Looking for Armor Class Analysis

[Ghost Nappa]

I recall a while back (maybe even 2014?) there was a thread analyzing PC AC Vs. Monster-to-Hit bonuses to see at what point the diminishing returns kicked in. I've poked around the useful threads a couple of times but can't find it. Does anyone know where I can find it or a similar resource?

[Silkensword]

http://www.giantitp.com/forums/showthread.php?517267-Comprehensive-DPR-Calculator

You can use this to determine diminishing returns VS any given attack bonus.

In general, if you're planning an ultra AC build, aim to get an amount of AC where the enemy needs to get a critical hit to get you. At that point, more points will be wasted. a grave cleric can make it so that if you get critted but your AC would catch the attack if it wasn't a crit you get missed instead, since they turn the crit into a regular attack, which your AC could cover, but that is a niche application.

[DrowPiratRobrts]

http://www.giantitp.com/forums/showthread.php?558345-Does-Increasing-AC-Work-on-a-Curve

I asked a similar question not too long ago. Our discussion might be helpful. Warning that it gets pretty mathy at some points though.

TLDR: This is really campaign dependent. It's totally based on what sorts of creatures you're fighting and how many at which CRs. That would give you an average "To hit" bonus across the campaign and you'd be able to find the most efficient AC, the threshold of unhitability, etc.

That's all really theory based and hypothetical though. You're not going to know the average "To hit" bonus of a campaign even as the DM, so my advice is to just get AC as high as you can without ruining your build by investing too many resources into it.

Sidenote: Cloak of Displacement is a great way to help with AC at any level and typically costs less in resources/time than getting even a single extra point at higher levels of AC.

[Galithar]

http://www.giantitp.com/forums/showthread.php?517267-Comprehensive-DPR-Calculator

You can use this to determine diminishing returns VS any given attack bonus.

In general, if you're planning an ultra AC build, aim to get an amount of AC where the enemy needs to get a critical hit to get you. At that point, more points will be wasted. a grave cleric can make it so that if you get critted but your AC would catch the attack if it wasn't a crit you get missed instead, since they turn the crit into a regular attack, which your AC could cover, but that is a niche application.

That only protects you from the critical hit. A natural 20 (auto hit) and a critical hit (Double the damage dice rolled and can occur at times other than a natural 20, such as a Champion Fighters expanded crit range) are not the same thing. A 20 hits at all times regardless of any outside effects (exception: Lucky re-roll).

[LudicSavant]

Diminishing returns kicks in at the point that an enemy needs a 20 to hit you. Any +1 to AC before that actually has increasing returns (e.g. the difference between 20 and 21 AC is bigger than the difference between 19 and 20 AC).

To understand why this is, think of it this way:
To reduce an enemy's DPR by half (not counting critical hits), you need to reduce an enemy's hit chance from 100% to 50%. To halve it again you only need to reduce it from 50% to 25%. To halve it again you only have to go from 25% to 12.5%. And so forth. Crits tip this scale a bit, but the general pattern still holds. And that pattern gives us an exponential curve for AC's value.

So while each +1 represents -5% to hit chance, each 5% is worth more hit points than the last.


In practical play, the effect of this is most pronounced against mook swarms. If you're fighting like 30 bandits as a low-CR encounter, those guys can kill you in one turn if you have low AC but high health, or do practically nothing if you have a bit higher AC and grant Disadvantage (at which point your Cleric can just wade into the swarm while Dodging and watch Spirit Guardians eat all the mooks).

[JakOfAllTirades]

http://www.giantitp.com/forums/showthread.php?558345-Does-Increasing-AC-Work-on-a-Curve

I asked a similar question not too long ago. Our discussion might be helpful. Warning that it gets pretty mathy at some points though.

TLDR: This is really campaign dependent. It's totally based on what sorts of creatures you're fighting and how many at which CRs. That would give you an average "To hit" bonus across the campaign and you'd be able to find the most efficient AC, the threshold of unhitability, etc.

That's all really theory based and hypothetical though. You're not going to know the average "To hit" bonus of a campaign even as the DM, so my advice is to just get AC as high as you can without ruining your build by investing too many resources into it.

Sidenote: Cloak of Displacement is a great way to help with AC at any level and typically costs less in resources/time than getting even a single extra point at higher levels of AC.

One word of warning on the Cloak of Displacement: a low-AC member of our party got one recently and it hasn't helped much; it's quite common for monsters attacking with disadvantage to succeed on both rolls and just ignore its effects. For this item to work, you've got to have a decent AC to begin with.

[LudicSavant]

I recall a while back (maybe even 2014?) there was a thread analyzing PC AC Vs. Monster-to-Hit bonuses to see at what point the diminishing returns kicked in.

If an enemy hits you on 19 or 20, and you get just +1 AC, then you've straight up cut their hit chance in half. That's a much more dramatic increase than if they hit you on a 2-20 and you get +1 AC.

In other words, AC has increasing, rather than diminishing, returns for your effective durability up to a maximum point (e.g. where they have to roll 20s to hit you).

[Snails]

One word of warning on the Cloak of Displacement: a low-AC member of our party got one recently and it hasn't helped much; it's quite common for monsters attacking with disadvantage to succeed on both rolls and just ignore its effects. For this item to work, you've got to have a decent AC to begin with.

Both claims can be true. It may be the most cost effective defense for that PC, while also not helping much.

[MaxWilson]

If an enemy hits you on 19 or 20, and you get just +1 AC, then you've straight up cut their hit chance in half. That's a much more dramatic increase than if they hit you on a 2-20 and you get +1 AC.

In other words, AC has increasing, rather than diminishing, returns for your effective durability up to a maximum point (e.g. where they have to roll 20s to hit you).

It also depends upon how monster behavior changes with AC. In the most extreme case where the monsters exclusively target the PC against whom they achieve the highest effective DPR after accounting for AC and resistances, the utility of additional AC is zero unless you are already the weakest link.

Likewise if monsters will shift to using saving throw-based attacks.

All else being equal though, yes, defenses have positive synergies, and it makes sense to have one guy with sky-high AC who can block chokepoints/etc. instead of two guys with mediocre AC.

[Snails]

If an enemy hits you on 19 or 20, and you get just +1 AC, then you've straight up cut their hit chance in half. That's a much more dramatic increase than if they hit you on a 2-20 and you get +1 AC.

In other words, AC has increasing, rather than diminishing, returns for your effective durability up to a maximum point (e.g. where they have to roll 20s to hit you).

With flat math, it is only in the oddball 5e campaign with Ye Olde Magick Shoppes and hight wealth where this is plausible.

After all, even the wee pipsqueak monsters get a +3 or +4 to hit. You have to reach at least AC 24 to worry about spending too much on going higher. But at a practical level, when you are high enough level to wonder whether AC 26 is worth it, you are fighting stuff regularly with a +10 to hit, so you are shooting for AC 30 now.

The real question is whether an additional +1 AC or +2 AC can be compared to, say, Fire Resistance. But that is so highly campaign dependent that it is probably unanswerable as a general question.

[LudicSavant]

With flat math, it is only in the oddball 5e campaign with Ye Olde Magick Shoppes and hight wealth where this is plausible.

I think you misconstrued my point a bit. The point being made is that AC generally has increasing returns, rather than diminishing returns. Not that you should actually expect to have an AC so high that enemies only hit on a 20.

The most common error when thinking about AC is that some folks have wrong intuitions about percentages and probabilities.

Which will provide a larger increase to the number of attacks you can survive?
A) Reducing an enemy's hit chance from 20% to 10%,
OR
B) Reducing an enemy's hit chance from 90% to 80%?

The person who is bad at math might say "trick question, both are +10% improvements, so they provide the same boost."

The person who is good at math will instead notice that...
In case A, +2 AC reduced the enemy's number of successful attacks by 50%.
In case B, +2 AC only reduced the number of successful attacks by ~11%.


It also depends upon how monster behavior changes with AC.

Naturally.

[DrowPiratRobrts]

One word of warning on the Cloak of Displacement: a low-AC member of our party got one recently and it hasn't helped much; it's quite common for monsters attacking with disadvantage to succeed on both rolls and just ignore its effects. For this item to work, you've got to have a decent AC to begin with.


Both claims can be true. It may be the most cost effective defense for that PC, while also not helping much.

Yeah, that's my point. It's not that Cloak of Displacement is perfect or anything. It's just a great item because the numbers support an enemy having disadvantage on attacks as being a great boon to your survivability. This was more of a tangential recommendation though.

[DrowPiratRobrts]

What we've said about AC is true, but we also have to consider the resources spent for each specific point of AC. That's where it gets tricky and becomes apparent that this isn't a zero-sum game.

If the goal is optimizing your AC in order to increase survivability, then the true goal is just surviving. At that point you have to calculate every resource you spend on increasing your AC and find out if spending that resource in some other way would have a greater net effect on your survival. This isn't easy to calculate and it's completely class/campaign/party dependent. But it's a vital part of considering any build and deciding when it's okay to stop pursuing AC and start pursuing other methods of survival.

[MaxWilson]

Also worth mentioning that the value of AC scales with the strength of a party's at-will attacks. In a Nova-oriented party, high AC will keep you alive longer but won't prevent your combat effectiveness from degrading. A team of warlocks and GWM/PAM fighters benefits more from four sets of plate armor, four Shields +2 and four Cloaks of Displacement than the equivalent team of blasty Evokers and smitey Paladins does.

[Leith]

The most common error when thinking about AC is that some folks have wrong intuitions about percentages and probabilities.

Which will provide a larger increase to the number of attacks you can survive?
A) Reducing an enemy's hit chance from 20% to 10%,
OR
B) Reducing an enemy's hit chance from 90% to 80%?

The person who is bad at math might say "trick question, both are +10% improvements, so they provide the same boost."

The person who is good at math will instead notice that...
In case A, +2 AC reduced the enemy's number of successful attacks by 50%.
In case B, +2 AC only reduced the number of successful attacks by ~11%.

Naturally.

A person who is good at math would make this observation. But apparently they won't notice that the distinction between A and B is meaningless.
If I attack 10 times and hit twice, then you get +2 AC and I attack another 10 times only hitting once. 50%, right?
What if I attack 10 times and hit 9. Then you get +2 AC and I attack 10 more and hit 8. Smaller %? True. Utterly irrelevant to the practical and the total # in reduction of hits is the same.

This is not relevant to the topic, except to say that making yourself 5% harder to hit can seem more or less potent depending on what your actual AC is and what the attack bonuses of the attackers are, but it is still 5%.

[Galithar]

A person who is good at math would make this observation. But apparently they won't notice that the distinction between A and B is meaningless.
If I attack 10 times and hit twice, then you get +2 AC and I attack another 10 times only hitting once. 50%, right?
What if I attack 10 times and hit 9. Then you get +2 AC and I attack 10 more and hit 8. Smaller %? True. Utterly irrelevant to the practical and the total # in reduction of hits is the same.

This is not relevant to the topic, except to say that making yourself 5% harder to hit can seem more or less potent depending on what your actual AC is and what the attack bonuses of the attackers are, but it is still 5%.

Your premise is false and the statements you're claiming to refute stand. There is an appreciable difference in damage reduced between removing 1 out of 2 hits in ten attempts and 1 out of 9 in ten attempts. It doesn't 'seem' more potent, it is actually measurably and mechanically more potent to increase your AC when your enemies hit % is lower, but still higher then the minimum of 5%. Reducing hits doesn't matter, reducing effective damage does. Resistance to damage is ALWAYS better then increasing your AC against the same damage. But that's off topic.

So based on MM creatures there is no attainable sustained AC (so remove affects like shield's +5 because it's only sustainable for top tier wizards) that causes AC gains to become diminishing as it's extremely difficult to get AC 30+ BUT it isn't uncommon for very high CR creatures to have +to hit over 10

[LudicSavant]

A person who is good at math would make this observation. But apparently they won't notice that the distinction between A and B is meaningless.

You're mistaken about it being meaningless.

To illustrate:
If we leave out crits (since the amount it tips the scale is unique to each monster), then the curve for effective hit points looks roughly like so:

https://i.postimg.cc/MTvcMjvC/Arcane-Trickster5save-HP.png

As we can see, the progression isn't linear. Crits change it a little bit, but the curve retains its general shape.



*snip*Your premise is false and the statements you're claiming to refute stand.

Ninja'd! :smallsmile:

[Galithar]

Ninja'd! :smallsmile:

Your refutation is superior though. Pictures explain things in a way words just can't! Haha

[MaxWilson]

A person who is good at math would make this observation. But apparently they won't notice that the distinction between A and B is meaningless.
If I attack 10 times and hit twice, then you get +2 AC and I attack another 10 times only hitting once. 50%, right?
What if I attack 10 times and hit 9. Then you get +2 AC and I attack 10 more and hit 8. Smaller %? True. Utterly irrelevant to the practical and the total # in reduction of hits is the same.

But the difference in capabilities is not the same. The number of monsters you can take on and defeat without dying increases far more in the first case than the second. Damage ratios matter.

Unless you're in a game where the number N of monsters you can take on without dying is always larger than the number M of monsters you actually encounter, because the DM takes responsibility for making sure you always face favorable odds. But in that case, why even play the game? It's pointless. No meaningful challenge. Let that DM play by himself or write a book.

[RSP]

“Higher AC is good” is great, but I think a more helpful approach would be to model when a character’s AC has hit “good enough” to focus increases on other party members.

For instance, a group of 4 has a high AC melee fighter. When is the fighter’s AC high enough that the Ring of Protection the party finds is better suited on someone else?

I’m not even going to pretend to do any math or charts but I’d imagine this is a more relevant question to Players, as taken individually without group context, the answer to “how high should my AC be?” is always “higher AC is better.”

I understand “good” AC is relative but for informative purposes, I’m curious if the math helps with general ranges or anything within the context of a group, as opposed to individual white room AC.

[DrowPiratRobrts]

“Higher AC is good” is great, but I think a more helpful approach would be to model when a character’s AC has hit “good enough” to focus increases on other party members.

For instance, a group of 4 has a high AC melee fighter. When is the fighter’s AC high enough that the Ring of Protection the party finds is better suited on someone else?

I’m not even going to pretend to do any math or charts but I’d imagine this is a more relevant question to Players, as taken individually without group context, the answer to “how high should my AC be?” is always “higher AC is better.”

I understand “good” AC is relative but for informative purposes, I’m curious if the math helps with general ranges or anything within the context of a group, as opposed to individual white room AC.

That's difficult because you can answer two different ways. Those who choose to answer mathematically would say that as long as you're getting hit by more than just a 20, it's worthwhile to give the ring to the player with the highest AC. They will have the "most efficient" or "best" use for the item as depicted in the chart above (the jump from 10% hit chance to 5% hit chance is exponentially larger than the previous jumps, say 20% to 15%).

Then you could also answer the more realistic way and say...it depends. It's something the party has to evaluate at each stage of the campaign. It's even advisable to mix and match AC items depending on what the players know is coming. Are they about to fight a bunch of low CR goblins? Then the fighter with 27 AC can probably give a few things to the wizard even though he's lowering his AC to 24 or so. Is the party about to fight a big boss with a great "to hit" bonus? Maybe give the fighter as many boons to AC as possible. The same wizard might let the fighter attune to his Cloak of Displacement to help the fighter as he tries to draw agro from the single boss.

As with many things in life, the math helps us understand in a vacuum, but it's still incredibly nuanced and situational.

[DrowPiratRobrts]

A person who is good at math would make this observation. But apparently they won't notice that the distinction between A and B is meaningless.
If I attack 10 times and hit twice, then you get +2 AC and I attack another 10 times only hitting once. 50%, right?
What if I attack 10 times and hit 9. Then you get +2 AC and I attack 10 more and hit 8. Smaller %? True. Utterly irrelevant to the practical and the total # in reduction of hits is the same.

This is not relevant to the topic, except to say that making yourself 5% harder to hit can seem more or less potent depending on what your actual AC is and what the attack bonuses of the attackers are, but it is still 5%.

This isn't right as others have said. But if it helps I'd like to give an example that might help you or someone else understand the funky things math does. This isn't entirely related so if you understand and agree with the math in the graphs above, feel free to skip this comment. In basketball the 3-point shot is 50% more efficient than the 2-point shot (assuming you make both shots) at a 3:2 ratio. Many times when people play half-court on the playground they'll count by 1's and 2's instead of 2's and 3's. This makes a basket from behind the 3-point line 100% more efficient than the 2-point shot at a 2:1 ratio.

So if in the normal scoring rules Team A makes 8 2-point shots (16 points), then Team B only needs to make 6 3-point shots to win with a score of 18 pts-16 pts
.
If in the other scoring rules Team A makes 8 2-point shots (valued at 1 point) for 8 total points, then Team B only needs to make 5 3-point shots (valued at 2 points) to win with a score of 10 pts-8 pts.

Notice the difference in the final scores. Not only did Team B make less shots in the second game, but they're also further ahead. How can that be if they lead by the same number in both games? Well, if Team A makes just one more of their 2-point shots in the first game then they'd be tied at 18-18. In the second game Team A would have to make 2 additional 2-point shots just to tie. So even though it appears that the scores are basically the same, Team B is actually twice as far ahead in game two, roughly speaking, as they are in game one.

[Leith]

I'll try again. The game theory is strong, but the value judgement is off.
If I reduce an enemy's hit probability from 80% to 70% I have reduced the damage potential by 12.5%. But that 12.5% can be expressed as an actual number. If the monster does 10 damage per hit that number could be 1. If the monster does 20 damage that number is 2. (I like round numbers cuz they're easy) Regardless of what the number is it could spell the difference between my character dying or living to run away, or winning. And that remains true no matter what the damage ratios are. You will always be dealing with a finite number for damage and hit points, and you will always wish you had more. It is irrelevant how much more, because you have no way of knowing how much is sufficient. So the benefit to a character with an AC of 11 buffing it to 12 is the same as a character with 18 buffing it to 19 regardless of the math.

But also, getting hit doesn't just do damage. Paralysis, sleep, poison, max hit point loss, restraining, grappling, death, and alien parasites are all things that can happen to you just by getting hit with an attack. There isn't necessarily a corresponding damage potential to consider with these attacks because they make your character less effective at what they do (whatever that might be), not just easier to hurt.

Oh, @DrowPiratRobrts, I don't see how the basketball analogy fits here.

[RSP]

I'll try again. The game theory is strong, but the value judgement is off.


I was thinking more in terms of groups rather than white-room single character. For the single character, more AC is always better, in and of itself.

I’m just curious what the “advanced analytics” would be for group AC theory. I’m not a statistician, but I imagine there could be an AC that’s effective vs the average to hit from CRs for level ranges, that could then lead to a general statement of “once the primary tank has an AC of X, it’s more valuable to increase the secondary tank’s AC to X,” and on and on.

I know it’s not going to cover everything, but I have to imagine the info is there. Also, I’m not asking you, Leith, to do this, just expounding in what I believe is more helpful to questions like the OP’s.

That is, the answer of effective AC for level and party role, is a better way to judge than just “characters benefit from higher ACs”

[Chronos]

Quoth JakOfAllTirades:

One word of warning on the Cloak of Displacement: a low-AC member of our party got one recently and it hasn't helped much; it's quite common for monsters attacking with disadvantage to succeed on both rolls and just ignore its effects. For this item to work, you've got to have a decent AC to begin with.
Advantage/disadvantage in general is most significant, compared to a numerical plus or minus, when the base chance of success is as close as possible to 50%. At a base 50% chance, it's equivalent to a ±5. But if a task is so hard that you only succeed on a 20, or so easy that you only fail on a 1, it's worth less than a ±1. The only exception to this is if you're making a roll that auto-succeeds or auto-fails on a natural, and you're at the point where an additional +1 literally doesn't matter at all.

[OverLordOcelot]

I was thinking more in terms of groups rather than white-room single character. For the single character, more AC is always better, in and of itself.

I’m just curious what the “advanced analytics” would be for group AC theory. I’m not a statistician, but I imagine there could be an AC that’s effective vs the average to hit from CRs for level ranges, that could then lead to a general statement of “once the primary tank has an AC of X, it’s more valuable to increase the secondary tank’s AC to X,” and on and on.

I don't think this is something you can answer by doing any reasonable math. You have to figure out how hits are generally distributed around the group based on the player's actions and enemy actions, and since this isn't an MMO with deliberately 'dumb' enemies using aggro scores, you don't really end up with a simple set of rules to model. For example, exactly what do you mean by a 'primary tank' and 'secondary tank'? How does one qualify as a 'tank' versus a 'damage dealer' (or whatever the distinction is) or 'primary' vs 'secondary'? Since enemies aren't forced to follow the aggro leader like in MMOs, you can't generally force them to attack anyone in particular. Feats like sentinel help direct attacks, but don't force them, there's no 'this is who will take hits from the boss, this is who will take them from minions, everyone else will just get the AOE' that you can rely on.

The end result of any model is going to be extremely tied to group composition and tactics, I don't think you're going to get a simple 'once a primary tank's AC hits 20, buff the secondary tank until they hit 20' rule, you're going to have basically an entire discussion of 'what does the party do', then do some math on hits with that, and have to bear in mind that if anything in the mix changes your math becomes irrelevant.

[RSP]

I don't think this is something you can answer by doing any reasonable math.

Every situation has its own nuisance, sure, but there’s still reasonable conclusions onecan draw. I think it’s easiest to show with the example of the Ring of Protection example: anyone could use +1 AC, but at what point does the +1 AC benefit the high AC character vs the high AC character has “high enough” AC at the current level and so there’s a greater advantage to giving it to another character.

And, yes, not every group has a designated “tank,” but most groups have varied roles between their characters, and the character who most often is involved in melee is probably eating more attacks than others because they’re pretty much always getting right in the monster’s face; whereas a ranged character or caster is intentionally staying away from threats.

So again, I think one could use an average of appropriate-CR-monsters-to-PC-level to-hit bonus compared to ACs to find approximate levels where characters hit the point where a lower AC benefits more from increases.

[LudicSavant]

I don't think this is something you can answer by doing any reasonable math. You have to figure out how hits are generally distributed around the group based on the player's actions and enemy actions, and since this isn't an MMO with deliberately 'dumb' enemies using aggro scores, you don't really end up with a simple set of rules to model.

Not only is math relevant here, but we actually have an entire branch of mathematics (called game theory (https://en.wikipedia.org/wiki/Game_theory), incidentally) that deals specifically with this kind of problem: accounting for the behavior of an intelligent foe and optimizing your choices accordingly.

[PeteNutButter]

Not only is math relevant here...

Couldn't agree more. As the guy who almost always plays the party "tank" I've had to explain this math countless times. The short answer is if the party finds a ring of protection it should go first to the guy with the high AC. Here's why:

1.) Not only is every point better "until they have to crit to hit," but since D&D is a game where the DM will always try to present a challenge, that "crit to be hit" goal is almost never attained, except against mooks. Any competent DM will have their bosses/BBEGs scaled in CR or + to hit so that they can at least have a small chance of being a threat. Not being hit by a boss, is about as important as it can get. If your local sentinel feat having tank can stand toe to toe with the big bad, a few rounds longer the party is in better shape.

2.) It becomes increasingly harder to find items and abilities to stack AC higher and higher, as any character currently stacking AC is going to be actively using as many as they can get their hands on. In contrast the wizard for instance can do a variety of things to boost his AC.

3.) Specialization is very valuable. Just as the AC tank should expect first choice of AC boosting items, so too should the GWM fighter expect first pick at the belt of MOAR strength. Would you offer the Oathbow to the strength fighter when there is a dex ranger in the party wanting it? Characters should have priority when picking things that will benefit them most.

[Chronos]

But additional AC doesn't become useless at the point where the enemy can't hit you. It becomes useless at the point where the enemy stops trying to hit you. Any reasonably intelligent enemy is going to ignore the guy in the full plate and instead focus attacks on the guy in the dress, until the guy in the dress is taken out. If they do try anything on the guy in full plate, it's going to be something with a saving throw, not an attack roll. So the +1 AC item is going to be much more useful on the wizard, where it does something, than on the fighter, where it does nothing.

[DrowPiratRobrts]

Oh, @DrowPiratRobrts, I don't see how the basketball analogy fits here.

That's fine. It's not a 1 to 1 correlation. I was just explaining the math in a different way for those who may not get it. I couldn't tell if you weren't understanding the math or if you understood what they were saying about AC being more efficient/valuable at higher AC and were just making a different point.


But additional AC doesn't become useless at the point where the enemy can't hit you. It becomes useless at the point where the enemy stops trying to hit you. Any reasonably intelligent enemy is going to ignore the guy in the full plate and instead focus attacks on the guy in the dress, until the guy in the dress is taken out. If they do try anything on the guy in full plate, it's going to be something with a saving throw, not an attack roll. So the +1 AC item is going to be much more useful on the wizard, where it does something, than on the fighter, where it does nothing.

That's highly DM/game dependent. Some DMs roll randomly every round to decide which NPC attacks whom. I typically have NPCs target the player they perceive as the most immediate threat to them. A lot of times that's the huge guy with a sword standing right in front of them and not the guy in a dress 50 ft away. It doesn't do them any good to stop the Wizard that may or may not hurt them if the Fighter right in front of them with the axe just chops off their head.

There's not one way to know for sure when enemies will stop targeting you because of your high AC. It's highly game and situation dependent.

[LudicSavant]

Answering who to give a Ring of Protection to is a balancing act requiring you to weigh a variety of different factors, including but not limited to...

1) Effective durability gain. A +1 AC might be 3 effective hit points for one character, and 60 effective hit points for another in the same party (see the graph above for why this is). It can potentially be a pretty massive discrepancy. Resistance to status effects riding on AC-targeting attacks scales on a similar curve.

2) The ability of a tank to leverage their defenses to help others. There are two main facets of this. The first is that a truly good tank makes themselves very difficult to ignore; going around them is not free and an enemy needing to go around them to hit a squishy is not useless. The second is that even if a frontliner isn't dying, them taking less damage / status effects might divert resources elsewhere (for example, a slot that might have been used on healing or buffing the tank's defense might instead go to a control spell or enemy debuff). Both of these factors can translate into effective durability for party members.

3) Enemy behavior. Rational enemies will prefer to target whatever will give them the most rewarding outcome, and that can and should be accounted for.


This is all doable, but you aren't going to get an oversimplified universal answer like "AC 25 is enough, after that give the Ring to someone else, regardless of class abilities and the like." You're going to get an answer like "in this specific party composition, with these specific abilities, assuming this kind of opposition, you will probably get the most benefit from giving the Ring to party member X."

[Silkensword]

Also, extremely high AC characters don't necessarily look like they have extreme AC's- Bladesinger, Monk, Barbarian come to mind.

[OverLordOcelot]

Every situation has its own nuisance, sure, but there’s still reasonable conclusions onecan draw. I think it’s easiest to show with the example of the Ring of Protection example: anyone could use +1 AC, but at what point does the +1 AC benefit the high AC character vs the high AC character has “high enough” AC at the current level and so there’s a greater advantage to giving it to another character.

If you think that there's reasonable math that will show this, then I'm perfectly fine with you using the easiest example (the Ring of Protection) to show that I'm wrong. Just do the reasonable math, the way LudicSavant posted the math for effective HP, and it will show that my assertion is silly and unrealistic.

But I will note that:

And, yes, not every group has a designated “tank,” but most groups have varied roles between their characters, and the character who most often is involved in melee is probably eating more attacks than others because they’re pretty much always getting right in the monster’s face; whereas a ranged character or caster is intentionally staying away from threats.

The thing is, this doesn't actually tell me what the difference is between a "Primary Tank" and "Secondary Tank". Before you can actually calculate the answer to whether the ring of protection should go to the secondary tank or primary tank, you're going to have to say what each one even is! I'm skeptical that there's a good, universal definition for this, but without a solid definition of primary vs secondary tank, there's no way to come up with something that lets you plug in primary and secondary tank ACs and see which one benefits more.

And this is just the first step on constructing a model to see who benefits most from the ring. You're going to need to model enemy target selection - whatever numbers you come up with are going to look really different if you assume the DM only ever targets high AC people or always tries to ignore high-AC people in favor of low AC people, or if the follows some simple rule (nonintelligent creatures always target whoever hit them last), or tries to roleplay in some specific way ('they prefer to go after the guy in robes because he looks easy to kill'). And you're going to need to model people's time as a possible target - rogues are often going to try to spend a bunch of the fight hidden, casters and other ranged casters will use flight and distance to avoid being a target (especially at high levels when teleport and flight abilities/items become common), and so on.

I don't think you're going to come up with a simple way to model all that, but I'd love to see see someone prove me wrong and post the model.


Not only is math relevant here, but we actually have an entire branch of mathematics (called game theory, incidentally) that deals specifically with this kind of problem: accounting for the behavior of an intelligent foe and optimizing your choices accordingly.

If you think there's a reasonable mathematical answer to the question using game theory, then why not post it like you did for effective HP and prove me wrong? I'm well aware that game theory exists, and said nothing to indicate that it doesn't. What I think doesn't exist is a reasonable mathematical model that someone can actually put together in pseudocode form or on a website that will take AC numbers give a meaningful answer of which one will benefit most from a +1 to AC. Constructing some model that will spit out some answer is easy, but constructing one where the answer will actually be reasonable is, I think, an extremely hard problem that no one is going to just post a solution for.

[PeteNutButter]

If you think there's a reasonable mathematical answer to the question using game theory, then why not post it like you did for effective HP and prove me wrong? I'm well aware that game theory exists, and said nothing to indicate that it doesn't. What I think doesn't exist is a reasonable mathematical model that someone can actually put together in pseudocode form or on a website that will take AC numbers give a meaningful answer of which one will benefit most from a +1 to AC. Constructing some model that will spit out some answer is easy, but constructing one where the answer will actually be reasonable is, I think, an extremely hard problem that no one is going to just post a solution for.

D&D games are pretty complicated with loads of variables. Even so, you could take a simple shot at it with DPR (average Damage Per Round) and RTD (Rounds to Death). So in the simpliest representation of D&D 5e, your total character effectiveness is your DPR times your RTD. In other words, you'll do your damage every round until you die. Whichever character has the most % increase in their total effectiveness should get the item, which in this case is a Ring of Protection.

For an example I'll compare two characters that are in my AL group. Both PCs are 15th level, against a DMG-assumed-stats CR 15. One is a GWM Fighter with a barbarian dip for rage and reckless attack, the other an AC tank S&B Paladin. I'll just do sustained DPR for simplicity, but I could also factor in a burst DPR.

The GWM Fighter that will put out a sustained DPR of 57.5 when factoring for hit chance, but will take ~6.4 rounds to die (generously assuming rage is active and physical damage). So his total effectiveness is 368, which is the expected amount of damage he'd do before going down if he were toe to toe with a CR 15. The other is a AC Tank Paladin with a sustained DPR of 35.7, but an insane RTD of 11.1, making is total combat effectiveness 396.7, meaning he can expect to deal almost 400 damage before a CR 15 would take him down (using just attacks).

Should the GWM Fighter pick up the Ring of Protection his AC boost will go up one, and raise his RTD to ~6.6. It's not a big step up, but it does bring his total effectiveness up to 379.9 which is a 3% increase.

Should the AC tank pick up the Ring of Protection it, takes that CR 15 into "crit to hit territory" on the paladin. Since he has disadvantage to be hit almost always that makes his RTD skyrocket to ~55. That should raise his total effectiveness to 1963, which is almost a 400% increase. A 400% increase makes that GWM Fighter's 3% just laughable.

I use these extreme examples to mathematically demonstrate my point. The characters shown above are relatively equal in total combat effectiveness at the start just around 400, but the ring has a much greater effect on the paladin. Before anyone says, "but, Pete, fights don't go 55 rounds." Keep in mind, that 55 rounds is just a baseline against a single CR 15. If the DM tosses 10 of them at the paladin, the "unkillable paladin" goes down in just 5.5 rounds (as opposed to 1.1 rounds without the ring).

Just because the math has many variables, doesn't mean that it can't be done. It just needs to be solved on a case by case basis. What you're asking for is an answer, when really the answer is an equation that can provide future answers. So DPR * RTD is your springboard. Go from there.

EDIT: I would point out that this could potentially create a spiral situation in the extreme scenario where one player is optimized to the point of doing way more damage and being able to survive way more, and thus should get all lewts. Be reasonable, it's a team game.

-It's assumed that the goal of any encounter is to first avoid a TPK. The RTD feature will irrelevant in most fights, meaning the GWM Fighter will do significantly more for the party. It's only when a TPK is possible that the tank is even needed. It doesn't matter that the fight won't go 55 rounds. It matters that the tank can live through enough that he can end fights by himself if he has to (hopefully taking his stabilized allies to safety afterwards).
-This is all assuming monsters are attacking willy nilly. If your tank character has some sort of "aggro-control" (sentinel, grapple, etc) his RTD are effectively added to the rest of the party's as he needs to die first, thereby making his RTD even more critical.
-You could factor in spells as well. CC spells are just a % chance to add to everyone's RTD. They prevent incoming damage.
-These assumptions can greatly shift based on foes, but they are a good ballpark for most encounters.
-The reason the numbers were such a huge difference is the GWM Fighter is almost always recklessly attacking, making the + 1 AC almost worthless, while the paladin rocks a cloak of displacement making that AC priceless. These are not crafted example characters to prove my point. They are real characters that are played by me and my nephew, mine being the tank. He tends to do a ton more damage, but I can save the day when the party is in threat of TPK.

[Chronos]

So, if the Unkillable Tank just keeps going and going and going, and isn't down until he's fought for 55 rounds, but in that time he eventually kills all of the enemies, do you count that as a victory? I wouldn't. Because in that time, all of the other party members whose RTD is less than 55 are already dead. All but one of the party members dying is a terrible outcome.

For a truly successful fight, one where none of the heroes die, you need to multiply everyone's DPR by the RTD of the least durable character, and therefore, whenever you find an item that boosts survivability, you want to give it to the character who's currently least durable.

[OverLordOcelot]

D&D games are pretty complicated with loads of variables. Even so, you could take a simple shot at it with DPR (average Damage Per Round) and RTD (Rounds to Death). So in the simpliest representation of D&D 5e, your total character effectiveness is your DPR times your RTD. In other words, you'll do your damage every round until you die. Whichever character has the most % increase in their total effectiveness should get the item, which in this case is a Ring of Protection.


Like I said in the material you quoted, you can slap together some model that will spit out some answer, but making one where the answer will be reasonable is an extremely hard problem that no one is going to be able to just post a solution for. You have slapped together a model that spits out an answer, but the model you posted doesn't spit out anything like a reasonable answer, and makes a lot of bad assumptions.

At a basic level, your model is based around trying to salvage a near TPK situation for level 15 characters facing CR15 enemies. A CR 15 boss like a mummy lord or adult green dragon makes a rough fight where 'avoid a TPK' is relevant for Tier 2 characters, but they're just a mild resource drain for level 15s. Your model also relies on the DM using laughably bad tactics against the characters in that you assume that enemies "almost always" have disadvantage on attacks against the paladin. Enemies can get around the disadvantage by grappling the paladin (uses STR instead of an attack roll, so your AC is irrelevant), doing small AOE damage (extremely common for boss/lair fights in my experience), or just using the 'help' action. As a quick example, an adult green dragon (CR15) can negate your cloak for a round with his breath weapon, and can force you to make a Dc19 dex save every round or lose the cloak's benefit and be knocked prone (switching to advantage from disadvantage) by spamming his wing buffet legendary action.

"We should give the ring to the paladin because if we manage to almost lose to a speed bump but he's still up then he would get the biggest jump in his ability to take down the enemy that shouldn't have been a real threat in the first place as long as it doesn't do anything sensible when fighting him" is not a reasonable answer to the question, it's just laughable. Your analysis based on DPR and RTD doesn't actually show anything remotely applicable to people trying to allocate the item in a sensible way, it just spits out some numbers disconnected from the reality of the game.

[PeteNutButter]

Like I said in the material you quoted, you can slap together some model that will spit out some answer, but making one where the answer will be reasonable is an extremely hard problem that no one is going to be able to just post a solution for. You have slapped together a model that spits out an answer, but the model you posted doesn't spit out anything like a reasonable answer, and makes a lot of bad assumptions.

At a basic level, your model is based around trying to salvage a near TPK situation for level 15 characters facing CR15 enemies. A CR 15 boss like a mummy lord or adult green dragon makes a rough fight where 'avoid a TPK' is relevant for Tier 2 characters, but they're just a mild resource drain for level 15s. Your model also relies on the DM using laughably bad tactics against the characters in that you assume that enemies "almost always" have disadvantage on attacks against the paladin. Enemies can get around the disadvantage by grappling the paladin (uses STR instead of an attack roll, so your AC is irrelevant), doing small AOE damage (extremely common for boss/lair fights in my experience), or just using the 'help' action. As a quick example, an adult green dragon (CR15) can negate your cloak for a round with his breath weapon, and can force you to make a Dc19 dex save every round or lose the cloak's benefit and be knocked prone (switching to advantage from disadvantage) by spamming his wing buffet legendary action.

"We should give the ring to the paladin because if we manage to almost lose to a speed bump but he's still up then he would get the biggest jump in his ability to take down the enemy that shouldn't have been a real threat in the first place as long as it doesn't do anything sensible when fighting him" is not a reasonable answer to the question, it's just laughable. Your analysis based on DPR and RTD doesn't actually show anything remotely applicable to people trying to allocate the item in a sensible way, it just spits out some numbers disconnected from the reality of the game.

You miss the point of it being a baseline. I compare CR 15 for simplicity, but the math scales for any CR or group of CRs. The only scenario where it wouldn’t be better to give to the paladin is if the enemy being faced hits the paladin on anything but a 1, at which point the AC boost is pointless for both characters. As it turns out, no such enemy exists. The Terrasque’s +19 missing the paladin’s 27 AC on an a 7 or lower. The CR 15 is a baseline, but the math stays accurate against any enemy.

While it’s true that enemies can use tactics to go around AC, that is irrelevant when comparing an AC item. The ring doesn’t help either character against grappling, although both characters have great athletics anyways. It does help in saves but while that is important it comes up less often than AC, so is a secondary consideration. Even if we consider saves, the paladin is giving all nearby teammates a massive boost to all their saves. So when fighting enemies that attack with saves primarily, keeping the paladin up and on the party’s side is priority as he gives them all that boost. So again it should go to the paladin.

Yes, enemies can mitigate the disadvantage by finding a way to grant advantage (effectively making it a wash/straight roll not advantage). But that just makes the math a little bit less overwhelming. I can run the math against every enemy in the MM and every time the paladin should get the ring. That’s why my baseline tactic works.

[OverLordOcelot]

You miss the point of it being a baseline. I compare CR 15 for simplicity, but the math scales for any CR or group of CRs. The only scenario where it wouldn’t be better to give to the paladin is if the enemy being faced hits the paladin on anything but a 1, at which point the AC boost is pointless for both characters. As it turns out, no such enemy exists. The Terrasque’s +19 missing the paladin’s 27 AC on an a 7 or lower. The CR 15 is a baseline, but the math stays accurate against any enemy.

You miss the point of needing to model something useful. The math models a single stupid opponent not using any of its special abilities, it doesn't say anything sensible about how effective the paladin would be in a significant fight against an opponent actually taking advantage of its abilities and the environment. This kind of model works for MMO boss fights (at least the ones I'm familiar with) because the enemy is forced to play by a very dumb set of aggro rules, but falls apart with a reasonably competent opponent using actual abilities.


While it’s true that enemies can use tactics to go around AC, that is irrelevant when comparing an AC item.

There are two problems with this. The first is that when I listed things like grappling and saves, they weren't for getting 'around AC', they were for removing the effect of the cloak of displacement. Your model treats that treats it as having 100% uptime, which is simply not reasonable. Using tactics to go around AC and negate one of the assumptions of your model is very relevant to how useful your model is.

The second is that it doesn't make any sense to make an argument based on how well your character can survive fighting alone if just you ignore the effective attacks that can be made against him. The idea that your paladin would last 55 rounds against a green dragon while dishing out 37.5 damage per round is just silly. Why wouldn't the dragon use its wing buffet to negate your cloak of displacement during the 55 rounds, or hang out of reach and use its breath weapon? Good luck averaging 37.5 damage when you're hurling 1 javelin per round, and better hope you're carrying 55 of them to last out the fight. (If you're going to say 'but I can fly', I'm going to have to ask just how many attunement items you're packing, because at this point you've filled two of your three attunements)


Yes, enemies can mitigate the disadvantage by finding a way to grant advantage (effectively making it a wash/straight roll not advantage). But that just makes the math a little bit less overwhelming.

And they can also mitigate it by doing a single point of damage to you, like with small 'save for half' AOEs that are common on boss fights in AL in my experience, or with abilities like dragon's breath and wing buffet that the first CR15 creature I found has available. It completely negates one of the core assumptions that makes your example produce numbers so absurdly skewed towards 'give this to the paladin', and the fact that your model treats an effect that is likely to have 50% or less uptime as having 100% uptime undescores the fact that you're not modeling something useful.


I can run the math against every enemy in the MM and every time the paladin should get the ring. That’s why my baseline tactic works.

Your baseline tactic won't work against an intelligent party. They will see that your math is not modeling anything actually useful, and realize that you're just trying to stockpile items for yourself and not actually distribute them in a way that provides the most benefit to the party. I don't think "If you give this to me, in the situation where everyone else has dropped I will be able to last even longer and then will be able to drag your bodies back to town and get them all ressurected" has ever actually worked; the near-TPK situation where you last 55 rounds and finally win is the kind of thing most parties optimize to avoid, and that requires keeping the whole party alive.

[RSP]

Your model also relies on the DM using laughably bad tactics against the characters in that you assume that enemies "almost always" have disadvantage on attacks against the paladin. Enemies can get around the disadvantage by grappling the paladin (uses STR instead of an attack roll, so your AC is irrelevant), doing small AOE damage (extremely common for boss/lair fights in my experience), or just using the 'help' action. As a quick example, an adult green dragon (CR15) can negate your cloak for a round with his breath weapon, and can force you to make a Dc19 dex save every round or lose the cloak's benefit and be knocked prone (switching to advantage from disadvantage) by spamming his wing buffet legendary action.

I have to disagree with this: you’re arguing that creatures can just not attack and chose a different tactic, and calling it a “laughably bad tactic” for DMs to have enemies attack high AC characters, yet your position is basically “the DM should have all creatures the party encounters metagame away any advantages the party has and always focus on a character’s weakness.”

You use the example of grappling, but why would any given enemy know that grappling the blurry combatant takes away the blurry-ness?

Your arguments basically are meaningless because a DM can always focus on any character’s weakness, regardless of what that is, but it’ll almost always have nothing to do with in-game factors.

“Don’t give the +AC item to the Paladin because the DM will just spam grapples against him every combat” isn’t an argument against what Pete posted, it’s an argument for why that’s a horrible way to DM.

Basically, if a DM wants to beat the players, they can, easily, whenever they want; so arguing that as the reason to not use stats is a ridiculous argument.

[OverLordOcelot]

You use the example of grappling, but why would any given enemy know that grappling the blurry combatant takes away the blurry-ness?

Your arguments basically are meaningless because a DM can always focus on any character’s weakness, regardless of what that is, but it’ll almost always have nothing to do with in-game factors.

I certainly don't think it's unreasonable to have a green dragon with an 18 intelligence and over a century of life figure out that it should use a ranged AOE against the guy who's hard to hit with claw/bite attacks, or that he should use wing buffets to make melee opponents easier to attack. And that doesn't even require him to know the specifics of a cloak of displacement - but I'd expect anyone with an 18 int and centuries of living in a D&D world being attacked by adventurers to pick up the basics of how common magic items work. And I don't think that having someone say "he's hard to pin down, have the zombies/cultists/snakes/whatever pile on him so he can't move around so much" is some unreasonable metagaming nonsense, a hero being dragged down by a swarm of monsters is a pretty common iconic image (notably on the cover of Doom)

Calling it metagaming and bad DMing if a dragon uses it's wing buffet ability is... an interesting position to take.
EDIT: I also went back and looked at the other CR15s listed on roll20. Bronze dragon also has breath weapon and wing buffet, and Mummy lord has an AOE stun with a DC16 con save plus 2 casts of insect plague (unavoidable AOE damage, lasts for concentration). So of the four CR15 creatures listed on D20, only one of them does not have a listed ability that will routinely nullify the cloak of displacement, and the ability is something they would reasonably keep doing against a melee opponent even if they didn't know specifically about the cloak of displacement, and all three are long-lived, intelligent creatures that likely know how standard magic items work.


“Don’t give the +AC item to the Paladin because the DM will just spam grapples against him every combat” isn’t an argument against what Pete posted, it’s an argument for why that’s a horrible way to DM.

It also isn't an argument that I ever made, so I'm not sure why you're acting like I did. I did make the argument that "The effects of the cloak of displacement can be nullified by a variety of common abilities, one of which is grappling, another of which is AOE damage, and another of which is save-based magic, therefore treating it as having 100% uptime is not reasonable", but you haven't refuted that.

[OverLordOcelot]

So, if the Unkillable Tank just keeps going and going and going, and isn't down until he's fought for 55 rounds, but in that time he eventually kills all of the enemies, do you count that as a victory? I wouldn't. Because in that time, all of the other party members whose RTD is less than 55 are already dead. All but one of the party members dying is a terrible outcome.

Yeah, even if I was using a RTD model I would rather attempt to push everyone's RTD higher than the length of a typical fight (5 rounds or less) than to shove one character's RTD to astronomical levels. Once someone's RTD is such that they're probably going to survive any reasonable fight, pushing their RTD higher is really pretty pointless. The party that manages to have everyone to survive the boss fight with injuries is generally going to be considered more successful than the one where everyone but one guy dies, but that guy eventually manages to finish off the boss, drag all the bodies back to a town where a 9th level cleric lives, then spend a few thousand to a few hundred thousand gold to get raise dead to true resurrection cast on all of the bodies.

[strangebloke]

There are two halves to tanking.

The first is to be really resilient yourself. If you aren't resilient, then why is the enemy attacking you better than attacking someone else? Being able to dodge and/or have a spell like sanctuary helps a lot here, but the biggest thing is to have good HP and AC. Everything else is much more niche.

The second is to be able to direct attacks toward yourself. By default, every character can do the following.

allow an ally to take cover behind them against ranged attacks
Close with an enemy to discourage ranged attacks and use an OA to discourage moving out of melee range.
shove a melee-focused enemy prone so that their mobility is limited and they can't get to other targets.
Taunting them, the effectiveness of which is up to your DM.


So any character with a high AC can reasonably tank against certain kinds of opponents. Highly mobile foes, or foes fighting at long range, are much harder to 'tank' against. It's obviously also hard to tank against large mobs of people at once, but generally, it isn't the job of the tank to do crowd control.

There isn't a single 'tank' in the game that can do that well, actually. Rangers can sorta manage large groups with martial abilities, but they're better at killing large groups than tanking them. Crowd control is very much a caster thing in this edition.

[PeteNutButter]

You miss the point of needing to model something useful. The math models a single stupid opponent not using any of its special abilities, it doesn't say anything sensible about how effective the paladin would be in a significant fight against an opponent actually taking advantage of its abilities and the environment. This kind of model works for MMO boss fights (at least the ones I'm familiar with) because the enemy is forced to play by a very dumb set of aggro rules, but falls apart with a reasonably competent opponent using actual abilities.



There are two problems with this. The first is that when I listed things like grappling and saves, they weren't for getting 'around AC', they were for removing the effect of the cloak of displacement. Your model treats that treats it as having 100% uptime, which is simply not reasonable. Using tactics to go around AC and negate one of the assumptions of your model is very relevant to how useful your model is.

The second is that it doesn't make any sense to make an argument based on how well your character can survive fighting alone if just you ignore the effective attacks that can be made against him. The idea that your paladin would last 55 rounds against a green dragon while dishing out 37.5 damage per round is just silly. Why wouldn't the dragon use its wing buffet to negate your cloak of displacement during the 55 rounds, or hang out of reach and use its breath weapon? Good luck averaging 37.5 damage when you're hurling 1 javelin per round, and better hope you're carrying 55 of them to last out the fight. (If you're going to say 'but I can fly', I'm going to have to ask just how many attunement items you're packing, because at this point you've filled two of your three attunements)



And they can also mitigate it by doing a single point of damage to you, like with small 'save for half' AOEs that are common on boss fights in AL in my experience, or with abilities like dragon's breath and wing buffet that the first CR15 creature I found has available. It completely negates one of the core assumptions that makes your example produce numbers so absurdly skewed towards 'give this to the paladin', and the fact that your model treats an effect that is likely to have 50% or less uptime as having 100% uptime undescores the fact that you're not modeling something useful.



Your baseline tactic won't work against an intelligent party. They will see that your math is not modeling anything actually useful, and realize that you're just trying to stockpile items for yourself and not actually distribute them in a way that provides the most benefit to the party. I don't think "If you give this to me, in the situation where everyone else has dropped I will be able to last even longer and then will be able to drag your bodies back to town and get them all ressurected" has ever actually worked; the near-TPK situation where you last 55 rounds and finally win is the kind of thing most parties optimize to avoid, and that requires keeping the whole party alive.

First of all, I don't appreciate the personal attacks. I'm trying to have a debate on math, not attacking your character.

All your points about the cloak of displacement or whatever are unnecessary. Even without that, the math still massively favors the paladin getting the ring. I just used the example to show how bad it could be. Without that it still is about 30 RTD on the paladin.

Bringing up other enemy tactics is entirely off subject. Any enemy that is doing things to get around AC is outside the scope of where to distribute an AC boosting item. It doesn't matter at that point. If you're enemy is casting Power Word Kill, that ring of protection doesn't even matter. You'd need more max hp/healing to stop that. The debate is on where to distribute an AC boosting item. AC attacks are the only thing relevant in this debate. That is after the point of the thread.

"AC gets increasingly better for each point you have."

"But what if you get hit with a fireball?"

It just isn't part of the debate.


Yeah, even if I was using a RTD model I would rather attempt to push everyone's RTD higher than the length of a typical fight (5 rounds or less) than to shove one character's RTD to astronomical levels. Once someone's RTD is such that they're probably going to survive any reasonable fight, pushing their RTD higher is really pretty pointless. The party that manages to have everyone to survive the boss fight with injuries is generally going to be considered more successful than the one where everyone but one guy dies, but that guy eventually manages to finish off the boss, drag all the bodies back to a town where a 9th level cleric lives, then spend a few thousand to a few hundred thousand gold to get raise dead to true resurrection cast on all of the bodies.

This is a point I could potentially agree with. Although unless there is mindless AoE going on, its pretty easy to stabilize the party. In actual gameplay, in these situations I usually spend my turns splashing some heals on the squishier allies, so they can keep fighting, thus effectively extending everyone's RTD.

This is also where we get into the debate on tanking and aggro control. Remember that point I mentioned about sentinel feat or grappling. If the tank can keep the big bad in place, that greatly nullifies that foe's ability to threaten others, and makes "tanking" a lot more useful.

How about this: Show me a scenario or foe where the fighter would be better off taking the ring, as opposed to the paladin?

Edit: One more point to help put my math into perspective. It’s not that the paladin will last 55 rounds or 30 or whatever. It is that the paladin will last roughly 9 times as long as the fighter with the ring. Without the ring he only lasts about 40% longer. No fight will last that long. It just means the party can handle a lot more enemies before falling when the paladin gets it.

[LudicSavant]

Figuring out who would get the most benefit out of a Ring of Protection is something we can evaluate on a case by case basis, but one of the big problems in this discussion is that we haven't been given a specific case to solve. We haven't been given 4 character sheets and asked "who in this party should get the Ring of Protection."

I think one of the big problems here is that people seem to want there to be some oversimplified axiomatic answer that fits in a single sentence and covers all possible party compositions while requiring a minimum of variables as inputs, and that's just not going to be sufficient in this case. Neither "always give to most durable party member" nor "always give to least durable party member" is correct.

"Always give to most durable party member" is incorrect because of point 3 in my last post (http://www.giantitp.com/forums/showsinglepost.php?p=23617941&postcount=32), as well as the fact that if your allies go down, your personal survivability tends to go down too. "Always give to least durable party member" is incorrect because of a combination of points 1 and 2.

[Ghost Nappa]

I've seen a couple of comments regarding "increasing" returns on AC and I should point out for the record, that I am currently playing a tank.

I've been fortunate enough to find +1 Chain Mail as a Paladin with the Defense Fighting Style: I'm sitting at a Base 20 AC at the moment.

We've fought against some monsters (The Froghemoth) that have +10 to hit already.

I want to get a sense of at what point further investing in AC is going to be a waste of resources for "most situations" because I suspect my DM will arms-race me a bit with monsters if I rush it, which means that the squishies in my party (who are not investing in their AC) are going to be guaranteed hit with basically everything; a situation I don't want.

Until recently, I was under the assumption that 22 was probably a good point to stop, but I don't have an actual way to check that or compare that to. What is the average monster's to-hit?

My current party contains a couple of characters that like to WWE onto other monsters despite their fragility making it difficult for me to tank in a more traditional fashion and while I'm rather certain that some smarter play would help, I don't want to get grumpy at my teammates for having fun.

For additional context, we recently fought a Werewolf that specifically *refused* to try and swipe at me because of my higher AC (and it was a plot character with Legendary resistances so it stopped my attempts to Compelled Duel it). I'm trying to sort of lock down what I should doing as a Tank and I'm kind of unsure at the moment.

[PeteNutButter]

I've seen a couple of comments regarding "increasing" returns on AC and I should point out for the record, that I am currently playing a tank.

I've been fortunate enough to find +1 Chain Mail as a Paladin with the Defense Fighting Style: I'm sitting at a Base 20 AC at the moment.

We've fought against some monsters (The Froghemoth) that have +10 to hit already.

I want to get a sense of at what point further investing in AC is going to be a waste of resources for "most situations" because I suspect my DM will arms-race me a bit with monsters if I rush it, which means that the squishies in my party (who are not investing in their AC) are going to be guaranteed hit with basically everything; a situation I don't want.

Until recently, I was under the assumption that 22 was probably a good point to stop, but I don't have an actual way to check that or compare that to. What is the average monster's to-hit?

My current party contains a couple of characters that like to WWE onto other monsters despite their fragility making it difficult for me to tank in a more traditional fashion and while I'm rather certain that some smarter play would help, I don't want to get grumpy at my teammates for having fun.

For additional context, we recently fought a Werewolf that specifically *refused* to try and swipe at me because of my higher AC (and it was a plot character with Legendary resistances so it stopped my attempts to Compelled Duel it). I'm trying to sort of lock down what I should doing as a Tank and I'm kind of unsure at the moment.

In general, yes the more AC the better. If the enemies are ok with ignoring you than you are at risk of running into a what I call a "lump" situation. I define a lump as a character that is significantly harder to kill than his or her companions, but also doesn't contribute enough offense that enemies feel like they can ignore them. Two main ways to get around this are work on your aggro control and to increase your offense. Sentinel feat is usually a great way to start. As it both encourages enemies to hit you while also increasing your offense if they don't.

To avoid being the lump, I usually favor dueling fighting style over defense fighting style, especially if the character can later pick up the latter. Without knowing the specifics of your character and the rest of your party it's tough to say more, but don't misconstrue incentive to hit you with a tanking. Smart monsters should go after squishy targets. It's your job to dissuade that as mechanically as possible, while also being as unkillable as possible. That means keep increasing your AC as much as you can within reason. Buy plate mail when you can, but consider taking the sentinel feat to make you a bigger threat. If given mutually exclusive options, start with increasing damage.

[Galithar]

I've seen a couple of comments regarding "increasing" returns on AC and I should point out for the record, that I am currently playing a tank.

I've been fortunate enough to find +1 Chain Mail as a Paladin with the Defense Fighting Style: I'm sitting at a Base 20 AC at the moment.

We've fought against some monsters (The Froghemoth) that have +10 to hit already.

I want to get a sense of at what point further investing in AC is going to be a waste of resources for "most situations" because I suspect my DM will arms-race me a bit with monsters if I rush it, which means that the squishies in my party (who are not investing in their AC) are going to be guaranteed hit with basically everything; a situation I don't want.

Until recently, I was under the assumption that 22 was probably a good point to stop, but I don't have an actual way to check that or compare that to. What is the average monster's to-hit?

My current party contains a couple of characters that like to WWE onto other monsters despite their fragility making it difficult for me to tank in a more traditional fashion and while I'm rather certain that some smarter play would help, I don't want to get grumpy at my teammates for having fun.

For additional context, we recently fought a Werewolf that specifically *refused* to try and swipe at me because of my higher AC (and it was a plot character with Legendary resistances so it stopped my attempts to Compelled Duel it). I'm trying to sort of lock down what I should doing as a Tank and I'm kind of unsure at the moment.

If your DM is literally having things not attack you due to your AC you've hit the point of AC increases decreasing in value. A high AC character NEEDS to be attacked or their AC is pointless. You personally should invest in making yourself do more damage so that you're a bigger threat, when you start getting attacked regularly again THEN invest in AC again.

[OverLordOcelot]

First of all, I don't appreciate the personal attacks. I'm trying to have a debate on math, not attacking your character.

I'm not attacking your character, I'm attacking your model. If you're going to take someone pointing out the myriad flaws of your model as a personal attack, that's a problem in your own head.

At the most basic level, your model fails because fights don't last the 55 rounds you used, so saying "wow I could last 55 rounds" is irrelevant. Actual fights that last more than six rounds of melee are vanishingly rare, and the difficulty of post-combat healing is irrelevant in any situation where a TPK-1 could be conisdered a victory. This means that both the 9 rounds the paladin lasts and the 7 rounds the fighter lasts without the ring are sufficient for them to stay alive through any actual combat, so giving it to either one provides no significant benefit for the party if we accept RTD as a metric.


All your points about the cloak of displacement or whatever are unnecessary. Even without that, the math still massively favors the paladin getting the ring. I just used the example to show how bad it could be. Without that it still is about 30 RTD on the paladin.

The model that you claimed showed the ring should go to the paladin included the assumption that the cloak of displacement would have 100% uptime. My points about the cloak of displacement show that your model was badly flawed and incredibly unrealisitc. The fact that you included 100% uptime in your model when 3 of the 4 CR 15 opponents in the SRD have abilities that would routinely switch off the cloak highlights how unrealistic the assumptions you used are.


Bringing up other enemy tactics is entirely off subject. Any enemy that is doing things to get around AC is outside the scope of where to distribute an AC boosting item.

No, bringing up enemy tactics is entirely on subject. An enemy that nullifies your characters defenses makes him easier to kill. If some minions grapple you, you lose the cloak. If they then knock you prone, they get advantage to hit you in melee and you get disadvantage to hit them. All of which completely the alter your damage output and damage taken far, far away from the model.

An actual CR 15 green dragon dueling you in melee will force you to make a DC 19 dex save at the end of your turn or fall prone, negating your cloak by giving it advantage on it's three attacks (four if another party member is still considered to have turns). On rounds where you don't fall prone it can use it's breath weapon to do 16d6 (save for half) that doesn't care about your AC. It's only going to make attacks against your full defenses when it doesn't have breath recharged and you make the dex save. A CR15 green dragon using it's 18 INT and centuries of life experience instead of blindly dueling will figure out if you've even got the ability to fly, and if you don't will ruthlessly strafe you from afar, negating your AC entirely by using its breath weapon every time and loitering out of range while it charges.

Which shows that your idea that you can sit there for 30 or 50 rounds dealing 37.5 DPR against the enemy if you get the ring just doesn't fly.


"AC gets increasingly better for each point you have."

"But what if you get hit with a fireball?"

It just isn't part of the debate.

"My character can fight for 55 rounds against a CR15 enemy if he gets the ring and the enemy only attacks him with straight attack rolls!"

"But all 4 of the CR15 enemies in the SRD have attacks and abilites that aren't just straight attack rolls, and that nullify the key ability that produces the silly 55 round result"

"I declare that actual monsters are not part of the debate, why would any situation you'd actually encounter be relevant to how you'd allocate the ring for best effect?"

A party should base their tactics on stuff that actually happens, not on theoretical constructs that will never occur.


How about this: Show me a scenario or foe where the fighter would be better off taking the ring, as opposed to the paladin?

Show me where I've asserted that the group should give the ring to the fighter? My guess is that the group would be best off giving the ring to someone who is not either of the two front line people, as it doesn't seem to provide any real benefit for either of them. And in practice a group of well-equipped level 15s will probably end up giving it to whoever has the free attunement slot to wear it (rendering the AC analysis moot), because by that point the limit of three attuned items per character is really severe.

Also, challenging me to provie a specific scenario or opponent that leads to a specific result when I don't even know basic information like most of the stats for the two party members you've mentioned, or the stats for the other party memebers, or even how many people are supposed to be in the party you're doing the analysis for is just silly.


Edit: One more point to help put my math into perspective. It’s not that the paladin will last 55 rounds or 30 or whatever. It is that the paladin will last roughly 9 times as long as the fighter with the ring. Without the ring he only lasts about 40% longer. No fight will last that long. It just means the party can handle a lot more enemies before falling when the paladin gets it.

You said the key words: No fight will last that long. That's what means that the extra point of AC on the paladin is irrelevant, the only benefit you've shown from giving it to him is that he'd last longer in some theoretical long fight that won't actually happen. Distributing magic items on the basis that they'll help a fight that won't ever occur does not make sense.

[PeteNutButter]

I'm not attacking your character, I'm attacking your model. If you're going to take someone pointing out the myriad flaws of your model as a personal attack, that's a problem in your own head.


They will see that your math is not modeling anything actually useful, and realize that you're just trying to stockpile items for yourself and not actually distribute them in a way that provides the most benefit to the party.

This is an attack. It's impugning my motives, by implying that I am not doing accurate math, but instead trying to justify hording items. I don't mind being called greedy, but saying my math is bad is the worst. :smalltongue:



At the most basic level, your model fails because fights don't last the 55 rounds you used, so saying "wow I could last 55 rounds" is irrelevant. Actual fights that last more than six rounds of melee are vanishingly rare, and the difficulty of post-combat healing is irrelevant in any situation where a TPK-1 could be conisdered a victory. This means that both the 9 rounds the paladin lasts and the 7 rounds the fighter lasts without the ring are sufficient for them to stay alive through any actual combat, so giving it to either one provides no significant benefit for the party if we accept RTD as a metric.


I feel like I must not explaining this well enough. Let me try one more time.

The DMG shows average stats for monsters by CR. I compare all characters against a monster of their own level CR as a baseline. (You could just as easily compare characters against their level + 3 CR, but the math would be equally worse for them both so that doesn't matter.) For level 15, monsters should have around +8 to hit and do an average of 95.5 damage a round. The math there is simple to figure out roughly how long it'd take that to bring the character down ( PC HP/[monster hit chance * average damage]). You can then compare the two characters on this simplified dimension.

This is only comparing them on relative AC and hit points. That's it. It doesn't say anything about flying enemies or team tactics, but most importantly it says all you need to know about AC.

Now while it is true that a single DMG "blank" monster at CR 15 might take 55 rounds to take down the paladin that isn't the useful information. The useful information is what we can extrapolate from that. Forget RTD, consider it an abstract number, call it durability (against AC attacks).

In practical combat there are going to be all sorts of things going, various monsters, terrain etc. They might be fighting a gajillion orcs, or a pair of red dragons. No matter the enemy, the paladin lasts significantly longer. But it is the paladin's job to get in the face of as many enemies as possible and try to keep them near him. The Fighter typically sticks to the edge, engaging with just one enemy at a time, to try and keep himself from getting splatted. D&D aggro sucks though, so even with sentinel, the paladin can't always keep enemies on him. In these situations, if the fight is deadly, the fighter usually goes down.

Now the paladin is given a choice of healing the fighter or attacking. The fighter does significantly more damage, so he is going to heal the fighter. He can LoH for 1 point all day (ready it, if the enemies go before the fighter to make sure the fighter gets a turn). This is the teamwork aspect of the game. The paladin accepts that the fighter does more damage, so he becomes a heal bot as long as the fighter is down. But the enemies quickly learn that they cannot win the fight without killing the paladin. (In this scenario I'm assuming the rest of the party is ranged and hopefully in less danger, but the paladin may have to peel off to go heal them as well.)

Back to my model. I know you had your criticisms of it, like me assuming enemy disadvantage for the paladin, but note that I was equally generous in my assumptions for the fighter, assuming he had rage up (2/day at 2 barbarian levels) and assuming enemy damage was resisted (physical only), which is not at all a sure thing at level 15. The model told us that the paladin's durability increased by 3-5 times (200-400%) when his AC increased by 1 point, while the fighter only gained about 3-5%. There are a number of reasons why the difference is so huge, such as assumed disadvantage on enemy attacks, and the nature of rage defense being reduction instead of avoidance. My point was that even if we assume worse case scenario, the math doesn't get much different. We are talking about a difference of 100 times. We could fiddle with the numbers a bit, but the magnitude is too big to think they'd even come close to parity, comparing 5% to 200%.

The model tells us who gets the most mathematically out of the ring, combining that with actual table knowledge/common sense is still required. It's a springboard. In this party, the paladin gets the most mathematically and it also makes sense for him to stay alive as he can keep the rest of the team alive.

There are in fact scenarios where the fighter should get the ring. If the DM almost exclusively uses hordes and foes are already having to get 20 to hit the paladin. In that case the AC boost does nothing for the paladin. Or maybe if the DM is particularly brutal and likes to attack PCs while they are at zero.

I think I may actually do up a spreadsheet for this model, where I can incorporate more complex things like saves and melee up time. Really, all my current model does is take the existing AC distribution method (give it to the PC that'll get the most durability from it) and add the damage dimension. This change only skews the math away from giving the AC boosting items to the lump, who does nothing but not die.

[OverLordOcelot]

This is an attack. It's impugning my motives, by implying that I am not doing accurate math, but instead trying to justify hording items. I don't mind being called greedy, but saying my math is bad is the worst. :smalltongue:

This is simply untrue as anyone reading the quoted statement can see. I have not claimed that your math is inaccurate, but rather that model is useless; that it, that the math is accurately spitting out a meaningless number that doesn't correspond to anything useful in game terms.


I feel like I must not explaining this well enough. Let me try one more time.

No, you're explaining fine. The problem is that your model doesn't correspond to reality, and that you simply don't get to tell other people 'you are not allowed to take actual survivability into account, you must accept my simplified model and allocate the treasure according to it because I said so'.


This is only comparing them on relative AC and hit points. That's it. It doesn't say anything about flying enemies or team tactics, but most importantly it says all you need to know about AC.

No, it DOES NOT SAY ALL I NEED TO KNOW ABOUT AC. You don't get to declare "This is all you need to know, shut up and accept my decision", and have anyone accept it. You're not some kind of authority figure who gets to tell people they must accept what you say.


The model tells us who gets the most mathematically out of the ring, combining that with actual table knowledge/common sense is still required. It's a springboard. In this party, the paladin gets the most mathematically and it also makes sense for him to stay alive as he can keep the rest of the team alive.

No. The model tells us that a particular set of equations that you set up that don't have much relation to the actual situation spits out a higher number for the paladin than the fighter. That's it. You declaring ad nauseum that this particular number is the only one that needs to be considered doesn't make it so.

As I pointed out and you ignored, if for the sake of argument we assume that the model shows us a sensible survivability metric, then what your model tells me is that giving the ring to the paladin is a waste. Since he's already so durable against melee attacks that he will last longer than any combat will go, lasting 9 rounds on average, he's not going to go down in any combat. So taking his expected lifetime from 1.5x what it needs to be to 6x what it needs to be provides no actual benefit, as he'll still be up at the end of the combat either way.

[OverLordOcelot]

For additional context, we recently fought a Werewolf that specifically *refused* to try and swipe at me because of my higher AC (and it was a plot character with Legendary resistances so it stopped my attempts to Compelled Duel it). I'm trying to sort of lock down what I should doing as a Tank and I'm kind of unsure at the moment.

As a tank, you need to be enough of a threat that enemies regret ignoring you to focus on non-tank party members. The MMO concept where you rack up defense and spam taunt just isn't there. I don't think there's a simple target number to look for, what I would do is try to balance offense and defense so that enemies don't feel they can safely ignore you. If, like is happening now, enemies are ignoring you to hit other party members, then it's time to ignore defense for a bit. What's actually balanced is going to depend primarily on the DM and to a lesser extent on the rest of the party, contrary to what certain people here think you're not going to get a simple breakpoint number. This is especially true since, while there is an average to hit by Cr listed in the DMG, actual creatures don't follow it very tightly, and higher CR creatures often have a plethora of special maneuvers listed.

As far as the werewolf specifically, legendary resistances are generally enough to shut down one player, like happened to you. To deal with legendary resistance, you need to have multiple party members forcing saves, and players need to not do the weird thing where they land enough 'save or suck' effects for the boss to use up all of its legendary resistances, then stop casting save or suck spells. (Seriously, I've seen that so many times). You can land some of your particular spells if you get another player to join you in spamming low level spells that the monster won't like - dissonant whispers is a great one that often gets overlooked (1st level, on a failed save the enemy has to run away from the caster and provoke AOOs).

Also, wrathful smite is a significantly better spell for that situation than compelled duel. Compelled duel means your party can't attack the werewolf or the spell ends, and it can attack them at disadvantage, so you have to bring it down yourself. Wrathful smite gives the werewolf disadvantage attacking anyone, including you, and allows other party members to attack just fine. Both spells require the enemy to fail a wisdom save and for you to maintain concentration while the enemy is hitting you, but one gives him disadvantage while doing so. Smite does require an attack roll to deliver, but it 'sticks' to your attacks until you get a hit, it's not lost on a miss. Compelled duel IMO is for taking out 'lieutenant' types or for bosses who have a lot of minions (so you take the boss and the rest of the party deals with the rest), it's not really good if there's only one enemy.

[malachi]

Wouldn't the correct analysis here, rather than strict RTD, be to look at A) how much resource expenditure is prevented by the ring, and B) whether giving the ring to any character would prevent them from dying in combat?

(A) determines how much out-of-combat healing and/or concentration spells are saved by the ring, and (B) determines how much in-combat healing is saved.

Thus, from this perspective, never give the ring to someone that is off the scales of bounded accuracy (either only missed on a 1, or only hit on a 20). If we discount constant advantage / disadvantage, the simplest answer is to give the ring to the character who gets attacked (note: I didn't say hit, since the ring will make 1/20 attacks miss) the most and/or dies the most. Complicating this analysis is whether a certain character needs to keep a concentration spell up to limit party resource-expenditure, yet is losing that concentration due to AC-based attacks.


Doing the math shows me that +1 AC always causes 1 attack in 20 to change from a hit to a miss for attacks without Advantage / Disadvantage (assuming the attack originally needed somewhere between a 2 and a 19 to hit).

For attacks with Advantage, +1 AC causes 1.1 attacks in 20 to change from a hit to a miss if the attacker needs an 11 to hit (to a maximum of 1.9 attacks in 20 changed to a miss when the attacker needs a 19 to hit).

For attacks with Disadvantage, +1 causes 1.1 attacks in 20 to change from a hit to a miss if the attacker needs a 10 to hit (to a maximum of 1.9 attacks in 20 changed to a miss when the attacker needs a 2 to hit).

[PeteNutButter]

No, you're explaining fine. The problem is that your model doesn't correspond to reality, and that you simply don't get to tell other people 'you are not allowed to take actual survivability into account, you must accept my simplified model and allocate the treasure according to it because I said so'.

The problem is you are trying to make my model represent all of D&D, when it doesn't do that. It just measures AC and damage and allocates accordingly. Never did I say that it is the gospel of D&D. I very clearly said it's a springboard.



No, it DOES NOT SAY ALL I NEED TO KNOW ABOUT AC. You don't get to declare "This is all you need to know, shut up and accept my decision", and have anyone accept it. You're not some kind of authority figure who gets to tell people they must accept what you say.

Such hostility. Chill out man. I am not making an appeal to authority which I never claimed to have. I am merely asserting a fact. All you need to know about AC is how much damage it mitigates. You are welcome to disagree for some reason, but I'm not sure how or why.



No. The model tells us that a particular set of equations that you set up that don't have much relation to the actual situation spits out a higher number for the paladin than the fighter. That's it. You declaring ad nauseum that this particular number is the only one that needs to be considered doesn't make it so.

A particular set of equations? It's essentially effective hp, I've just added in a damage concern which only helps the case for non-tank characters getting the item.

At any rate, you declaring that my declarations just aren't so is not an argument. All of your attempts to refute my points in game terms are you describing things that are outside the scope of AC.



As I pointed out and you ignored, if for the sake of argument we assume that the model shows us a sensible survivability metric, then what your model tells me is that giving the ring to the paladin is a waste. Since he's already so durable against melee attacks that he will last longer than any combat will go, lasting 9 rounds on average, he's not going to go down in any combat. So taking his expected lifetime from 1.5x what it needs to be to 6x what it needs to be provides no actual benefit, as he'll still be up at the end of the combat either way.

At this point it is clear you're just not getting it. The paladin living longer draws those attacks from the team. It's not like everyone is being attacked equally at all times. If that were the case, the wizard goes down first in every fight, and defensive characters would be completely pointless. The paladin takes the hits (or mostly misses) so the fighter doesn't have to. If a tank cannot control any aggro (which everyone can do a little just via opportunity attacks) then he isn't a tank.

If the tank can live 6 times as long as the damage dealer, than the tank can get in the face of 6 enemies, while the damage dealer chops at the 7th on the edge. The tank isn't living 6 times as long as the damage dealer, he is taking 6 times the heat. It's basic strategy. As long as the tank is taking more attacks than the others, the party is doing well. How well all this works is terrain and DM dependent or with a good DM, monster dependent. It works swimmingly with things like the ancestral guardian barb and cavalier.

[Azgeroth]

+AC items should go to who-ever the healer/rezer is first, then the 'tank' or whoever is getting hit the most..

edit : and i dont mean who is attacked the most, i mean who is HIT the most.

ill accept that going from 11-12 ac is less of a gain mathematically than going from 18-19.. but the character with an ac of 11, really needs help with their AC more than the char with an 18..

[OverLordOcelot]

The problem is you are trying to make my model represent all of D&D, when it doesn't do that. It just measures AC and damage and allocates accordingly. Never did I say that it is the gospel of D&D. I very clearly said it's a springboard.

No, I am not expecting the model to represent all of D&D. I do, however, expect that a model that claims to give an answer for where to allocate a magic item provide a good answer for where to allocate that magic item. Yours fails.


Such hostility. Chill out man. I am not making an appeal to authority which I never claimed to have. I am merely asserting a fact. All you need to know about AC is how much damage it mitigates. You are welcome to disagree for some reason, but I'm not sure how or why.

I responded in bold text to your bold text. If that level of emphasis is 'hostility', and I need to 'chill out man', then take your own advice - which you didn't, because you went right on to use the same bold 'hostile' text immediately after the 'chill out' nonsense. The fact that you asserted something does not make it a fact, no matter how often you use bold text then complain if someone uses the same formatting in response. And asserting that something is a fact when it is clearly not even true is most definitely attempting to exert an authority that you lack.


A particular set of equations? It's essentially effective hp, I've just added in a damage concern which only helps the case for non-tank characters getting the item.

Yes, your effective HP equations are just a particular set of equations, that don't model D&D combat well at all, and do not provide any reasonable answer for how to distribute the item.


At any rate, you declaring that my declarations just aren't so is not an argument. All of your attempts to refute my points in game terms are you describing things that are outside the scope of AC.

I have repeatedly provided arguments. The fact that you ignore them and continue to make false declarations and use 'hostile' bolded text doesn't change that.

[Chronos]

A model that assumes that enemies won't use tactics will give good answers, if we're playing a game where enemies won't use tactics. If we're playing a game where enemies do use tactics, then it won't give good answers.

Think about what the players themselves would do. If you're playing D&D, and your enemies include a heavily-armored warrior and a spellcaster, which one do you attack first? In all the groups I've been in, everyone tries to take out the enemy spellcaster first. Alternately, if we can't tell the difference between the enemies, but one is already hurt, everyone will try to finish off that one first. Why wouldn't the enemies plan the same way?

And it's not just CR 15 monsters that have ways of going around the tank's high AC. All enemies have ways of going around the tank's high AC. All you have to do is go after someone else's AC instead.

Sure, some monsters, like most animals, will just go after whichever target is closest. Tanking works for them. But that's not most monsters. Most are smarter than that.

[OverLordOcelot]

Think about what the players themselves would do. If you're playing D&D, and your enemies include a heavily-armored warrior and a spellcaster, which one do you attack first? In all the groups I've been in, everyone tries to take out the enemy spellcaster first. Alternately, if we can't tell the difference between the enemies, but one is already hurt, everyone will try to finish off that one first. Why wouldn't the enemies plan the same way?

Or if you do need to kill the heavily armored warrior, you negate his armor instead of blindly smacking at him with regular attacks. Grapple him and knock him prone to negate his cloak of displacement, give your melee attacks advantage, and let everyone but the people grappling walk away from him. Hit him with a CC spell so that you can ignore him and finish off his allies first, or at least debuffs like slow so he can't sentinel you when you kill allies. Focus you damaging abilities that don't go against AC on him to kill him. Yeah, someone can respond "Oh, but that's not an attack against AC so it doesn't count when distributing an AC item" - but once the non-AC abilities have the tank out of commission, the opponents are then free to make regular attacks against the party member that the model says will never be attacked. And none of this requires the DM to tailor the encounter to take out this one character specifically!



To get an idea of how divorced from 'reality' the model is, it says that if the group faced half a dozen green dragons (CR 15 opponent from the SRD), they would do melee attacks against the paladin because the paladin would use the sentinel feat and grappling to tank the dragons, and that no attacks would go against the rest of the group. And that "If the tank can live 6 times as long as the damage dealer, than the tank can get in the face of 6 enemies, while the damage dealer chops at the 7th on the edge." If use the math from the model, giving him the ring means it will take him 55 rounds to die against a CR 15 opponent, which converts to better than 9 rounds against 6. That's pretty safe, I think our party can kill enemies in the 9 rounds before he dies. Obviously we should give the guy who can last longest the ring if enemies always attack him and he shrugs off their attacks!

But if you try to put that on a map, it's hard to just get all six dragons attacking the paladin, because they're huge creatures and crowd each other out. They'd have to work just to be able to focus melee attacks in the first place! Sentinel can only hit one of them a round, so it doesn't seem likely to bother the other five. Grappling is out since they're huge, two sizes larger than a human (and impractical for an outnumbered person). It's also not really believable that even unintelligent enemies would work that hard to just attack one guy, and these are long lived dragons with 18 INT who can probably figure that working hard to attack the worst possible opponent is a bad idea. So the model already seems to have problems.

However, to make the model work lets pretend the paladin is carrying the Plot Item of Must Attackness, and the dragons must kill him before attacking anyone else. And they'll work around any positioning issues. This way he gets to soak the damage from 6 CR 15 enemies while damage dealers attack on the periphery. So the half a dozen green dragons start, as even young and dumb dragons are wont to do, with their breath weapons. If we assume the paladin has a 16 CON and 20 CHA, that results in an average of 252 poison damage. Even if he gets lucky and makes all 6 saves, he's still taking 168 damage. He's now got a slight problem, in that a level 15 paladin with a 16 con has only 139 hit points. If we assume he is a dwarf (resistant to poison) he now has 13 HP left. Unless he goes immediately after the dragons, they will all use their legendary action to do wing buffets at the end of whoever is next's turn. Fortunately, he only needs an 11 to make the DC 18 dex saves if we assume a 14 dex, unfortunately he needs to make 6 of them at once, as a single failure finishes off his last 13 HP.

So what we're told about the model is "the tank can get in the face of 6 enemies, while the damage dealer chops at the 7th on the edge", "As long as the tank is taking more attacks than the others, the party is doing well," and if we apply the numbers from the model we see that he should last a staggering 9 rounds tanking 6 CR15 enemies. But if we put those claims to the test and see what happens if he tries it against six of the first CR15 creature in the SRD, we first see that he's unlikely to be able to force them all to attack him and avoid the rest of the party. And if we handwave them into doing exactly what he wants, the only way he survives the first round of enemy attacks is if he happens to be a particular race, and even then he's not even likely to make it to his next round. Instead of laughing and shrugging off enemy attacks while healing the party, he's going to either be dead or soaking up an absolutely huge amount of healing.

[MaxWilson]

A model that assumes that enemies won't use tactics will give good answers, if we're playing a game where enemies won't use tactics. If we're playing a game where enemies do use tactics, then it won't give good answers.

Think about what the players themselves would do. If you're playing D&D, and your enemies include a heavily-armored warrior and a spellcaster, which one do you attack first? In all the groups I've been in, everyone tries to take out the enemy spellcaster first. Alternately, if we can't tell the difference between the enemies, but one is already hurt, everyone will try to finish off that one first. Why wouldn't the enemies plan the same way?

And it's not just CR 15 monsters that have ways of going around the tank's high AC. All enemies have ways of going around the tank's high AC. All you have to do is go after someone else's AC instead.

Sure, some monsters, like most animals, will just go after whichever target is closest. Tanking works for them. But that's not most monsters. Most are smarter than that.

There's also counterplay to consider: if the heavily-armored warrior is first to kick in your door, and the spellcaster is 200' away on top of a 30' tall stone monolith blasting away with Spell Sniper Agonizing Repelling Blast, then both of your choices are bad. As a player you ideally want a situation where all of an enemy's choices are bad. You can settle for a situation where all of the choices they are likely actually make are wrong--e.g. when fighting zombies, it might be okay if you have no plan for what to do if the zombies scatter and regroup later to harass you with ambush tactics, because zombies usually don't do that. But it's even better to have a plan that would work even if these zombies turn out to be tactical geniuses.

Because PCs have a wide range of capabilities and are usually in the position of initiating action declarations (due to how the typical dungeon crawl is structured), they have a lot of room to stack the deck in their favor by choosing builds that synergize with each other, choosing or creating favorable terrain (with caltrops, Mold Earth, etc.), choosing formations intelligently (keeping PCs widely-dispersed but within mutual support range, like an army squad in real life, allows more control over which PCs take the brunt of melee monster attacks except when you're horrifically outnumbered and being attacked from all directions; taking advantage of cover allows more control over which PCs take the brunt of ranged attacks), preparing spells for contingencies (like Dimension Door for when you're horrifically outnumbered and being attacked from all sides), pre-casting certain spells (like Pass Without Trace, Conjure Animals), etc.

Net effect is that against most monsters from the MM or published WotC products, PCs have a great deal of ability to choose the terms of engagement and force monsters to attack strong PCs instead of weak ones. Monsters typically do not have the opposite ability, if the players play intelligently. This is a deliberate aspect of the game's design.

[OverLordOcelot]

Net effect is that against most monsters from the MM or published WotC products, PCs have a great deal of ability to choose the terms of engagement and force monsters to attack strong PCs instead of weak ones. Monsters typically do not have the opposite ability, if the players play intelligently. This is a deliberate aspect of the game's design.

To use the example that's come up in this thread, do you think a party of level 15s can force six CR15 enemies (I'll look at the four in the SRD) to attack a particular high armor PC instead of weaker characters (drop it to four if you think six is overkill, but six was the subject of a specific claim). This has to take place in terrain that they'd reasonably fight in, and the scenario has to be one that would qualify as a challenge; if you're fighting purple worms where the party can just hover out of reach then you get no XP and it doesn't count as any kind of achievement. For Bronze and Green dragons, you've got to deal with flight, aoe stuns, and legendary saves, plus they are not grappleable by size M creatures. Mummy lords have a variety of spells including dispel magic, and the ability to go immune to grapples and attacks and move 60'. Purple worms tunnel through any cave walls you try to use to channel their movement.

And then once you do that, how do you force them to use regular attacks that have a hard time hitting instead of their variety of abilities? Only the purple worms are forced to use regular attack rolls, all of the others have lots of abilities that get by the stacked AC. Breath and buffet for dragons, and harm and other spells for mummy lord don't have to hit 'AC 28 with disadvantage'. Even the purple worms can go for a grapple instead of attack once they get tired of missing!

Players definitely have some ability to channel enemy attacks, especially less fantastic or lower level monsters, but the idea being floated here that high level characters can easily force high level monsters to focus on one 'unhittable' character (and therefore it's always best to just stack AC on the one guy) doesn't hold water.

[Ninevehn]

Maybe y'all ought to build a specific scenario and test it with the Ring of Protection on the best AC and then on the worst AC. Just to establish a ground floor. To be honest, if, as OverLordOcelot is positing, the CR15's are just gonna ignore the Paladin's AC whether or not he has the ring, I'm not sure why'd they be attacking the Wizard's (or whomever's) AC either.

[OverLordOcelot]

Maybe y'all ought to build a specific scenario and test it with the Ring of Protection on the best AC and then on the worst AC. Just to establish a ground floor. To be honest, if, as OverLordOcelot is positing, the CR15's are just gonna ignore the Paladin's AC whether or not he has the ring, I'm not sure why'd they be attacking the Wizard's (or whomever's) AC either.

I don't see what you'd establish as a ground floor by making up a scenario. Testing a scenario that a model shows should be easy and finding that it's more of a TPK can highlight problems with a model, but saying 'well the ring would be better here in this particular case' doesn't really provide much information. This is especially true since any scenario is going to depend on how the DM would handle the encounter, which varies from DM to DM and encounter to encounter, and whatever you post will lead to accusations of you biasing for the players and/or against the players, of using metagame knowledge and/or of playing the enemy too dumb, of making assumptions that are too generous and/or strict, and often all of these for the same scenario.

I'd also probably use a more balanced encounter - according to the DMG, 6 green dragons is 4x the xp value that should provide a deadly encounter for 6 level 15 charaters, though according to Xanathar's chart it should be challenging but winnable. I would not really expect those PCs to be able to beat the dragons unless they were able to do specific prep beforehand like putting up resistance spells against poison, it's pretty easy to hit a bunch of people with those breath weapons, and 6 of them is extremely deadly. I only used 6x CR 15s for the encounter because the person specifically bragged about how easily his character could tank 6x CR 15s based on his model.

As far as why they'd use AC based attacks, for the 4 example CR15s it's pretty simple. The purple worm is stuck with AC based attacks, other than trying to grapple and knock prone to get advantage. The dragons have a breath weapon that recharges on a 5-6, so they only have their big AC ignoring attack available about 1/3 of the time, and need to use AC attacks the rest of the time, again unless they feel the need to grapple/prone. Mummy lords have the harm spell, but don't have unlimited castings of it.

[Ninevehn]

Well, you know, keep all that in mind when talking about how they can work around the Paladin's AC. Purple Worms have to attack AC to do damage, dragons might not have all their breath weapons available, mummy lords don't always have Harm available. I just don't see any of this stuff cutting one way but not the other.

[malachi]

+AC items should go to who-ever the healer/rezer is first, then the 'tank' or whoever is getting hit the most..

edit : and i dont mean who is attacked the most, i mean who is HIT the most.

ill accept that going from 11-12 ac is less of a gain mathematically than going from 18-19.. but the character with an ac of 11, really needs help with their AC more than the char with an 18..

I want to dispute this reasoning (not picking you out in particular, you're just the person I randomly chose to quote on this line of thought). A lot of people are looking at this through the framework of "how many hits can I take before I die", but I think it's important to realize that the typical combat only lasts so many rounds, and thus only so many attacks are dealt by the enemies. As long as your AC is in normal ranges (enemy needs somewhere between a 2 and 19 to hit), getting +1 AC means that 1/20 attacks changes from a hit to a miss. Thus, it lowers the amount of damage dealt by each attack by 5% of the damage roll.

Example:
Albert has AC 21, and is getting attacked by an enemy with +2 to hit, and deals an average of 10 damage per attack (15 on a crit). The enemy needs to roll a 19 or 20 to hit, so deals an average of 1.25 damage per attack to Albert. With the ring, the enemy needs a 20 to hit, so deals 0.75 damage per attack to Albert. The ring reduced the average damage per attack taken by Albert by 0.5 damage.

Bob has AC 4, and is getting attacked by the same enemy. The enemy needs a 2+ to hit, so it deals 9.75 damage per attack, or 9.25 with the ring. The ring reduces the average damage taken by Bob per attack by 0.5 damage.

In this example, the ring stops 0.5 damage per attack, despite the 17 point difference in AC between the two characters. Thus, the limit the number of resources expended to fully heal up, the only important factor is to decide who is going to get hit more often, as then the ring will prevent more damage and thus require fewer resources to heal up.

My previous post gets into where the breakpoints are of who to give the ring to depending on characters who typically grant Advantage / Disadvantage to attackers (assuming that characters get attacked the same amount).

[PhantomSoul]

I want to dispute this reasoning (not picking you out in particular, you're just the person I randomly chose to quote on this line of thought). A lot of people are looking at this through the framework of "how many hits can I take before I die", but I think it's important to realize that the typical combat only lasts so many rounds, and thus only so many attacks are dealt by the enemies. As long as your AC is in normal ranges (enemy needs somewhere between a 2 and 19 to hit), getting +1 AC means that 1/20 attacks changes from a hit to a miss. Thus, it lowers the amount of damage dealt by each attack by 5% of the damage roll.


Agreed, but with one slight change: it's minimally about 5.26% (crit fails never hit)... but it's normally going to be better than that. To use a relatively extreme case, if the attacker can only hit on a 17, 18, 19 or 20 and you increase your AC by 1, you're actually reducing the damage by about 22% (using 1d8+4 as the damage for the toy value to have a damage estimate on a crit).

[OverLordOcelot]

And then once you do that, how do you force them to use regular attacks that have a hard time hitting instead of their variety of abilities? Only the purple worms are forced to use regular attack rolls, all of the others have lots of abilities that get by the stacked AC.

Well, you know, keep all that in mind when talking about how they can work around the Paladin's AC. Purple Worms have to attack AC to do damage, dragons might not have all their breath weapons available, mummy lords don't always have Harm available. I just don't see any of this stuff cutting one way but not the other.

Well, you know, I already explicitly mentioned it. And I don't bother to mention 'water is wet' stuff like 'The DM can arbitrarily change any or all abilities that a creature has at any time for any reason', because that's the way RPGs work. However, I do think that if someone claims 'this character could tank six CR15 enemies at once', one should assume that they're tanking CR15 enemies as listed in RAW, and not CR15 enemies that have their iconic abilities arbitrarily disabled. The simple fact is that if the paladin tries to tank 6x CR15 enemies the way the model said he could and the poster bragged he would, 3/4 of the actual CR15 opponents simply squash him right off the bat if they use their RAW abilities.

[malachi]

Agreed, but with one slight change: it's minimally about 5.26% (crit fails never hit)... but it's normally going to be better than that. To use a relatively extreme case, if the attacker can only hit on a 17, 18, 19 or 20 and you increase your AC by 1, you're actually reducing the damage by about 22% (using 1d8+4 as the damage for the toy value to have a damage estimate on a crit).

We're looking at this in two different ways. I'm going to agree that yes, increasing high AC reduces the damage taken by a higher percentage. I do not dispute this fact, but merely whether that perspective is relevant.

If an enemy hits on Albert on a 17, 18, 19, or 20 and you give him a ring of protection, the enemy now hits on an 18, 19, or 20. 1 out of 20 attacks (when a 17 is rolled) will have reduced damage (1d8+4 less damage taken). Every other roll will deal exactly the same amount of damage as it would have without the ring.

If an enemy hits Bob on a 2 - 20 and you give him a ring of protection, the enemy now hits on a 3 - 20. 1 out of 20 attacks (when a 3 is rolled) will have reduced damage (1d8+4 less damage taken). Every other roll will deal exactly the same amount of damage as it would have without the ring.

When assigning who in the party gets the ring, the main goal is to optimize party survival and the minimize resource expenditure, correct? It takes the same amount of resources to heal 1d8+4 damage that Albert has taken as it does to heal Bob the same amount of damage (exceptions being characters who generate temp HP, lvl 20 champions, wild shape / polymorph, self-healing effects, etc). Therefore, if the ring prevents 1d8+4 / 20 damage per attack, doesn't it reduce resource expenditure the most to give the ring to the character who receives the most attacks?

Albert: lvl 1 fighter, AC 18, HP 13.
Bob: lvl 1 wizard, AC 13, HP 8.

Enemy has +3 to hit, deals 1d8+4 (8.5) damage on a hit, and 2d8+4 (13) damage on a crit. Accounting for misses, hits, and crits, Albert will have an average of 10.23 health left (took 2.78 damage) after one attack, while Bob will have an average of 3.1 health left (took 4.9 damage).

Giving each of them a ring of protection leaves Albert with 10.65 health left (2.35 dmg) and Bob with 3.54 health left (0.425 dmg).
Albert took 0.425 less damage from the attack, taking 15.32% less damage, and has 4% more health left.
Bob took 0.425 less damage from the attack, taking 8.67% less damage, and has 12% more health left.

If I tell you that Bob is going to get attacked five times over the adventuring day, while Albert is going to get attacked three times over the same period, is it meaningful to ask if whether you'll spend more resources to heal the party if you give the ring to Albert (who would take 3 attacks at -15.32% damage each) or to Bob (who would take 5 attacks at -8.67% damage each)?
Or is it more meaningful to ask whether you'll spend more resources to heal the party if you give the ring to Albert (who would take 3 attacks at -0.425 damage each) or to Bob (who would take 5 attacks at -0.425 damage each)?

Granted, the wizard getting attacked more than the fighter is indicative of a silly situation or bad party coordination, but I used that as an example to show that the percentages don't seem like a helpful comparison to me.

Honest question, given how widespread the "increasing AC gives larger returns the higher your AC is" thought is on these forums, is there some reason that I am missing when asking who in the party should get a ring for +1 AC?

[OverLordOcelot]

I want to dispute this reasoning (not picking you out in particular, you're just the person I randomly chose to quote on this line of thought). A lot of people are looking at this through the framework of "how many hits can I take before I die", but I think it's important to realize that the typical combat only lasts so many rounds, and thus only so many attacks are dealt by the enemies. As long as your AC is in normal ranges (enemy needs somewhere between a 2 and 19 to hit), getting +1 AC means that 1/20 attacks changes from a hit to a miss. Thus, it lowers the amount of damage dealt by each attack by 5% of the damage roll.

No, it really, really doesn't. Attacks and (most) healing occur in discrete chunks in D&D, and there catastrophic effects of hits beyond HP attrition, which can be worse for some characters than others. Modeling hits by treating them as a slow, steady stream based on average damage misses all of that. If a monster

In the CR15 example, the purple worms get a swallow attack anytime they hit, which forces a dex save, if the character fails they are restrained, blinded, and take unavoidable damage each round.

For a simple example, let's say there's two characters Sam with 50 HP and an AC of 18 and Tom with 75HP and an AC of 22, both of whom are worried about fire giants, who take 2 swings per round at +11 to hit and do 28 damage per hit. We determine that Sam takes an average of 19.6 damage per round and will expect to survive 3 rounds, while Tom takes an average of 14 damage per round and will expect to survive 6 rounds. This appears to put both of them in a pretty safe zone, since the party should be able to burn the giant on Sam before it gets 3 rounds of attacks and kills him, while Tom can tank his for a while. If we examine giving a ring to each one, we would drop Sam's average damage to 18.2 and Tom's average damage to 12.6. So it appears to be a wash as far as average damage, and we'd give the ring to Tom since we expect to focus-fire and kill the giant on Sam first, so Tom will soak more rounds saving more chunks of that 1.4 out of combat healing.

But looking just at average damage misses that if a giant hits Sam twice in a round, he's going down and will probably lose an action, plus require either a major burst of healing or be effectively out of the fight (if someone just tosses a healing word to get him on his feet, he probably needs to go full defense or run away). The giant will land both hits in a round 49% of the time if Sam doesn't have the ring, or 42% if we give it to him. There isn't a chance of catastrophe for Tom either way. And what if there are concentration spells involved? Well, they would both have to make a DC 13 concentration save for each hit, but also Sam will automatically lose concentration on the spell if he takes two hits in a round. So every time Sam gets his, there's a 50% chance of him losing concentration (assuming +2 con bonus and no war caster/resilient), and he automatically loses concentration if he goes down. The average damage calculation doesn't capture that at all, and in fact might lead you to think that all concentration checks would be DC 10.

If we put that math into a scenario like "Party of four sees four hostile Fire Giants. Sam casts hypnotic pattern and hypnotizes two of them, Tom charges in and engages one. Now the loose giant charges Sam, and the party tries to kill/incapacitate the Sam giant, then the Tom giant before the hypnotized giants wake up.", we see that we REALLY want the ring on Sam and not Tom. If we put it on Tom we will expect to need 10 less out of combat healing, but if we put it on Sam we reduce the chance of having to fight twice as many giants at once, which will almost certainly do more than 10 additional combat damage.

This is why building a model that gives a good answer on who benefits the most is so difficult that I don't expect anyone to do so - it's easy to come up with a way to find some kind of average hit points taken based on attacks vs AC, and then declare that boosting the AC of the highest AC party member is better. But there are multiple abilities in play with effects that aren't modeled with average HP, which tend to be more important than 1.4 HP/round differences and which depend a LOT on specific scenarios, parties, and playstyles.

[MaxWilson]

I want to dispute this reasoning (not picking you out in particular, you're just the person I randomly chose to quote on this line of thought). A lot of people are looking at this through the framework of "how many hits can I take before I die", but I think it's important to realize that the typical combat only lasts so many rounds, and thus only so many attacks are dealt by the enemies.

What matters is damage ratios, not the number of rounds a combat lasts, and the number of rounds is not fixed. Against a given set of monsters, if party A takes half as much damage as party B every round while inflicting 80% as much damage, it doesn't matter that party A will take 25% longer to finish any given combat--A will come out of each combat in better shape than party B.

[PeteNutButter]

Well, you know, I already explicitly mentioned it. And I don't bother to mention 'water is wet' stuff like 'The DM can arbitrarily change any or all abilities that a creature has at any time for any reason', because that's the way RPGs work. However, I do think that if someone claims 'this character could tank six CR15 enemies at once', one should assume that they're tanking CR15 enemies as listed in RAW, and not CR15 enemies that have their iconic abilities arbitrarily disabled. The simple fact is that if the paladin tries to tank 6x CR15 enemies the way the model said he could and the poster bragged he would, 3/4 of the actual CR15 opponents simply squash him right off the bat if they use their RAW abilities.

Repeatedly accusing me of bragging is both false and petty. I was merely comparing two separate characters (both of which I planned) and determining by my methods that the ring should go to character A. I never even claimed that the character could tank 6 CR 15s. I merely claimed that character A could tank 6 times what character B could tank when compared against an average DMG CR 15.

Ironically, and slightly off topic, the character in question would be at a standstill against flying green dragons. He has a periapt or proof against poison. I was awarded any rare item, but attuned out. It seemed the best item available. The dragons would be forced to engage in melee or to just leave (which the character couldn’t stop). If the dragon(s) did engage in melee they’d be a fatal fight. Their blindsight negates any common way to give them disadvantage, so the character would probably use their first bonus action to cast SoF to bring the 27 AC to 29 (+11 to con saves keeps it up automatically against anything but a massive crit). That means the dragons full stack does an average of 8.1 damage a round. At that rate the characters 150 hp is depleted in 18.5 Dragon rounds, or on the fourth dragon turn if there are six of them. Well probably the third depending on their legendary actions. Even if I used every resource on heals and smites, I’d end up as dragon lunch.

See that, that’s bragging about the character.

Edit: Forgot about the wing buffet. With the almost inevitable advantage they’d have 6 dragons would only take 9.8 dragon rounds so the character would get maybe 2 turns depending on initiative. Hey that’s not too shabby.

We're looking at this in two different ways. I'm going to agree that yes, increasing high AC reduces the damage taken by a higher percentage. I do not dispute this fact, but merely whether that perspective is relevant.

If an enemy hits on Albert on a 17, 18, 19, or 20 and you give him a ring of protection, the enemy now hits on an 18, 19, or 20. 1 out of 20 attacks (when a 17 is rolled) will have reduced damage (1d8+4 less damage taken). Every other roll will deal exactly the same amount of damage as it would have without the ring.

If an enemy hits Bob on a 2 - 20 and you give him a ring of protection, the enemy now hits on a 3 - 20. 1 out of 20 attacks (when a 3 is rolled) will have reduced damage (1d8+4 less damage taken). Every other roll will deal exactly the same amount of damage as it would have without the ring.

When assigning who in the party gets the ring, the main goal is to optimize party survival and the minimize resource expenditure, correct? It takes the same amount of resources to heal 1d8+4 damage that Albert has taken as it does to heal Bob the same amount of damage (exceptions being characters who generate temp HP, lvl 20 champions, wild shape / polymorph, self-healing effects, etc). Therefore, if the ring prevents 1d8+4 / 20 damage per attack, doesn't it reduce resource expenditure the most to give the ring to the character who receives the most attacks?

Albert: lvl 1 fighter, AC 18, HP 13.
Bob: lvl 1 wizard, AC 13, HP 8.

Enemy has +3 to hit, deals 1d8+4 (8.5) damage on a hit, and 2d8+4 (13) damage on a crit. Accounting for misses, hits, and crits, Albert will have an average of 10.23 health left (took 2.78 damage) after one attack, while Bob will have an average of 3.1 health left (took 4.9 damage).

Giving each of them a ring of protection leaves Albert with 10.65 health left (2.35 dmg) and Bob with 3.54 health left (0.425 dmg).
Albert took 0.425 less damage from the attack, taking 15.32% less damage, and has 4% more health left.
Bob took 0.425 less damage from the attack, taking 8.67% less damage, and has 12% more health left.

If I tell you that Bob is going to get attacked five times over the adventuring day, while Albert is going to get attacked three times over the same period, is it meaningful to ask if whether you'll spend more resources to heal the party if you give the ring to Albert (who would take 3 attacks at -15.32% damage each) or to Bob (who would take 5 attacks at -8.67% damage each)?
Or is it more meaningful to ask whether you'll spend more resources to heal the party if you give the ring to Albert (who would take 3 attacks at -0.425 damage each) or to Bob (who would take 5 attacks at -0.425 damage each)?

Granted, the wizard getting attacked more than the fighter is indicative of a silly situation or bad party coordination, but I used that as an example to show that the percentages don't seem like a helpful comparison to me.

Honest question, given how widespread the "increasing AC gives larger returns the higher your AC is" thought is on these forums, is there some reason that I am missing when asking who in the party should get a ring for +1 AC?

I liked your other model and it was good math showing that the ring only prevents .5 damage either way. Increasing AC gives larger returns as an accepted model is because that .5 damage is a larger portion of overall damage on the AC character, thus increasing his effective health by a much larger percentage.

Now as you also point out I think in actual play it has more to do with roles. Characters designed to withstand damage tend to go out of their way to make or encourage enemies to attack them, with both mechanics and/or with simple tactics.

[MaxWilson]

To use the example that's come up in this thread, do you think a party of level 15s can force six CR15 enemies (I'll look at the four in the SRD) to attack a particular high armor PC instead of weaker characters (drop it to four if you think six is overkill, but six was the subject of a specific claim). This has to take place in terrain that they'd reasonably fight in, and the scenario has to be one that would qualify as a challenge; if you're fighting purple worms where the party can just hover out of reach then you get no XP and it doesn't count as any kind of achievement. For Bronze and Green dragons, you've got to deal with flight, aoe stuns, and legendary saves, plus they are not grappleable by size M creatures. Mummy lords have a variety of spells including dispel magic, and the ability to go immune to grapples and attacks and move 60'. Purple worms tunnel through any cave walls you try to use to channel their movement.

And then once you do that, how do you force them to use regular attacks that have a hard time hitting instead of their variety of abilities? Only the purple worms are forced to use regular attack rolls, all of the others have lots of abilities that get by the stacked AC. Breath and buffet for dragons, and harm and other spells for mummy lord don't have to hit 'AC 28 with disadvantage'. Even the purple worms can go for a grapple instead of attack once they get tired of missing!

Players definitely have some ability to channel enemy attacks, especially less fantastic or lower level monsters, but the idea being floated here that high level characters can easily force high level monsters to focus on one 'unhittable' character (and therefore it's always best to just stack AC on the one guy) doesn't hold water.

On the one hand, let's note that a party of six CR 15s is a Deadly x6 encounter for a party of four level 15 PCs. Even just four green dragons is Deadly x4. As a player and a DM I love high-stakes combats like that--lots of dramatic tension--but I think I am in the minority, and even for me I actually would prefer a little bit more variety than just a bunch of high-level monsters in a brute squad.

On the other hand, I agree with your basic point that high-CR enemies are probably not going to be restricted to attacking purely AC. If you review the thread you'll see I am on record as agreeing with this point.

On the gripping hand, PCs have the upper hand against most monsters, not just because of AC but because they have so many options for controlling terms of engagement. It depends on the details of the party but for example, against six (vanilla, non-spellcasting) Adult Green Dragons, well, Heroes' Feast alone is going to absolutely wreck their game plan. Pass Without Trace is likely to grant grant surprise rounds; a Simulacrum of the fighter will add to the party's DPR and let them exploit those surprise rounds better. Planar Bound Air Elementals and/or a skeleton army can engage individual dragons; Wall of Force + Forcecage can let the wizard take two dragons out of the fight almost unassisted. Even the Paladin can get into the game (despite his likely melee specialization) with his Find Greater Steed Pegasus.

The main tool the PCs have for forcing the dragons to concentrate on the tanks (probably summoned creatures in this case, plus the hypothetical Paladin) is in this case the superior ranged weaponry of the PCs (Sharpshooter Fighter or Spell Sniper Agonizing Eldritch Spear Warlock or similar, plus maybe a Simulacrum of same) which lets them maintain a dispersed formation at long range and retain the strategic initiative. They may use illusions or invisibility to force further tactical mistakes on the part of the dragons, but without the ranged superiority, the PCs would be restricted to hoping that the dragons choose to take the bait. Ranged superiority lets them punish the dragons until they take the bait, and then when they've spent a couple of rounds taking fire from the Sharpshooter (or Warlock) and breaking through the hypothetical screen of Planar Bound Air Elementals, and are just about to start inflicting punishment on the PCs in return, wham! the wizard decloaks from Invisibility V and Forcecages a dragon and then retreats 100' on his Phantom Steed (preparing for a Wall of Force next round), the Sharpshooter continues to kill off the weakest dragon with his arrows, and the other two PCs do whatever they do (e.g. Conjure Animals, or activating Antipathy (Green Dragons)), and the dragon's situation is looking pretty bad now even before they start using their breath weapons only to discover that they don't really work on these PCs (Heroes' Feast or at least Protection From Poison).

But it's not really because of AC in this case. You're right about that. The broader point is that PCs have abilities aimed at seizing the initiative, and WotC-created monsters are basically just sacks of HP which are designed to die in hopeless combat, with very little tactical depth. Some of the better monsters might have some counterplay available to them, but they aren't supposed to be able to enage in counter-counter-counterplay.


So, if the Unkillable Tank just keeps going and going and going, and isn't down until he's fought for 55 rounds, but in that time he eventually kills all of the enemies, do you count that as a victory? I wouldn't. Because in that time, all of the other party members whose RTD is less than 55 are already dead. All but one of the party members dying is a terrible outcome.

For a truly successful fight, one where none of the heroes die, you need to multiply everyone's DPR by the RTD of the least durable character, and therefore, whenever you find an item that boosts survivability, you want to give it to the character who's currently least durable.

What makes you think the other PCs didn't just spend 55 rounds Hiding/Disengaging/Dashing away, or hiding behind Unkillable Tank in a chokepoint? If you know Korath the Unkillable can kill all of the Grues without you, but the Grues can kill you, the courteous thing to do for Korath is to get out of dodge and let them do his job. Then later on you'll help him kill the things that bypass AC, like those hypothetical Green Dragons, or a bunch of Flameskulls. Save your strength for when it matters.

People like to say "don't split the party, ever" but it's not really sound tactical advice. If your total-DPR:worst-RTD ratio is improved by removing some characters from combat, get them out of the combat.

[Ninevehn]

Well, you know, I already explicitly mentioned it. And I don't bother to mention 'water is wet' stuff like 'The DM can arbitrarily change any or all abilities that a creature has at any time for any reason', because that's the way RPGs work.

I didn't say anything about the DM arbitrarily changing things, funnily enough. My point is, putting the ring on the Wizard is unlikely to actually make things better for him if it won't help the tank either.


However, I do think that if someone claims 'this character could tank six CR15 enemies at once', one should assume that they're tanking CR15 enemies as listed in RAW.

Like purple worms. I know, I know, you mentioned them, but continue on anyways. What CR or monsters would you suggest as a baseline for analysis?

[Chronos]

Quoth MaxWilson:

There's also counterplay to consider: if the heavily-armored warrior is first to kick in your door, and the spellcaster is 200' away on top of a 30' tall stone monolith blasting away with Spell Sniper Agonizing Repelling Blast, then both of your choices are bad.
Well, obviously, if the heavily-armored warrior is the only one capable of doing anything to you, then you take out the heavily-armored warrior first, and then maybe after you're done with him you go seek out his ally who for some reason was staying out of the fight. I thought that was taken for granted. Of course, you'll still prefer to use non-attack-roll abilities against the warrior, whenever possible. And incidentally, why is your warlock just leaving it up to the warrior, instead of contributing himself?

[OverLordOcelot]

I didn't say anything about the DM arbitrarily changing things, funnily enough. My point is, putting the ring on the Wizard is unlikely to actually make things better for him if it won't help the tank either.

dragons might not have all their breath weapons available, mummy lords don't always have Harm available. I just don't see any of this stuff cutting one way but not the other.

You did, in fact, talk about the DM arbitrarily deciding that dragons would not have their breath weapons available or that mummy lords would not always have harm available.


Like purple worms. I know, I know, you mentioned them, but continue on anyways. What CR or monsters would you suggest as a baseline for analysis?

I wouldn't. As I said on page one, I don't think that mathematical modeling is going to provide good answers "I don't think this is something you can answer by doing any reasonable math. You have to figure out how hits are generally distributed around the group based on the player's actions and enemy actions, and since this isn't an MMO with deliberately 'dumb' enemies using aggro scores, you don't really end up with a simple set of rules to model."

[OverLordOcelot]

On the gripping hand, PCs have the upper hand against most monsters, not just because of AC but because they have so many options for controlling terms of engagement. It depends on the details of the party but for example...But it's not really because of AC in this case. You're right about that. The broader point is that PCs have abilities aimed at seizing the initiative,

These examples serve to illustrate how absurdly useless simple models are at high levels. If you're actually going to try to come up with a model that says anything sensible about where to allocate items for a party doing all of this, then you need to fit all of this stuff into the model. A simple calculation of average HP taken per round really doesn't touch any of that. At lower levels the array of abilities is simpler, but is still much more complicated than a couple of averages.

Also, I will note that almost all of your examples involve the players getting extreme amounts of prep time, foreknowledge, and choice of terrain. I really don't think that the big bad guy who has a lair with a conveniently placed monolith 200' from his door with LOS into his room that isn't a trap so the warlock can perch on it and snipe, who also built his room narrow enough that he can't just step to one the side and break the monolock's LOS, doesn't have any alarms or underlings to warn him that the party is setting up outside, and doesn't have any traps in his lair to make the guy kicking down the door regret it is really that great of an opponent. If that's the kind of thing you enjoy, then enjoy it, but if you're getting that much control over fights and have that much resources available, I don't think you need to spend a lot of time on a mathematical model to allocate a ring of protection. Coming up with a model in this case just seems like a couple of billionaires spending all afternoon clipping coupons to figure out how to save $23.67 at the grocery store, I don't think the +1 AC is actually going to make a real difference.

I do have two questions:

The main tool the PCs have for forcing the dragons to concentrate on the tanks (probably summoned creatures in this case, plus the hypothetical Paladin) is in this case the superior ranged weaponry of the PCs (Sharpshooter Fighter or Spell Sniper Agonizing Eldritch Spear Warlock or similar, plus maybe a Simulacrum of same) which lets them maintain a dispersed formation at long range and retain the strategic initiative.

How would using ranged weapons force the dragons to concentrate on the tanks; why wouldn't the dragons either bypass the tanks or just disengage entirely?


What makes you think the other PCs didn't just spend 55 rounds Hiding/Disengaging/Dashing away, or hiding behind Unkillable Tank in a chokepoint? If you know Korath the Unkillable can kill all of the Grues without you, but the Grues can kill you, the courteous thing to do for Korath is to get out of dodge and let them do his job. Then later on you'll help him kill the things that bypass AC, like those hypothetical Green Dragons, or a bunch of Flameskulls. Save your strength for when it matters.

I would like to know what exactly is the layout of a choke point for 6 huge or gargantuan creatures, who incidentally can just move through the tank's space because they're at least 2 sizes bigger, that allows 6 of them to engage the tank but doesn't let them easily pass by. Also what stops the purple worms tunneling past, mummy lords turn into dust and whirling past, and dragons just flying over?

[MaxWilson]

Well, obviously, if the heavily-armored warrior is the only one capable of doing anything to you, then you take out the heavily-armored warrior first, and then maybe after you're done with him you go seek out his ally who for some reason was staying out of the fight. I thought that was taken for granted. Of course, you'll still prefer to use non-attack-roll abilities against the warrior, whenever possible. And incidentally, why is your warlock just leaving it up to the warrior, instead of contributing himself?

The warlock is contributing with Eldritch Blast.

[MaxWilson]

These examples serve to illustrate how absurdly useless simple models are at high levels. If you're actually going to try to come up with a model that says anything sensible about where to allocate items for a party doing all of this, then you need to fit all of this stuff into the model. A simple calculation of average HP taken per round really doesn't touch any of that. At lower levels the array of abilities is simpler, but is still much more complicated than a couple of averages.

Also, I will note that almost all of your examples involve the players getting extreme amounts of prep time, foreknowledge, and choice of terrain. I really don't think that the big bad guy who has a lair with a conveniently placed monolith 200' from his door with LOS into his room that isn't a trap so the warlock can perch on it and snipe, who also built his room narrow enough that he can't just step to one the side and break the monolock's LOS, doesn't have any alarms or underlings to warn him that the party is setting up outside, and doesn't have any traps in his lair to make the guy kicking down the door regret it is really that great of an opponent. If that's the kind of thing you enjoy, then enjoy it, but if you're getting that much control over fights and have that much resources available, I don't think you need to spend a lot of time on a mathematical model to allocate a ring of protection. Coming up with a model in this case just seems like a couple of billionaires spending all afternoon clipping coupons to figure out how to save $23.67 at the grocery store, I don't think the +1 AC is actually going to make a real difference.

I do have two questions:


How would using ranged weapons force the dragons to concentrate on the tanks; why wouldn't the dragons either bypass the tanks or just disengage entirely?



I would like to know what exactly is the layout of a choke point for 6 huge or gargantuan creatures, who incidentally can just move through the tank's space because they're at least 2 sizes bigger, that allows 6 of them to engage the tank but doesn't let them easily pass by. Also what stops the purple worms tunneling past, mummy lords turn into dust and whirling past, and dragons just flying over?

I'll come back and answer these later after work. For now I'll just note that PCs *do* have recon abilities like Arcane Eye and Divination and so can indeed have lots of foresight and preparation especially against passive opponents like a typical D&D adventure module, that stepping to the side to get out of the warlock's line of fire is fine but means the warlock is restricting your movement options and you're ceding some of the initiative exactly as I claimed, and that the dragon fight doesn't rely on chokepoints--you're conflating the dragons and the hypothetical 55-round Grues.

I agree that active opponents are better and more fun than passive ones, but they are extremely rare in practice because taken to a logical extreme, they imply that all the monsters in the adventure should descend on the PCs at once and stomp them all together, and most DMs don't like doing that. It's bad enough to prepare a Deadly x6 encounter for the PC with six adult green dragons; if you make those adult green dragons ambush the PCs at their inn in the middle of the night before the PCs have even learned of the dragons' involvement in the adventure, you're likely to feel like a jerk who didn't give players a fun game. But if players have advance warning, they can turn the tables just as I've already shown. It's tricky to get the balance right between active and passive and I am still working on it, but WotC for one writes almost entirely passive enemies anyway.



Also, I will note that almost all of your examples involve the players getting extreme amounts of prep time, foreknowledge, and choice of terrain. I really don't think that the big bad guy who has a lair with a conveniently placed monolith 200' from his door with LOS into his room that isn't a trap so the warlock can perch on it and snipe, who also built his room narrow enough that he can't just step to one the side and break the monolock's LOS, doesn't have any alarms or underlings to warn him that the party is setting up outside, and doesn't have any traps in his lair to make the guy kicking down the door regret it is really that great of an opponent.

Suppose for a second that the bad guy had all of these advantages you postulate: guards, no monolith, traps inside, wide room. Would that be enough to prevent a party of PCs from offering him a menu of only bad choices? (Answer: no. It barely even changes the way the PCs kick down the door. The warlock is *already* out of range of any traps--triggering them is the melee guy's raison d'etre, that's basically what kicking down the door consists of, that and dealing with anyone behind total cover--and if guards have alerted anyone inside, that may let them ready actions but it doesn't change their fundamental tactical menu.)

[PeteNutButter]

I would like to know what exactly is the layout of a choke point for 6 huge or gargantuan creatures, who incidentally can just move through the tank's space because they're at least 2 sizes bigger, that allows 6 of them to engage the tank but doesn't let them easily pass by. Also what stops the purple worms tunneling past, mummy lords turn into dust and whirling past, and dragons just flying over?

I think this has dissolved into a discussion about how tanking 5e sucks. If a DM is set on ignoring the front line, that becomes easier and easier as the characters increase in level. Higher CR foes are a lot more mobile and slippery. Many monsters with legendary actions have one that is a movement ability that doesn't provoke OAs, completely nullifying things like sentinel. I think the design team realized this, and added in the ancestral guardian barb and cavalier, but even they only provide disadvantage to the foes.

I think it comes down to how a DM manages "aggro" control. A common method is for mindless or low int things to just attack nearest. Another method is for them to attack whatever did the most damage last turn. For smart foes, like dragons, usually they should be able to attack whatever they perceive is the biggest threat based on a variety of factors, including visible equipment, past actions, and circumstance.

When facing foes that are both smart and mobile, in my experience it usually becomes a damage race, as the team tries to kill the foe before it can take out the party squishies. Deadly fights against these foes almost always result in downed (but hopefully not dead squishies). In these fights is worth considering if the wizard or whatever should have the AC items, but then there are breakpoints. If that +1 AC isn't enough to make the wizard last an extra turn against the particular foe then it's not really helping.

DMs that are regularly having foes that consistently only target the squishy characters are IMO doing it wrong and at risk of having the game devolve into rocket tag as the PCs adapt. If enemies only attack the squishiest target, your optimal goal is to focus almost exclusively on offense (unless you are the squishiest).

Edit: Aggravating the problem of foes getting more mobile and intelligent at higher CRs is the widening gap of character specialization. A character focused on being hard to kill at level 1 is only marginally harder to kill than his or her team. At level 20, the tanky character is many times over harder to kill. That means in order to create a truly deadly encounter for the party the DM has to use foes that can seriously threaten the tanks. These foes are then by their nature capable of shattering the weaker characters in a round. This is one of the many reasons, I find game play begins to breakdown in tier 3, in 5e.

[MaxWilson]

I do have two questions:

How would using ranged weapons force the dragons to concentrate on the tanks; why wouldn't the dragons either bypass the tanks or just disengage entirely?

I would like to know what exactly is the layout of a choke point for 6 huge or gargantuan creatures, who incidentally can just move through the tank's space because they're at least 2 sizes bigger, that allows 6 of them to engage the tank but doesn't let them easily pass by. Also what stops the purple worms tunneling past, mummy lords turn into dust and whirling past, and dragons just flying over?

I have a lull at work so answering this now:

Using ranged weapons forces the dragons to take some kind of action: you can't just ignore the PCs if they're hitting you for 30ish HP per round. At that point you can either run and hide (which maybe would appeal to green dragons' predator psychology, but still, you're six adult dragons hiding from four puny little humans--are you really going to let them make you scatter? here's where illusions (Seeming) and misdirection can be tactically important, because if you see a bunch of air elementals and giant owls before you commit, you are more likely to scatter) or you can engage. It's the tanks' job to buy more time for the ranged guys to make their attacks, and to punish the dragons as they fly by--this involves both making opportunity attacks and also just generally getting in the way. I noted that the tanks are likely mostly summoned creatures, and the ones who are e.g. Air Elementals or Giant Owls are large enough to block a Huge creature's movement, and even Medium tanks are at least difficult terrain for the dragons. So they buy time for the ranged guys to degrade the dragons further, especially if the dragons bite the hook and waste time killing the tanks.

The other job of the tanks is to prevent the dragons from escaping after the trap springs, or at least to harry them in pursuit. A nice Longstrider'ed Air Elemental has 100' of movement, which is enough to keep up with an adult green even if it uses its legendary action to Wing Buffet for an extra 40' of speed. Even if the dragon flees at maximum speed, it can't break contact with the Air Elemental, and it takes an opportunity attack every round of Dashing away per elemental (though at least it gets its Wing Buffet and Tail Attack against the elemental), all the while it's still more than likely taking ranged attacks. You may not be able to finish off all six dragons, but it's reasonable to expect to get at least three of them, maybe four, with a 15th level party using the aforementioned tactics (e.g. invisible wizard on a Phantom Steed springing Forcecage + Wall of Force to catch two dragons), leaving only two to seek revenge.

RE: mummy lords/purple worms/etc., nothing stops them from trying to bypass the tanks, but the ranged attackers' mobility is what stops that from mattering. E.g. a Dashing Purple Worm burrows only 60' per round, 100' if it's on the surface instead of burrowing; a Dashing Mummy Lord using Whirlwind of Sand moves only 100'. It's easy to get enough mobility that those guys will never catch up to you, and even easier to get enough mobility that they'll catch up only after they're already dead, and meanwhile the tanks can punish them with opportunity attacks (e.g. Warcaster: Booming Blade) to make it a winning strategy instead of a stalemate. Combined arms for the win!

[OverLordOcelot]

I have a lull at work so answering this now:

Using ranged weapons forces the dragons to take some kind of action: you can't just ignore the PCs if they're hitting you for 30ish HP per round.

There are a lot of options other than 'just ignore the PCs' and 'cluster on top of each other and start launching melee attacks against a single high AC target who's not the one doing 30hp/round'. Like leaving or flying to attack the archers. I'm not really sure why you're posting so much text instead of just agreeing that the idea of getting the dragons to launch a bunch of melee attacks on the high AC guy is just silly and isn't something you can reasonably make happen.


At that point you can either run and hide (which maybe would appeal to green dragons' predator psychology, but still, you're six adult dragons hiding from four puny little humans--are you really going to let them make you scatter?

a Simulacrum of the fighter will add to the party's DPR and let them exploit those surprise rounds better. Planar Bound Air Elementals and/or a skeleton army can engage individual dragons... Even the Paladin can get into the game (despite his likely melee specialization) with his Find Greater Steed Pegasus.

Every single dragon has 18 int, 15 wis, 17 CHA; that's likely double what the party has for average INT, and include an intellect at least on par with great generals of history like Alexander the Great or Julius Caesar. I think that's smart enough to get the concept that powerful humans can and do kill dragons, and to evaluate the relative battlefield capabilities of the two groups. The idea that the dragons cannot possibly react to the party as if they're a threat is just silly. Especially since it's not four puny humans - as I quoted, you've talked about an extra copy of a human, a skeleton army, planar bound air elementals, and a pegasus.


It's the tanks' job to buy more time for the ranged guys to make their attacks, and to punish the dragons as they fly by--this involves both making opportunity attacks and also just generally getting in the way. I noted that the tanks are likely mostly summoned creatures, and the ones who are e.g. Air Elementals or Giant Owls are large enough to block a Huge creature's movement, and even Medium tanks are at least difficult terrain for the dragons. So they buy time for the ranged guys to degrade the dragons further, especially if the dragons bite the hook and waste time killing the tanks.


I would really like to hear a description the layout of this battlefield. It's large enough that people are sniping from hundreds of feet away with hundreds of feet behind them to move away and keep firing, but that's also constrained enough that you can summon enough 57+ hp large creatures to block flying movement - the giant eagles you mentioned won't do much as they only have 26HP, even if you double their HP and they make their save, a single dragon can clear a giant hole in the wall for the others with a single breath attack. (And if they spread out, the dragons are not going to be very slowed by OP attacks that hit on a 14 for 10 damage). It's bad enough that you keep picking battlefields that exactly suit the PCs, but this seems to need to be both extremely open and extremely constrained at the same time, which doesn't work topologically.


Suppose for a second that the bad guy had all of these advantages you postulate: guards, no monolith, traps inside, wide room. Would that be enough to prevent a party of PCs from offering him a menu of only bad choices? (Answer: no. It barely even changes the way the PCs kick down the door. The warlock is *already* out of range of any traps--triggering them is the melee guy's raison d'etre, that's basically what kicking down the door consists of, that and dealing with anyone behind total cover--and if guards have alerted anyone inside, that may let them ready actions but it doesn't change their fundamental tactical menu.)

Honestly, the more I read of your posts it sounds like you're either talking pure theorycraft or your DM plays absurdly softball with your party. The fact that you see a conveniently placed monolith and think "I'll stand on it because I'm obviously our of range of any possible traps, there's no way this could ever do anything nasty or be the focus of any kind of summoned minion" instead of "yeah, I'm not falling for that obvious trap" is honestly amusing to me. And how is the warlock out of range of 'I'll trigger the secondary stone door to the room to close, forcing the warlock to either sit out the fight or suffer the side effects of my anti-teleport ward'? If your DM runs enemies as passive bags of HP that never require quick action from you, never have plans of their own, and never set really obvious traps on really obvious places like the overly-conveniently placed monolith, then yeah you're going to kick them all over the place. But that's not really part of the game system itself, and it's not a safe assumption that an arbitrary DM running games for level 15s is going to play with such a light touch towards players.

Seriously, do you really not see why I find it utterly hilarious to hear someone say "I'm going to stand on the sniper platform that overlooks the big bad guy's room while the fighter breaks down the door, I'm obviously safe from any traps here" with a straight face?

[MaxWilson]

There are a lot of options other than 'just ignore the PCs' and 'cluster on top of each other and start launching melee attacks against a single high AC target who's not the one doing 30hp/round'. Like leaving or flying to attack the archers. I'm not really sure why you're posting so much text instead of just agreeing that the idea of getting the dragons to launch a bunch of melee attacks on the high AC guy is just silly and isn't something you can reasonably make happen.

What in the world? What in my post makes you think this was in any way what the PCs are doing? Didn't I describe in detail an entirely different approach to the dragons?

[OverLordOcelot]

I think this has dissolved into a discussion about how tanking 5e sucks. If a DM is set on ignoring the front line, that becomes easier and easier as the characters increase in level. Higher CR foes are a lot more mobile and slippery. Many monsters with legendary actions have one that is a movement ability that doesn't provoke OAs, completely nullifying things like sentinel. I think the design team realized this, and added in the ancestral guardian barb and cavalier, but even they only provide disadvantage to the foes.

Attempting MMO-style tanking in a game that isn't an MMO designed for that style of tanking sucks, because you're trying to do a thing that only works with a specific rule set but don't have that rule set forcing it to happen. D&D grew out of tactical games where you fought humans using actual tactics against you, the 'tank/healer/dps' thing is decades more recent than this game. The fact that MMO-style tanking sucks in D&D is one of its strengths, if I wanted to grind down dumb opponents who will just keep attacking the hardest guy to hurt I would play MMOs or CPRPGS that use that mechanic and avoid having to coordinate schedules or leave the house.


Edit: Aggravating the problem of foes getting more mobile and intelligent at higher CRs is the widening gap of character specialization. A character focused on being hard to kill at level 1 is only marginally harder to kill than his or her team. At level 20, the tanky character is many times over harder to kill. That means in order to create a truly deadly encounter for the party the DM has to use foes that can seriously threaten the tanks. These foes are then by their nature capable of shattering the weaker characters in a round. This is one of the many reasons, I find game play begins to breakdown in tier 3, in 5e.

The actual problem is that you have people trying to treat D&D like it follows MMO rules instead of dealing with how it actually works. Characters should be focused on being generally effective, not on specializing to the point that they are useless if the enemies are role played as having anything remotely near human intelligence (or even with animal intelligence, like 'attack whoever hurt me the worst'). If you decide that you have 'tanks' who you stack defensive item after defensive item on, and 'non-tanks' who completely ignore defense and hope that enemies focus on the tanks, of course you'll have problems.

[OverLordOcelot]

What in the world? What in my post makes you think this was in any way what the PCs are doing? Didn't I describe in detail an entirely different approach to the dragons?

You explicitly said that "The main tool the PCs have for forcing the dragons to concentrate on the tanks is in this case the superior ranged weaponry of the PCs" I asked "How would using ranged weapons force the dragons to concentrate on the tanks; why wouldn't the dragons either bypass the tanks or just disengage entirely?" you responded "Using ranged weapons forces the dragons to take some kind of action: you can't just ignore the PCs if they're hitting you for 30ish HP per round." And then in the post that you're 'what in the world' about, I pointed out that the dragons have many choices other than "ignore the PCs hitting you for 30ish HP per round" and "focus on the toughest guy". If you're conceding that the PCs don't have any way to force the dragons to concentrate on the tanks, then we're in agreement, if not then I'm not sure what your confusion is.

[MaxWilson]

You explicitly said that "The main tool the PCs have for forcing the dragons to concentrate on the tanks is in this case the superior ranged weaponry of the PCs"

Specifically this: "The main tool the PCs have for forcing the dragons to concentrate on the tanks (probably summoned creatures in this case, plus the hypothetical Paladin)". That's clearly the opposite of what you're postulating: they're not focusing primarily on the paladin.

I expanded on this in other posts, but bottom line is that the tactic you're mocking has nothing to do what what I posted.


I asked "How would using ranged weapons force the dragons to concentrate on the tanks; why wouldn't the dragons either bypass the tanks or just disengage entirely?" you responded "Using ranged weapons forces the dragons to take some kind of action: you can't just ignore the PCs if they're hitting you for 30ish HP per round." And then in the post that you're 'what in the world' about, I pointed out that the dragons have many choices other than "ignore the PCs hitting you for 30ish HP per round" and "focus on the toughest guy". If you're conceding that the PCs don't have any way to force the dragons to concentrate on the tanks, then we're in agreement, if not then I'm not sure what your confusion is.

You can't, strictly speaking, force opponents to do anything. You can ideally present with a menu of choices that are all bad. If you'll concede that flying through the tanks to try to engage the ranged specialists is a bad move for the dragons, I'll gladly concede that the dragons have other bad options including fighting the tanks and scattering and fleeing to fight another day, and that if the dragons had adequate intelligence about the PCs' forces, they'd quickly determine that scattering and fleeing was in fact their best action. So yeah, the PCs can't force the dragons to stand and fight the tanks.

[Chronos]

Quoth MaxWilson:

The warlock is contributing with Eldritch Blast.
Not if his enemies are all inside of a room that he's 200 feet from the door of, like in your example.

And if the PCs are doing a lot of ranged damage to a dragon, then yes, the dragon has to do something about that, and yes, if they're all spread out, it's going to have to focus on one. But why would the one it focuses on be the one with the highest AC? And why would it ever get close enough to the tank to allow opportunity attacks, since dragons fly at high speed?

Unless you're using flying summoned monsters (which don't have Sentinel) for tanks, in which case you're advocating... that you give the Ring of Protection to the monster you summoned?

[OverLordOcelot]

You can't, strictly speaking, force opponents to do anything. You can ideally present with a menu of choices that are all bad. If you'll concede that flying through the tanks to try to engage the ranged specialists is a bad move for the dragons, I'll gladly concede that the dragons have other bad options including fighting the tanks and scattering and fleeing to fight another day, and that if the dragons had adequate intelligence about the PCs' forces, they'd quickly determine that scattering and fleeing was in fact their best action. So yeah, the PCs can't force the dragons to stand and fight the tanks.

No, I most certainly will NOT concede that. Flying through a bunch of giant eagle tanks you described means risking an op attack that hits on a 14 for 10 damage. And they could also have one of the six dragons breathe on the wall of 26HP giant eagle tanks to clear a giant hole for the others. Even soaking one attack from a tougher tank really isn't that big of a deal, and getting enough tanks to block a wide open 3d area seems a bit tricky (And again, I would really like to hear a description the layout of the battlefield that is large enough that people are sniping from hundreds of feet away with hundreds of feet behind them to move away and keep firing, but that's also constrained enough that you can summon enough 57+ hp large creatures to block flying movement.)

For that matter, I also don't concede that fleeing to fight another day is a bad move for the dragons. If the DM hands that to the players, obviously he can do so. But I would have things set so that there is something to motivate the players to act, that there is some kind of pressure where just letting the dragons fly away is a bad move for the players. Maybe the dragons are part of a 'time bomb' plot and letting them live lets the timer advance, maybe the dragons will lay waste to the city as soon as the players leave the area, maybe the survivors will go join up with an enemy. Fights where the players get to exactly decide all of the terms of engagement, including the exact layout of the battlefield, timing, and positioning, and where the enemy can't just go 'whoa, looks like a trap, I think I won't stick my face in it today' are pretty unusual in my experience. I mean, you can set up a big ambush outside of Acerak's tomb (whether it's horrors or Annihilation) but that's not going to make him engage you.

I've tried to avoid discussing the lopsidedness of your scenarios since this thread was really about AC analysis, but the degree of 'everything goes right for the players, and the enemies have no choice but to fall into the trap' just keeps getting sillier. And come on, how can "I will stand on the conveniently placed monolith 200' from the bad guy's door that has perfect lines of sight, because it means I am safe from any attacks or traps" not be amusing to a veteran adventurer? If that turned out NOT to be a trap i'd be surprised!

[MaxWilson]

Not if his enemies are all inside of a room that he's 200 feet from the door of, like in your example.

Depends on the geometry and what the other PCs are doing. You're focusing on the minutiae and not on the point, but even if all of the enemies clear the doorway, the warlock can e.g. Ready Eldritch Blasts to kill anyone who crosses the line of the doorway, which effectively splits the enemy in two. Or the Paladin can grapple whoever is in inside and drag them out into the open, or the warlock can maneuver to gain a better angle, or the Paladin can remain in the doorway and just shoot enemies with a bow or attack them with a halberd while the warlock continues to provide covering fire. As long as the enemy can't just blithely ignore the warlock, he's contributing, and the Paladin will cooperate with him to make sure that happens.


And if the PCs are doing a lot of ranged damage to a dragon, then yes, the dragon has to do something about that, and yes, if they're all spread out, it's going to have to focus on one. But why would the one it focuses on be the one with the highest AC?

It probably won't be, if there are six dragons. The one with the highest AC can maneuver to make one dragon's life inconvenient (via Sentinel or whatnot), but he can't do that for all six.


And why would it ever get close enough to the tank to allow opportunity attacks, since dragons fly at high speed?

Pegasus + mounted combat rules + PC spells (like Longstrider) means the Paladin has a speed potentially equal to or greater than the dragons' speed, and additionally the geometry of the attack means the paladin can intercept the dragons. It's like asking why the defending football team can tackle the guy catching the ball even though they're all about the same speed: of course it's hard to penetrate a line of defenders unless you are much, much faster than any of them. It's slightly easier in 3D but still not going to happen in this case.


Unless you're using flying summoned monsters (which don't have Sentinel) for tanks, in which case you're advocating... that you give the Ring of Protection to the monster you summoned?

Nope. At this point we're not discussing Rings of Protection at all, we're discussing principles of strategy. I already said way back that AC isn't the best primary defense, and that the Ring of Protection super-tank is only relevant in certain scenarios. This isn't one of them.

[OverLordOcelot]

Not if his enemies are all inside of a room that he's 200 feet from the door of, like in your example."

I think Panic! At the Disco has a song that's relevant here; something about 'Haven't you people ever heard of...'

[MaxWilson]

No, I most certainly will NOT concede that.

Well, then you're wrong.


Flying through a bunch of giant eagle tanks you described means risking an op attack that hits on a 14 for 10 damage.

Yep.


And they could also have one of the six dragons breathe on the wall of 26HP giant eagle tanks to clear a giant hole for the others.

A medium-sized hole which doesn't even clear out any elementals in the first place, and which won't last long enough for all the dragons to penetrate the formation--so you're really just splitting your forces. And the fact that you have to slow down to make the hole (breathing instead of Dashing) is exactly the point. It's a bad tactic, and I'm happy for you to think it's a good tactic because trapping you and destroying you is the whole point.


Even soaking one attack from a tougher tank really isn't that big of a deal, and getting enough tanks to block a wide open 3d area seems a bit tricky (And again, I would really like to hear a description the layout of the battlefield that is large enough that people are sniping from hundreds of feet away with hundreds of feet behind them to move away and keep firing, but that's also constrained enough that you can summon enough 57+ hp large creatures to block flying movement.)

"Hundreds of feet" isn't nearly as big as you think it is. If you want me to come up with a battlefield I can, but it would be harder to come up with a realistic battlefield where flying is relevant but long-ranged attacks aren't. Gary Gygax recommended battlespaces on the order of 1200' x 1800'.


For that matter, I also don't concede that fleeing to fight another day is a bad move for the dragons.

It's the best of their bad moves but it means conceding whatever objective they were covering. Since dragons are kind of known for hoarding stuff, that could potentially be quite bad. Naturally hoards are underground, so the PCs would still need to leverage their ability to force the dragons to flee open terrain, into an ability to force them to free from an underground objective--but even if all they can do is hold open terrain, you're still letting humans blockade you in your underground lair. If you think that's not a negative outcome for the dragons, well, I think you're wrong. It's just better than the other bad options.


If the DM hands that to the players, obviously he can do so. But I would have things set so that there is something to motivate the players to act, that there is some kind of pressure where just letting the dragons fly away is a bad move for the players. Maybe the dragons are part of a 'time bomb' plot and letting them live lets the timer advance, maybe the dragons will lay waste to the city as soon as the players leave the area, maybe the survivors will go join up with an enemy. Fights where the players get to exactly decide all of the terms of engagement, including the exact layout of the battlefield, timing, and positioning, and where the enemy can't just go 'whoa, looks like a trap, I think I won't stick my face in it today' are pretty unusual in my experience.

I already addressed this upthread: so are Deadly x6 encounters like six adult green dragons against four level 15 PCs. DMs who arrange those kinds of encounters and stack the deck against the PCs tactically by making the dragons ambush them with no warning will probably just end up feeling like jerks who give the PCs an un-fun game to play.

Nothing is ever as simple in play as it is in a forum post, but I didn't claim that it was. I didn't say there won't be counterplay. Having scattering dragons go and devastate civilians in exchange for losing their hoard is exactly the kind of consequence I'd reach for as a DM. That's what gameplay is all about, and that is also why the PCs would be wise to entrap and destroy as many of the dragons up front as they can--chasing down two vengeful dragons is a lot better than chasing down six. That's how I'd set up the adventure anyway: once you've ambushed the dragons and taken their stuff, it's now completely fair game for them to insert themselves into future scenarios at horribly inconvenient times until you deal with them once and for all.

WotC doesn't tend to write adventures with active opponents though.


I've tried to avoid discussing the lopsidedness of your scenarios since this thread was really about AC analysis

You've harped on it a lot, to the degree that I suspect you're mostly a Combat As Sport DM and not a Combat As War or sandbox DM, but that's not unusual for GITP.


but the degree of 'everything goes right for the players, and the enemies have no choice but to fall into the trap' just keeps getting sillier. And come on, how can "I will stand on the conveniently placed monolith 200' from the bad guy's door that has perfect lines of sight, because it means I am safe from any attacks or traps" not be amusing to a veteran adventurer? If that turned out NOT to be a trap i'd be surprised!

I suppose it's not your fault that you can't read my mind: please understand that in the scenario in my head when I wrote that post, that particular stone monolith is not noteworthy, that there are lots of buildings and statues around the area (which is why it's only 200' range and not 600'), that the warlock isn't particularly hampered if there is no monolith at all (Ascendant Step or climbing a tree or even just standing on flat ground all work perfectly well). Maybe I should have written "foobar" or "XYZ" instead of Stone Monolith--I was assuming that you had some experience with realistic terrain and Combat As War scenarios, but from your posts I can see that that's not the case and you're very much of the more gamey, CAS type.

It is odd to me though that you think things like the stone monolith are important when I've clarified more than once that they're not, and that you are welcome to build a scenario where the villain has all the advantages you have asked for (guards and no monolith and traps, etc.) but it won't change anything significant about how the party should kick down that door.

[Ninevehn]

The dragons have a breath weapon that recharges on a 5-6, so they only have their big AC ignoring attack available about 1/3 of the time, and need to use AC attacks the rest of the time, again unless they feel the need to grapple/prone. Mummy lords have the harm spell, but don't have unlimited castings of it.


You did, in fact, talk about the DM arbitrarily deciding that dragons would not have their breath weapons available or that mummy lords would not always have harm available.

See what I'm saying?

[OverLordOcelot]

I suppose it's not your fault that you can't read my mind: please understand that in the scenario in my head when I wrote that post, that particular stone monolith is not noteworthy, that there are lots of buildings and statues around the area (which is why it's only 200' range and not 600'), that the warlock isn't particularly hampered if there is no monolith at all (Ascendant Step or climbing a tree or even just standing on flat ground all work perfectly well). Maybe I should have written "foobar" or "XYZ" instead of Stone Monolith--I was assuming that you had some experience with realistic terrain and Combat As War scenarios, but from your posts I can see that that's not the case and you're very much of the more gamey, CAS type.

If your idea of 'combat as war' is that the PCs get to conveniently pick the terrain that suits them and enemies dwell in simple rooms with no traps where they're always taken by surprise and can't even close their own door to protect themselves, or where they always walk headlong into whatever trap you set and that a DM is bad if they have enemies do anything but let you win, then I certainly am not into that. Though I'm not sure how you can claim with a straight face that having an enemy react to you is more 'gamey' or more of 'combat as war' than having the enemy just let you get away with stuff.

And I've clearly had experience with realistic terrain like 'castles and keeps that don't give you a clear 200' LOS into the throne room from outside', or even 'ordinary houses that have a foyer in the front, which renders the sniper 200' away completely useless as the boss will be in the master bedroom that the front door doesn't open into. Or even 'buildings that have doors that are normal height, which is easily tall enough for a normal person to walk through but not tall enough that a person 200' from the door at an elevation of 30' can see anything very far behind the door'.

Your sneering tone and accusations of other people playing in a 'gamey' fashion really don't work when your scenarios all require enemies that act like MMO monsters and charge headlong into your trap or sit around in a 'lair' that has no defenses but walls and that don't respond when you set up ambushes, and definitely don't follow Panic! at the Disco's advice for dealing with a sniper. You seem to have some weird idea that if a DM doesn't have monsters just walk right into whatever theorycraft-style trap you set, you can say they're a bad DM and you win. If you want to play a game that way, more power to you - but D&D rules and published modules don't require that, and I don't think it's a situation that people should use as a guide in less softball games, or for outfitting parties to deal with published modules that don't generally work like that.


It is odd to me though that you think things like the stone monolith are important when I've clarified more than once that they're not, and that you are welcome to build a scenario where the villain has all the advantages you have asked for (guards and no monolith and traps, etc.) but it won't change anything significant about how the party should kick down that door.

So what you want to happen is: BurlyMan kicks down the door, falling block trap triggers and injures BurlyMan while blocking the door. Boss and allies leisurely defeat Burlyman while the rest of the party is trapped outside of the lair by the trap. At higher levels allies could try to teleport in, but at that point the boss is in a building out of a published module, and their teleport either results in nothing, or worse ends with them naked in a room with a demon while their equipment is in another building. I think if you played a bit of Tomb Of Annihilation, to use one published adventure that really punsihes this, you'd be a bit less keen on having one character inside of a building or room and the rest at a long distance.

About the dragon ambush scenario:

"Hundreds of feet" isn't nearly as big as you think it is. If you want me to come up with a battlefield I can, but it would be harder to come up with a realistic battlefield where flying is relevant but long-ranged attacks aren't. Gary Gygax recommended battlespaces on the order of 1200' x 1800'.

I think that hundreds of feet is hundreds of feet big, and that it's not nearly as easy to constrain people's movement in three dimensions as you think it is. The problem is not that whether you can get enough range, the problem is that you're trying to set up a front line to constrain the movement of flying creatures in a 3d space. Making a line to block movement on the ground is easy, and it's even possible to fill an area with fairly big creatures, but making such a wall in a 3d space is a lot harder. What you seem to need is something less like a 1200x1800 area, and something more like a long, narrow tube that the enemies are compelled to enter and engage you in. If the space is something silly like the compelled tube, then that's a rather large problem with the plan.


A medium-sized hole which doesn't even clear out any elementals in the first place, and which won't last long enough for all the dragons to penetrate the formation--so you're really just splitting your forces. And the fact that you have to slow down to make the hole (breathing instead of Dashing) is exactly the point. It's a bad tactic, and I'm happy for you to think it's a good tactic because trapping you and destroying you is the whole point.

Since the dragons move 160' in one round if they dash and a breath weapon is a 60' cone, you're going to need a really deep wall to have one that they can't just zip through in one round if one of them breathes a hole in it. So I'm really interested in what kind of formation you're envisioning here, exactly how many summoned creatures there are and how they are arranged. It's going to be a lot even if you have the silly cylinder scenario, and to make a sphere or hemisphere of them to have a wall in a 1200' x 1600' battle ground is going to take an astronomical number. So, again, the actual geometry would be of great interest here, as you're clearly not making up any kind of achievable scenario, but just stringing stuff along without the least effort at envisioning how it works.

And coming up with what amounts to 'if the DM lets us have a huge air elemental army, we can mob the enemy and win' just isn't the amazing tactical innovation you seem to think it is. Also a number of the things you're using (like giant eagles) can only be conjured for an hour at a time, so you either need to keep actions free after the start of the fight to create the wall, or come up with enough spell slots to maintain all of these hour-long spells indefinitely.


It's the best of their bad moves but it means conceding whatever objective they were covering. Since dragons are kind of known for hoarding stuff, that could potentially be quite bad. Naturally hoards are underground, so the PCs would still need to leverage their ability to force the dragons to flee open terrain, into an ability to force them to free from an underground objective--but even if all they can do is hold open terrain, you're still letting humans blockade you in your underground lair. If you think that's not a negative outcome for the dragons, well, I think you're wrong. It's just better than the other bad options.

No, they're conceding letting you do some nonsense in a field (or weird cylinder or whatever geometry it is) doing nothing effective while spending a lot of time and money to prepare. You're spending 1000gp/day for heroes feast and 1000gp/elemental/month for planar binding, and plus whatever it costs to assemble and bring in the fleet of air elementals. For all of this effort and expense you get to... stand in an ambush formation. Your 'blockade' costs you a lot of money and prep time, and can't even be maintained for a day - at some point you will have to break ambush formation, sleep, and get together for the heroes feast and whatever other buffs you want. In the long run, you've got to maintain a supply line of jewels and heroes feast materials to keep all of this going. It all sounds a bit silly, and certainly isn't the "I have given them only bad options to choose from" that you seem to think it is.

I'm not really aware of any published modules that you win by standing in a field burning through money, would modify one if it was that dumb, and would not set up such a scenario myself in the first place.