds don't have a need to boost Dex or Wis except to help cover being non-proficient in the save and that's a waste of build points.

[Unoriginal]

Outside of Monk, you'll only ever get 3 good saves, a max of 2 in the 'big 3'. So, not sure how much agency a specific character has in that regard. You're always going to be 'bad' in Con, Dex, or Wis.

That's what agency is.

Agency isn't "I'm good at everything", it's "my choices have a real, concrete impact on what is happening".

[stoutstien]

CR is like judging distance with your thumb. You can get really good at it with practice but it never going to very precise.

NPCs are designed to be relevant as long as possible so even if you need a lot more of them(both in density and frequency) you don't have the issue with them falling off.

This means when you look at any individual block they look really wonky but when you start putting them into encounters with other NPCs it's does work well enough for what it's designed for. It's horribly explained but what isn't.

[J-H]

I use the Blogofholding CR calculation method to check CRs. It's simple and generally accurate.

PC attack and damage also go up, so a monster that's hard to kill at level 5 will usually be just a speedbump at 11. DCs do go up, but how many monsters really force a lot of saving throws? It's only the caster-types (a minority), breath weapons, and then a few "boss" types like illithid and beholders, unless you do things not explicitly listed in the MM, like have high-strength brutes try to shove their targets prone.

My experience is that PCs saves do go up with levels. Warcaster, Resilient, Indomitable, Flash of Genius, Monk failed save reroll, Monk 14, Paladin aura, Samurai & Rogue extra save proficiencies.... most parties have 1-2 characters with good saves or that boost the saves of others, even leaving aside magic items. Despite what they may say officially, magic items are important in 5e, especially for characters that are not full casters.

[Beelzebub1111]

I watched the video you are talking about and generally agree with the premise. I do find that in practice the impossible tasks completed at low levels with no spent resources requires a lot of preplanning and coordination on the player's part, but I can see how tables with players who exploit that would find it to be a problem or annoying.

I also understand how the balancing system fails to take into account magic items in challenge ratings and adds another layer of difficulty to building challenging encounters first hand. I ran updated versions of 1e and 2e modules for 5e indended for player leveled 1-3 and there were a TON of magic items that threw off the balance of the game very early. Maybe a misconception of the history of how much the game expected you to have magic items in the early days (It's a lot, maybe they didn't start with shops but you got a LOT of them)

The whole design goal of "Monsters never phase out of difficulty you can still present a challenge of fighting more of them" absolutely failed. It just made fights a boring slog. Monster design also feels kind of lackluster as a result as bigger threats seem to just be more HP, More Damage. Anyone whose fought both frost giants and fire giants in 5e will tell you that it gets to be a pretty boring slog either way. To think that at higher levels that means you can just fight more of them at once while still being entertained means that you haven't tried.

[J-H]

It's not about the monsters...it's about the encounter design.

1: Fight 4 Frost Giants
2: Fight 4 Frost Giants; two are acolytes and can cast Bless or Bane. Add buffs/debuffs and a need to try to break enemy concentration;
3: Fight 4 Frost Giants alongside a cliff with an 80' drop. The Frost Giants use Shove and Grapple and try to throw the party off the cliff. Add terrain and both sides trying to maneuver and add forced movement.
4. Fight 4 Frost Giants alongside a cliff as above, one of them is an acolyte, and one of them climbs atop a 20' rock nearby and then stands there throwing a returning warhammer at the back line of the party while buffing his attacks with low-level Ranger spells (hail of thorns, etc.), and the giants have a pack of allied Dire Wolves running around their feet trying to knock PCs down. Adds all of the above, plus a high-threat ranged enemy who's hard to reach, minions who matter, and suddenly nobody wants to just stand still and smash the attack button.

[Witty Username]

I don't understand what you mean by "saves are busted". Saves do increase, if you invest into them. That's what agency is.

Two of your saves increase, the other four do not. And in this case this is a thing discussed in the video, that one of the goals of bounded accuracy is to allow participation of non-specialists. Saves don't really work like this because DCs scale beyond what most characters can reasonably resist.

I think the quote was something like, 'success and failure should be determined by the player's cleverness, not what's on their character sheet'
So that is me using agency wrong, what was more meant is saves only care about the character build step, not actions during play.

[Keltest]

Two of your saves increase, the other four do not. And in this case this is a thing discussed in the video, that one of the goals of bounded accuracy is to allow participation of non-specialists. Saves don't really work like this because DCs scale beyond what most characters can reasonably resist.

Maybe at the extreme high end of play, but in my experience most monsters wont have a DC for anything exceeding around 15. Thats rough if youre a barbarian making a wis save against hold person, but its not insurmountable, and there are tools in the game to allow you to work around it, up to and including just getting an ASI or feat to improve the save, if its that important to you.

[Witty Username]

PCs don't feel like they become weaker over time, because their HPs keep getting higher even if the AC does not.

I have actually seen this one a little in play,
But AC is static, HP goes up means that because monsters gain both greater damage to keep up with the increasing HP and greater attack bonuses, combat trends more deadly as level and CR go up.

My sampling is low though, we like our magic items at our table, so our AC tends to be a little higher than system expectations. I think the info I have on this may be on one hand counting in terms of sessions.

[Beelzebub1111]

Maybe at the extreme high end of play, but in my experience most monsters wont have a DC for anything exceeding around 15. Thats rough if youre a barbarian making a wis save against hold person, but its not insurmountable, and there are tools in the game to allow you to work around it, up to and including just getting an ASI or feat to improve the save, if its that important to you.

If the game doesn't work on the high ends of play where the last 25-33% of your game should be taking place are a problem then it's still a problem. In your example a Lich's save target is DC20. that means for the barbarian in your example it is insurmountable.

[Keltest]

If the game doesn't work on the high ends of play where the last 25-33% of your game should be taking place are a problem then it's still a problem. In your example a Lich's save target is DC20. that means for the barbarian in your example it is insurmountable.

A Lich is a CR 21 creature. If a barbarian has made it all the way to that point without doing ANYTHING to invest in protection against wisdom saves, thats because he doesnt want to, not because he can't.

A barbarian gets 5 ASIs. Even if youre playing in a featless game, thats enough to bump up your wisdom from 10 to 20 if its that bad of a problem for you. With feats, the options become significantly stronger and cheaper. And thats just the options available to the barbarian by themselves. Throw in teamwork and the solutions are plentiful.

[Unoriginal]

If the game doesn't work on the high ends of play where the last 25-33% of your game should be taking place are a problem then it's still a problem. In your example a Lich's save target is DC20. that means for the barbarian in your example it is insurmountable.

DC 20 is only insurmountable for a PC with a -1 modifier.

Should it not matter when a player selects which stat has a -1 mod?

[tokek]

I have actually seen this one a little in play,
But AC is static, HP goes up means that because monsters gain both greater damage to keep up with the increasing HP and greater attack bonuses, combat trends more deadly as level and CR go up.

My sampling is low though, we like our magic items at our table, so our AC tends to be a little higher than system expectations. I think the info I have on this may be on one hand counting in terms of sessions.

I would tend to agree that other than real skew builds the higher levels of the game render AC less valuable over time. At level 17+ I don't find relatively normal amounts of AC (22-24) do an awful lot. But then you will have other abilities by then - my Rune Knight basically just uses other tricks to not take the damage and the AC only really reduces the number of times I need to do that to something manageable.

Combats do trend shorter in number of rounds but longer in elapsed playtime in my experience. Because the action economy gets wild - everyone has high impact BA and Reactions and martial characters can have a lot of attack on their action while casters can have complex spells to resolve on their action.

[Darth Credence]

DC 20 is only insurmountable for a PC with a -1 modifier.

Should it not matter when a player selects which stat has a -1 mod?

It's not even insurmountable then. Paladin in the party? Could be up to a +5 to the roll from the aura. Artificer? Well, if they haven't used all their flashes of genius, it could be another +5, or they could cast resistance on the poor -1 character beforehand to at least give them a 1d4. Got a cleric? Are they not blessing the people with bad saving throws, to get a bonus 1d4? Have a bard that can give you inspiration? A ring or cloak of protection? The right Ioun stone to boost wisdom?

Plenty of ways to get past the insurmountable saving throw. Difficult, sure, but that's what happens when you build a character that is weak to those saves.

[tokek]

It's not even insurmountable then. Paladin in the party? Could be up to a +5 to the roll from the aura. Artificer? Well, if they haven't used all their flashes of genius, it could be another +5, or they could cast resistance on the poor -1 character beforehand to at least give them a 1d4. Got a cleric? Are they not blessing the people with bad saving throws, to get a bonus 1d4? Have a bard that can give you inspiration? A ring or cloak of protection? The right Ioun stone to boost wisdom?

Plenty of ways to get past the insurmountable saving throw. Difficult, sure, but that's what happens when you build a character that is weak to those saves.

Also - its fun to have a weakness

It became a bit of a group joke that my ranger was terrified of illithids because he had a terrible Int save and basically every time he got downed and possibly killed it was illithids that did it. Having a bad Int save became part of the story.

But yes also all the things you say. Having any one of Bard/Paladin/Cleric/Artificer in the party tends to mitigate bad saves and there are magic items that help with it as well.

[Amnestic]

At level 17+ I don't find relatively normal amounts of AC (22-24) do an awful lot.

It absolutely depends on what you're fighting, and in what numbers.

A party of 4 17ths fighting four Clay Golems (CR9) is a Deadly encounter, by raw CR. They've got a +8 on attack rolls. 24 AC means they're hitting on a roll of 16+ only, so less than 25% of the time. Your wizard in a robe of the magi with AC18 is getting hit only half of the time still (well, probably less, because they've got enough slots to burn Shield constantly).

A hypothetical dungeon delve for your tier 4 characters should ideally have a healthy mix of the sort of solo CR23 Ancient Dragons with +17 attack rolls and the party of CR6 Vrocks with a +6. If for no other reason than it's pretty cool to to get to stomp down some 'scary' demons that aren't so scary anymore.

One of the encounters described in a later section of Chains of Asmodeus (where the party is 17th-20th level) is "a half-dozen cambions (CR5) aided by ten bearded devils (CR3)". Through sheer numbers this likewise comes out to Deadly, but they've got attack rolls of +7 and +5 respectively. 24AC effectively neuters the bearded devils almost entirely, and the cambions are still only hitting 1/4th of the time.

Another encounter in the same area is a maelephant nomad (CR14, +11 attack rolls), four bone devils (CR9, +9), and four barbed devils (CR5, +6). Again, likewise 'Deadly', but +6 to attack rolls on the barbed means they're rarely hitting your 24AC Rune Knight, but the nomad is going to hit with decent enough frequency (13+ roll required, so 1/3rd of the time, so maybe twice every three rounds due to its two attacks/turn).

But this is also the adventure where the party can and possibly will end up fighting CR25-30 archdevils with +17 attacks and DC24 spell saves and whatnot. So it's a mixed bag.

[tokek]

It absolutely depends on what you're fighting, and in what numbers.

<snip>

.

There is a different challenge with lots of weaker monsters - as a frontliner how to stop them hitting the squishy backline. It is a challenge but realistically not one where taking damage is the primary thing you worry about.

Its when the threat is to the frontline with the high CR monster that can hit just about anything that you then need other tricks to mitigate damage.

As for the danger level the DMG says

When making this calculation, don’t count any monsters whose challenge rating is significantly below the average challenge rating of the other monsters in the group unless you think the weak monsters significantly contribute to the difficulty of the encounter.

Which I tend to extend to anything significantly below the level of the party. On the whole monsters with CR less than half the party level probably don't add much danger to the encounter. There will be times they do, but that is something I will consider on a case by case basis. CR 7 monsters vs a level 17 party are likely to just get wiped without doing much other than making the wizard look awesome for a turn.

[Amnestic]

There is a different challenge with lots of weaker monsters - as a frontliner how to stop them hitting the squishy backline. It is a challenge but realistically not one where taking damage is the primary thing you worry about.

So either AC24 isn't enough, or it's too much? :P

Which I tend to extend to anything significantly below the level of the party. On the whole monsters with CR less than half the party level probably don't add much danger to the encounter. There will be times they do, but that is something I will consider on a case by case basis. CR 7 monsters vs a level 17 party are likely to just get wiped without doing much other than making the wizard look awesome for a turn.

Well a team of 6 Vrocks (CR6) is 'Hard', not even 'Deadly' so yeah I wouldn't expect it to be super taxing to take them down, but between stuns and spores and attacks and 100HP a piece, they're gonna stick around for at least a turn or two to get some hits in and cost the party some resources. If the Wizard wants to blow her 9th level slot on meteor swarming them to get her "AHAHAHAHAHA BURN!" moment then more power to her, I say! It means she won't have it for the boss at the end of the dungeon.

And likewise, if the party manages to get the Vrocks into a position where they can only hit the AC24 rune knight due to terrain, then he gets to feel pretty great as attack after attack bounces off his armour and deals nothing to him.

[tokek]

So either AC24 isn't enough, or it's too much? :P

AC24 largely lets you roll over easy monsters and easy encounters but it makes very little difference in seriously challenging encounters - which were the only ones where your build was really going to matter much anyway

Or at least that is my experience. At the highest levels any sort of frontliner needs more tricks than just a decently good AC to stand up the the pounding they will take. But then at that level you have had the chance to take several feats so that's fine.

[Keltest]

I think its fair to say that pushing for AC past about 20 starts getting significant diminishing returns as far as survivability goes. You can do it, and its not worthless, but you do start to hit the point where you need more than just AC to survive.

[tokek]

I think its fair to say that pushing for AC past about 20 starts getting significant diminishing returns as far as survivability goes. You can do it, and its not worthless, but you do start to hit the point where you need more than just AC to survive.

High AC can dominate low tier games and pushing it up can keep it relevant into the mid tiers. But my personal experience is that its less relevant at the very highest levels of play - or at least it is in encounters where you are actually in danger.

[Amnestic]

AC24 largely lets you roll over easy monsters and easy encounters but it makes very little difference in seriously challenging encounters - which were the only ones where your build was really going to matter much anyway

On the other hand is the build mattering "quietly" because it turned an encounter into an easy one? AC24 is "too much" but...I mean, that's kinda the point! You geared and built for a high AC, and your payoff is encounters that might have been "medium" or "hard" are now "easy" and tax far less resource.

Or at least that is my experience. At the highest levels any sort of frontliner needs more tricks than just a decently good AC to stand up the the pounding they will take. But then at that level you have had the chance to take several feats so that's fine.

I think ultimately my point is that a game in which you're only facing CR=Character level or higher enemies is one where AC will taper off in use and that's debatably fine, but that DMs should be making sure to spread out CRs for the fights, because it'll let different party members shine better. Variety is the spice of life, etc. A group of CR3s might not be Deadly to a party of 10s, but they're an important part of the gameplay loop/ecosystem to help players feel satisfied with their characters.

[Unoriginal]

Two of your saves increase, the other four do not. And in this case this is a thing discussed in the video, that one of the goals of bounded accuracy is to allow participation of non-specialists. Saves don't really work like this because DCs scale beyond what most characters can reasonably resist.

As demonstrated by others above, that is really not the case.

I think the quote was something like, 'success and failure should be determined by the player's cleverness, not what's on their character sheet'
So that is me using agency wrong, what was more meant is saves only care about the character build step, not actions during play.

No offense, but the quoted claim is absurd.

D&D isn't and has never been like a Dark Soul game, where you can play the game as a level 1 naked hobo with all stats at minimum and not only beat the game, but beat it without being hit by any enemy attack.

What's written on your sheet are your main problem-solving tools. Player cleverness plays a role, of course, but it's mostly being clever about what is written on the sheet, or being clever about acquiring things you can write on the sheet for later.

Puzzle bosses can be great, and I'm always thrilled when a player has a clever idea that changes a whole encounter, but you still needs the stats and features and equipment to pull it off. Figuring out the BBEG is paralyzed by cold damage thanks to context clues is awesome, but you still need a way to inflict cold damage reliably and a way to muscle through the BBEG's other defenses if you want to capitalize on that weakness.

[Ignimortis]

In Short,
-Saves are busted because DCs are expected to increase over time but saves are not, removing one of the goals of increased player agency without being inhibited by needing to significantly plan out a character

True. Proficiency-related bonuses in general scale too slowly - you just barely keep up with DCs IF it's both your main stat and you're proficient. If you're non-proficient or it's not your main stat, you're still falling behind somewhat. If neither apply, you go from ~40% chances of success to 5% over 20 levels.

- The math of the game is fundamentally off because attack bonuses and damage values of monsters increase by CR, but not PC AC values, leading to a feeling of characters getting weaker as they level rather than stronger

True also. I have had characters with ACs of 25 to 28, but those were the exception, and they were not the characters who were supposed to be heavily armored - in fact, heavy armor wearers were the least armored and durable unless using a shield.

- the lack of need for mechanical bonuses from other sources leading to looser game design with multiple areas that are easy to exploit

Not sure what that refers to.

But the thing I am more interested in, Is this nugget. Its a blog from 2018 (so definitely not new information), but the idea it posits is interesting. It is essentially a list of observations comparing the DMG monster guidelines and Monster Manuel statblocks and noting some weird stuff, in short:
- Monsters tend to have lower HP and expected damage across the board than the DMG would have us believe
- Monsters tend to have higher attack values then what the DMG would have us believe
- Monster HP, AC, Damage and Attack bonuses have mild correlation. The DMG would have us believe high attack bonuses would be balanced by lower damage and vice versa. but in practice there is little to suggest that in the MM. High AC monsters have comparable HP to low AC monsters and High Attack bonus monsters have comparable damage to Low Attack Bonus monsters.
- Of particular personal validation, Resistances and Immunity have no correlation to a monsters HP to a given CR, which makes sense as vulnerability does not according to the DMG, despite being the same thing from a game perspective presented differently

Does this match up with what people have observed using these systems? What are your opinions on any of this?

MM statblocks are notoriously bad in several ways, yes.

[Witty Username]

No offense, but the quoted claim is absurd.

Non taken, its a paraphrase of Wotc's words in that case, not mine own.

The video cited an article towards the stated goals of using bounded accuracy, with the video noting it failed in that goal. I can try to find a link in case I have done it an injustice.

Whether it was a good idea to begin with? That is for someone above my pay grade.

I do think the save numbers could use touch ups, but I think that is probably true of any game with fiddlly numbers.

Not sure what that refers to.

At least the argument made in the video, was that 5e has generally trended bad on editing, and doesn't have systems in place to control stacking bonuses, the system doesn't expect bonuses often but when you can get them they can add up quickly.
The video gave an example of two of the Tasha's spell boosting items:
the wizard one that boosts wizard spells
and the cleric one that boosts spells (with the cleric portion omitted)
and so the items are stackable if the caster meets the criteria
This came off to me as more 5e has poor editing than indicative of a fundamental concern.

The more interesting example was talking about how the system math can be broken at low levels, giving an example of a fighter with the following effects:
Archery
Bardic Inspiration
Bless
and a couple other things

This one was more interesting to me because it comes off as less a writing mistake, and more how stacking bonuses can get out of hand very fast, since 5e doesn't have many but doesn't have the limits imposed by other systems (like how 3.5 limits the kind of bonuses that can stack).

[Witty Username]

Reading through the article and confirming my impressions, I do think the sources are better than my explanations are providing.

First of all,Here is the video that started this chain. I figure the link at this point is just an aid, not a sight to people who don't watch this stuff.

Second, The Wotc article discussing the why. On a reread, I think the video maker may be conflating the design goals of Ability checks and saves, or simply noting that saves don't work in accordance to the non-specialist design goal even if it wasn't explicitly included:

Nonspecialized characters can more easily participate in many scenes. While it's true that increases in accuracy are real and tangible, it also means that characters can achieve a basic level of competence just through how players assign their ability bonuses. Although a character who gains a +6 bonus to checks made to hide might do so with incredible ease, the character with only a naked ability bonus still has a chance to participate. We want to use the system to make it so that specialized characters find tasks increasingly trivial, while other characters can still make attempts without feeling they are wasting their time.

I think that saves do kinda fail at this, point buy stats don't really have a lot of range to cover gaps, as Con because everyone needs HP, a combat stat because everyone is expected to perform well in combat, and usually some dex investment because everyone needs AC are rather investment hungry, it can be done that you have a wide range of competent saves, but it requires something a bit odd, like dumping one save ability the class gets proficiency in: intelligence for rogues, wisdom for Warlocks, strength for rangers, etc.

And saves are somewhat difficult to trivialize without DM assistance. Take say an enemy mage, as the stat block, its CR 6 with a DC 14, even a hard point at this level (invested score + prof) is only going to be around +6, or a little over a 1/3 chance of failure.

And we already have discussed a bit on how increasing save DCs out scale characters with little investment, something unusual like strength or intelligence saves are very good at this as most characters have little incentive to invest in such, without deliberate effort on the DM anyway.

[Ignimortis]

Non taken, its a paraphrase of Wotc's words in that case, not mine own.

The video cited an article towards the stated goals of using bounded accuracy, with the video noting it failed in that goal. I can try to find a link in case I have done it an injustice.

Whether it was a good idea to begin with? That is for someone above my pay grade.

I do think the save numbers could use touch ups, but I think that is probably true of any game with fiddlly numbers.

Personally, I would agree. The only area of the game where bounded accuracy sort of works is lower levels (1-13, maybe) and to-hit/AC only. Saves are fundamentally broken, skills are probably even more broken due to the "everyone gets to try" philosophy which means that without expertise and high stats, you can't reliably do anything the other people can't do sometimes.

At least the argument made in the video, was that 5e has generally trended bad on editing, and doesn't have systems in place to control stacking bonuses, the system doesn't expect bonuses often but when you can get them they can add up quickly.
The video gave an example of two of the Tasha's spell boosting items:
the wizard one that boosts wizard spells
and the cleric one that boosts spells (with the cleric portion omitted)
and so the items are stackable if the caster meets the criteria
This came off to me as more 5e has poor editing than indicative of a fundamental concern.

The more interesting example was talking about how the system math can be broken at low levels, giving an example of a fighter with the following effects:
Archery
Bardic Inspiration
Bless
and a couple other things

This one was more interesting to me because it comes off as less a writing mistake, and more how stacking bonuses can get out of hand very fast, since 5e doesn't have many but doesn't have the limits imposed by other systems (like how 3.5 limits the kind of bonuses that can stack).

True in a certain way, which makes it all the funnier because those stacking bonuses are often larger than what regular improvements (PB/stat mods) provide.

And saves are somewhat difficult to trivialize without DM assistance. Take say an enemy mage, as the stat block, its CR 6 with a DC 14, even a hard point at this level (invested score + prof) is only going to be around +6, or a little over a 1/3 chance of failure.

And we already have discussed a bit on how increasing save DCs out scale characters with little investment, something unusual like strength or intelligence saves are very good at this as most characters have little incentive to invest in such, without deliberate effort on the DM anyway.

A CR1 or CR2 enemy generally has an ability DC of 10 to 12. A proficient+stat-focused level 1 character rolls at +5, succeeding on a 5 to 7 against those. This certainly feels like a decent save. A CR20 to CR24 enemy generally has an ability DC of 20 to 24. A proficient+stat-focused level 20 character rolls at +11, succeeding on an 9 to 13 against those. Somehow, their chances against an on-level enemies have gotten worse, not better. Any character who isn't proficient or not focused in that stat is getting way worse (from a circa 50 to 40% chance at level 1, to 5% chance at level 20).

Compare 3.5's good and bad saves, even on a character who doesn't improve the linked stats as a priority. A character with a good save and investing into the save stat normally (~14 dex, 14 con, though 14 wis shows up a lot less often) has a +4 good save, +1 or +2 in a bad save at level 1, vs common abilities that do something along the lines of DC12 to 14 - so it's around 60 to 50% success rate on a good save and 50 to 35% on a bad one.

Twenty levels pass, and going solely by expected progression (+5 cloak of resistance, single class levelling without focusing on those stats, enhancement items of at least +4 for all those stats, maybe a couple inherent bonuses through Wish or Tomes of X, so a final ability score of 20 perhaps), we end up with +23 for good saves, +17 for bad ones. Again, these are characters who are NOT investing in those stats primarily - a primary focus would be at 30 or above by that point, boosting the save by 5 or more.

A 3.5 balor (CR20) hurls around abilities with DCs ranging from 23 to 27 - good saves pass on 2 (nat 1 autofails) to 4, bad saves pass on 6 to 10. A marked improvement over level 1 versus the same-level opponent. A CR24 wyrm red dragon (notoriously underCR'd for some things they do) has two abilities at DC38 and 35 - their breath weapon and Frightful Presence respectively, otherwise they throw around spells at circa DC25. Even so, a good save with very basic investment passes DC vs breath on 15 and DC vs presence on a 12, and continues to pass vs dragon's spells on a 2 to 4. Bad save is the only one where we see the same situation as 5e - nat 20 is required to make the save vs breath, and 18+ to make the save vs Presence, except it also still passes vs dragon's saves on at least a 10, usually less.

TL;DR: Saves scale extremely poorly in 5e. There is no improvement, only gradual decline in chances of actually making the save. Previous editions have handled that better and I'm not sure what the issue was for 5e.

[Amnestic]

3.5 character gets decked out in expected magic gear but the poor 5e guy can't even scrounge up a Cloak of Protection :(