Offense > Defense?

[Casualoblivion]

Long time D&D player and character optimizer here. I've been playing 5E for a few weeks now, read up on optimization guides, and have built a number of characters from level 1-20, and have come to the general conclusion that focusing on offense over defense seems like the optimal character building strategy in 5E. Am I right or wrong about this?

In my character building studies, I've found it to be pretty trivial to get just about any character to have 14 con and 16-18 AC, and in play having such seems survivable enough. Combats are fast in 5E, and it truly seems better to focus on offense and survive by getting attacked less.

[brainface]

Long time D&D player and character optimizer here. I've been playing 5E for a few weeks now, read up on optimization guides, and have built a number of characters from level 1-20, and have come to the general conclusion that focusing on offense over defense seems like the optimal character building strategy in 5E. Am I right or wrong about this?

In my character building studies, I've found it to be pretty trivial to get just about any character to have 14 con and 16-18 AC, and in play having such seems survivable enough. Combats are fast in 5E, and it truly seems better to focus on offense and survive by getting attacked less.

Honestly, I feel focusing on offense is a more efficient use of your time, more or less regardless of how the math works out.

I play in a campaign where I've made a very tanky dude that's not so good with the damage, and I think it works out well since 2-4 other people are focused entirely on damage--I think combat might get grueling pretty fast if we were all focused on survivability, and I think we'd fall prey to spellcaster screwage and opponents crits much easier, simply because they'd have more time to bring spells to bear, roll a lucky crit against us, and so forth.

[pwykersotz]

It really depends on the style of campaign, but overall I would agree. The proactive approach has advantage.

[Casualoblivion]

I went proactive myself in my current Curse of Strahd campaign and it has seemed to work so far. I rolled a Paladin, and have gone for max offense at the expense of some defense. AC 16 from just Chainmail(hard to buy anything better in Barovia) while using a Greatsword with Great Weapon Style. I use my magic for Smite and Misty Step only, I don't bother with healing(outside of LoH) or buffs.

[Giant2005]

Mathematically, defense trumps offense in a significant way (you can test it yourself easily enough by comparing offensive and defensive builds against any creature with infinite hit points - the defensive build will always inflict more damage before being dropped).
However, Brainface has the right of it - combat is boring as hell and concentrating on defense over offense just extends that boredom. Nobody wants that regardless of what is more effective.

[Casualoblivion]

Mathematically, defense trumps offense in a significant way (you can test it yourself easily enough by comparing offensive and defensive builds against any creature with infinite hit points - the defensive build will always inflict more damage before being dropped).
However, Brainface has the right of it - combat is boring as hell and concentrating on defense over offense just extends that boredom. Nobody wants that regardless of what is more effective.

Monster HP are not infinite however, and a monster that is dead or disabled does not hit back. It's been my experience so far that defeating enemies as fast as possible so they get fewer turns is a strong defensive strategy.

[Easy_Lee]

In addition to 5e combat being fast, and monsters dropping relatively quickly compared to 4e, there are a wide variety of damage types and effects a defensive character needs to ward against.

Sample combat: I'm up against two unknown enemies, and I'm alone.

They might deal fire, cold, poison, thundering, radiant, etc. damage. I have to find a way to resist all of these.
They might force a save. I need to be proficient in all saving throws. Being immune to poison and disease will help.
They might deal heavy melee damage, I need to have high AC or avoid entering melee range
They might utilize ranged attacks. I need to find cover or use an ability to defend myself against these.
They might try to grapple or shove me into obstacles. I need a good defense against that.

The list goes on. As we can see, it's very difficult to build a character who is truly resistant to being hurt. Barbarians are great against raw damage, but their defenses have holes. Monks are great against spells, saves, and ranged attackers, but will fall to heavy melee damage, and their INT and CHA saves are pretty low even after 14. Paladins have healing and great saves, but can often be brought down with AoE effects and similar.

If you want a really defensive character, you might build a Monk 14 / Paladin 6. However, that character doesn't really come online until about level 16. Alternatively, you can make a shield mastery oath of devotion paladin and cover most of your bases. But that character doesn't deal very much damage, so you may lose a war of attrition.

Examine the above combat once more. What if I just play an Assassin 3 / BM Fighter 5 with the Alert and Observant feats, and expertise in stealth and perception? Enemies and traps will basically never escape my notice (observant, perception). I can't be surprised (Alert). I'm going to win initiative pretty much every time (Alert, Dex). And the first round of combat, if I win initiative, I get a full round of crits. Know how many enemies are resistant to crits from a magic weapon? Very few. And magic weapons are, generally, available in campaigns.

And if I'm worried about enemy defenses, I can play a blade-pact warlock. Between force damage, save-or-suck spells, and a guaranteed magic weapon, there is nothing in the monster manual which a bladelock can't hurt. Stick alert on that guy, and you can create basically the same sort of character. Never surprised, always acts first, always able to inflict damage.

In short, it's very easy to focus on offense, but very difficult to focus on defense. There are too many ways in which your character can get hurt, but there are also quite a few readily-available damage types which few enemies resist. And a dead enemy can't hurt you anymore.

tl;dr: strike first, strike hard, and strike with a small variety of damage types for best results in D&D 5e.

[MaxWilson]

Long time D&D player and character optimizer here. I've been playing 5E for a few weeks now, read up on optimization guides, and have built a number of characters from level 1-20, and have come to the general conclusion that focusing on offense over defense seems like the optimal character building strategy in 5E. Am I right or wrong about this?

In my character building studies, I've found it to be pretty trivial to get just about any character to have 14 con and 16-18 AC, and in play having such seems survivable enough. Combats are fast in 5E, and it truly seems better to focus on offense and survive by getting attacked less.

You're wrong about that, although conventional wisdom will tell you you're right.

It's much easier to halve the amount of damage you take than to double your DPR, because the designers kept a tighter rein on DPR than on defense. But they both have identical effects on combat: you will end the combat with more HP.

Combats are not a fixed length. If you focus on offensive novas instead of a balanced offensive/defensive capability, you'll never be able to handle hard fights. You'll probably die the first time you run into a platoon of drow warriors, or four Invisible Stalkers at the same time. Offensive strategies are brittle.

The one benefit to offensive strategies is that they don't really require cooperation from other PCs. Even if all of your teammates are tactical idiots, killing the bad guys ASAP will probably keep them alive. Ergo, offensive strategies are good (or at least better) for games with strangers and other people that you don't trust.

P.S. 14 Con and AC 16-18 is pretty pathetic defense. Try an AC 21 Paladin who grapples the Hexed (Str) enemies and then knocks them prone while everyone else kills them. AC 21 + disadvantage to (prone) attackers is three or four times as powerful defensively as AC 16 against CR 5-10 enemies; but you'll never get a character who is three or four times as powerful offensively.

Other good defensive tactics include chokepoints, caltrops, creating partial cover w/ a shovel or Mold Earth, creating killing grounds (archers like to be far away from each other and from enemies), Dodging when attacked while other ranged PCs do the killing, eating opportunity attacks to deny enemies real attacks, spells like Blur/Sanctuary/Shield of Faith, eating damage in favor of healing it after combat via Aura of Vitality, illusions including Disguise Self (Medusa) to make enemies act in counterproductive ways, gaining surprise via Pass Without Trace and/or regular Stealth and/or scouts (Chain Warlock familiar or a Rogue), leveraging mobility (Mobile feat, Expeditious Retreat, Longstrider, Fly, etc.), minions (skeletons) who throw nets to eat up the enemy's action economy, Conjure Animals, and hiding.

[Casualoblivion]

You're wrong about that, although conventional wisdom will tell you you're right.

It's much easier to halve the amount of damage you take than to double your DPR, because the designers kept a tighter rein on DPR than on defense. But they both have identical effects on combat: you will end the combat with more HP.

Combats are not a fixed length. If you focus on offensive novas instead of a balanced offensive/defensive capability, you'll never be able to handle hard fights. You'll probably die the first time you run into a platoon of drow warriors, or four Invisible Stalkers at the same time. Offensive strategies are brittle.

The one benefit to offensive strategies is that they don't really require cooperation from other PCs. Even if all of your teammates are tactical idiots, killing the bad guys ASAP will probably keep them alive. Ergo, offensive strategies are good (or at least better) for games with strangers and other people that you don't trust.

P.S. 14 Con and AC 16-18 is pretty pathetic defense. Try an AC 21 Paladin who grapples the Hexed (Str) enemies and then knocks them prone while everyone else kills them. AC 21 + disadvantage to (prone) attackers is three or four times as powerful defensively as AC 16 against CR 5-10 enemies; but you'll never get a character who is three or four times as powerful offensively.

Other good defensive tactics include chokepoints, caltrops, creating partial cover w/ a shovel or Mold Earth, creating killing grounds (archers like to be far away from each other and from enemies), Dodging when attacked while other ranged PCs do the killing, eating opportunity attacks to deny enemies real attacks, spells like Blur/Sanctuary/Shield of Faith, eating damage in favor of healing it after combat via Aura of Vitality, illusions including Disguise Self (Medusa) to make enemies act in counterproductive ways, gaining surprise via Pass Without Trace and/or regular Stealth and/or scouts (Chain Warlock familiar or a Rogue), leveraging mobility (Mobile feat, Expeditious Retreat, Longstrider, Fly, etc.), minions (skeletons) who throw nets to eat up the enemy's action economy, Conjure Animals, and hiding.

Being offensively focused doesn't mean you end up one-dimensional. My max offense Paladin can still cast Shield of Faith and spam dodge in a chokepoint. Being truly optimized includes a bit of resilience and flexibility.

I didn't say 14 Con and 16-18 AC was good defense, I said it was good enough and anybody can do it. How many rounds does combat need to go before having 3-4x better defense starts to matter? If combat only lasts 2-4 rounds most characters can take a hit or three just fine. Focusing on offense ends things faster and you get attacked less. An enemy swinging at you with better defenses isn't as good as a dead/disabled enemy not attacking at all.

As for the rest, what you are talking about sounds sketchy and situational to me. 5E combats have been quick and simple in my experience, and getting that technical seems like a big waste compared to the direct approach.

[MaxWilson]

Being offensively focused doesn't mean you end up one-dimensional. My max offense Paladin can still cast Shield of Faith and spam dodge in a chokepoint. Being truly optimized includes a bit of resilience and flexibility.

But the math doesn't work out as well--the marginal value of AC increases as you move higher on the AC scale, especially if you're imposing disadvantage on your enemies. Conversely, the marginal value of extra offense decreases as your to-hit or damage increases.

Your question was whether building for max offense was "optimal", and it's not. Your max offense Paladin AC 18 will be less valuable Dodging in that chokepoint than an AC 21; about 50% less valuable in fact.


I didn't say 14 Con and 16-18 AC was good defense, I said it was good enough and anybody can do it. How many rounds does combat need to go before having 3-4x better defense starts to matter? If combat only lasts 2-4 rounds most characters can take a hit or three just fine. Focusing on offense ends things faster and you get attacked less. An enemy swinging at you with better defenses isn't as good as a dead/disabled enemy not attacking at all.

It sounds like you're arguing against optimality: if you "can take a hit or three just fine", and you're going against DMG-built combats, then yes, no matter what you do you're going to win that combat, and you can afford the 20-odd HP from that hit. It would be better to lose only 6 or 7 HP, but you can afford that non-optimal 20 HP.

But your original question wasn't "should I optimize?", it was "is this optimal?"


As for the rest, what you are talking about sounds sketchy and situational to me. 5E combats have been quick and simple in my experience, and getting that technical seems like a big waste compared to the direct approach.

Right--you don't need to optimize (default difficulty) 5E for tactical effectiveness because the game is so easy. You can instead optimize for "fun", or for minimal table time, or whatever. 5E is designed for casual players to have fun with; if the DM sticks to DMG difficulty guidelines and isn't a jerk, the players basically can't lose.

Offensive focus is non-tactically-optimal, but if you enjoy offense, you can do it anyway and you'll be good enough. That's why GITP is full of people who do things like blow paladin spell slots on huge smites--smiting is almost always a poor use of spell slots compared to simply using those same slots on Aura of Vitality healing after combat, but it doesn't matter, because smiting for 20 HP of damage is more fun to those people than healing 70 HP of damage afterwards, and either strategy is good enough to beat up e.g. 2 Displacer Beasts and 2 Giant Boars (Medium encounter for 9th level party), six times per day (4600 out of 30,000 XP adventuring day budget).

It would be different if you were going to be facing down four Displacer Beasts, ten Hobgoblins, a Hobgoblin Captain, and a Flameskull, several times per day. But default 5E difficulty won't have you do that.

[pwykersotz]

It would be different if you were going to be facing down four Displacer Beasts, ten Hobgoblins, a Hobgoblin Captain, and a Flameskull, several times per day. But default 5E difficulty won't have you do that.

Max, I gotta say I find your posts like these fascinating. :smallsmile:

[MrFahrenheit]

Successful use of the banishment spell > offense :P

In all seriousness though...it's all equal in 5e, as long as you play your character the way they were meant to be played. Barbarians are far more defensive, IMO, and paladins far more offensive, with fighters in between...but they all make effective tanks at the end of the day.

[MaxWilson]

Max, I gotta say I find your posts like these fascinating. :smallsmile:

Well, I grew up on the Gold Box games. :) It just doesn't feel right to me unless the PCs are outnumbered.

[Hrugner]

Most of the guides are based on damage out, as that's the mathiest bit of optimization. Effectiveness by encounter type is probably a better metric, but it's harder to quantify and changes in value based on size of party. If you run a five player group you can afford specialists like burst damage people in a way that a two party group can't really. Damage mitigation is less complicated, though you do need to choose if you're blowing resources for defense or just want some good walking around defense.

I find most of the guides don't look closely enough at ensuring all your action types are being used, but class design seems to miss this as well so it's excusable for the player. The guides also often don't look closely at flexibility by encounter type or level range.

If you are optimizing for one type of thing, then damage out makes the most sense. You can't contribute with all your abilities tied up in defensive options in the same way that a player can contribute if they're optimized for damage. Optimizing for pure support seems neglected in 5e, so that's out. But generally, optimization for varied encounter types is going to work out better. You don't want to have your cool character suck simply because he always ends up using a bow he grudgingly bought after the first few encounters.

[Casualoblivion]

But the math doesn't work out as well--the marginal value of AC increases as you move higher on the AC scale, especially if you're imposing disadvantage on your enemies. Conversely, the marginal value of extra offense decreases as your to-hit or damage increases.

Your question was whether building for max offense was "optimal", and it's not. Your max offense Paladin AC 18 will be less valuable Dodging in that chokepoint than an AC 21; about 50% less valuable in fact.

It sounds like you're arguing against optimality: if you "can take a hit or three just fine", and you're going against DMG-built combats, then yes, no matter what you do you're going to win that combat, and you can afford the 20-odd HP from that hit. It would be better to lose only 6 or 7 HP, but you can afford that non-optimal 20 HP.

But your original question wasn't "should I optimize?", it was "is this optimal?"

One of the traps in terms of practical optimization is overkill. If you already survive just fine and aren't being a drain on resources, what does being tougher accomplish? Numbers are just numbers. If they don't have a practical impact on the game, it doesn't really matter how high they go. If I can afford losing 20 hp, it's kind of a waste not to use them.

And why wouldn't we be going against DMG-guidelines in terms of encounters? I would think that the DMG guidelines would be the objective standard. If your game doesn't go by those guidelines, I'm not really sure how your advice really applies.


Right--you don't need to optimize (default difficulty) 5E for tactical effectiveness because the game is so easy. You can instead optimize for "fun", or for minimal table time, or whatever. 5E is designed for casual players to have fun with; if the DM sticks to DMG difficulty guidelines and isn't a jerk, the players basically can't lose.

Offensive focus is non-tactically-optimal, but if you enjoy offense, you can do it anyway and you'll be good enough. That's why GITP is full of people who do things like blow paladin spell slots on huge smites--smiting is almost always a poor use of spell slots compared to simply using those same slots on Aura of Vitality healing after combat, but it doesn't matter, because smiting for 20 HP of damage is more fun to those people than healing 70 HP of damage afterwards, and either strategy is good enough to beat up e.g. 2 Displacer Beasts and 2 Giant Boars (Medium encounter for 9th level party), six times per day (4600 out of 30,000 XP adventuring day budget).

It would be different if you were going to be facing down four Displacer Beasts, ten Hobgoblins, a Hobgoblin Captain, and a Flameskull, several times per day. But default 5E difficulty won't have you do that.

I think you sell Smite a little short. Smite isn't an action, and it can't miss or fail. The most debilitating condition you can inflict on a monster is the dead condition. The dead don't hit back. By itself, it doesn't look like much on paper, but in practical play there is more to it than that. Smite doesn't compare to a spell by itself, but that's a false comparison because a spell takes an action to cast and can fail.

Now, I will say that I have seen in play that my offensive Paladin is less effective against large groups of weaker enemies(against a large group with a dangerous enemy accompanied by a bunch of weak ones, I can focus on the strong enemy). The casters pick up the slack in those fights. It sounds like your games are far more focused against large groups than the typical game, and as such I imagine things are different for you than they would be in a standard game.

[Casualoblivion]

To add to what I said above, spending resources on defense may have a bigger mathematical impact in theory, the impact is much smaller in practice. Smaller in practice when you factor in overkill, resources(your hp and hit dice are a resource, and if you aren't using them that resource is wasted), and the assumed guidelines of the system.

To the inverse of what was said above about buffing defenses having more mathematical impact, to my eyes that makes offensive boosts that much more precious. Defense is easy, offense is hard, and in practical play that makes an offensive buff, even though it's small, have more impact.

[MaxWilson]

One of the traps in terms of practical optimization is overkill. If you already survive just fine and aren't being a drain on resources, what does being tougher accomplish? Numbers are just numbers. If they don't have a practical impact on the game, it doesn't really matter how high they go. If I can afford losing 20 hp, it's kind of a waste not to use them.

That's an interesting way of thinking. I suspect the PC himself would have a different perspective on the value of not getting ripped/torn/shredded for those extra 13 HP of damage--but vanilla 5E doesn't model pain so you're right, in a sense it doesn't matter. (One fun variant rule: track cumulative HP loss over time. That's how much scarring you have. How do you think your players will react to an ogre who has survived well over 1200 HP of damage so far during his lifetime? They'll probably assume that this ogre is much tougher than the normal ogre, and perhaps they'll be right. NPCs can react similarly to the PCs--probably won't want a PC to marry the princess, but might hire him for a tough job. You can likewise say that any damage taken within the past week is a measure of the current physical pain the PC is experiencing, which helps explain why peasants don't just jump off buildings and such even though they can survive 1d6 damage from a 15' fall without problems.)

But that's not an argument for optimizing offense instead of defense, because the end result is actually worse: you spend 20 HP instead of 6 or 7. Optimizing for offense gains you exactly nothing in this hypothetical. That's how it works out in practice too: there's sometimes a few priority targets that are worth spiking damage for (spellcasters, mainly), but against 80% of the MM it doesn't matter if you kill the monster 30% faster via offensive focus; all that really matters is how much damage it deals before you kill it, and good defense gives you options for tilting the combat in your favor. For example, if your defense is sufficiently good, you can be the guy who grapples the Oni, forcing him to target you instead of the squishy wizard. That doesn't work so well if your AC is 16--in that case, you're probably actually squishier than the wizard! You can be the guy who ventures into the darkness to see if the displacer beast is hiding in that cave or if it's the lair of a purple worm instead; you can be the guy who holds the chokepoint so that the Chuuls can't get past to kill the archers; you can be the guy who plays rearguard when your party runs into something too tough for them to handle; you can be the guy who goes into the hut with the orc shaman to negotiate a ceasefire. Typically you want a multilayered defense that for something a bit better than raw AC, but the point is that if you're a glass cannon, you don't have any of these options--it's just "kill or be killed" at maximum ferocity because you can't afford to let anyone else (e.g. orcs) get in a good swing at you.

The best configuration IMO is to have lots of mobility, high AC when necessary, good at-will damage, good spike damage when necessary (via Quicken Spell/Action Surge/smites/whatever), high stealth and perception, and good crowd control options (Hypnotic Pattern, Wall of Force, etc.). Don't put all your eggs in the DPR basket. You'll be more likely to live to a ripe old age.

Unless your DM's game is so easy that everyone lives to a ripe old age--then it doesn't matter what you do, everything works.

[bid]

To the inverse of what was said above about buffing defenses having more mathematical impact, to my eyes that makes offensive boosts that much more precious. Defense is easy, offense is hard, and in practical play that makes an offensive buff, even though it's small, have more impact.
To add to this...

Lets say you have enough DPR to kill 1 monster per round. There are 5 monsters.

You will receive:
- 5 "monster damage" the 1st turn, then you will kill 1
- 4 "monster damage" the 2nd turn, then you will kill 1
- 3 "monster damage" the 3rd turn, then you will kill 1
- 2 "monster damage" the 4th turn, then you will kill 1
- 1 "monster damage" the 5th turn, then you will kill the last one.

Notice that you received 1 + 2 + 3 + 4 + 5 "monster damage". Total damage received is proportional to SQUARED {sum is x*(x+1)/2}. This means +10% offense has the same impact as +21% defense.


So no, defense has to look better because it wouldn't balance otherwise.

[Slipperychicken]

If you are optimizing for one type of thing, then damage out makes the most sense. You can't contribute with all your abilities tied up in defensive options in the same way that a player can contribute if they're optimized for damage. Optimizing for pure support seems neglected in 5e, so that's out. But generally, optimization for varied encounter types is going to work out better. You don't want to have your cool character suck simply because he always ends up using a bow he grudgingly bought after the first few encounters.

I think optimizing for support is fun. My life cleric 'healbot' makes me feel awesome because he keeps the offense-focused guys alive, healthy, and free of status effects, and also gives them bless so they do their job better. He contributes in a way the fighter and barbarian cannot, because he measurably improves their damage output (By about 12.5%), while keeping them alive so they can do their job. He does not deal that much damage on his own, but he is awesome because his work is crucial. Without him, the party would be dead, or else crippled by tetanus and multiple different kinds of insanity.


I think that people neglect optimization for support roles because while some nuances exist on the utility side of support, covering the basics sounds simpler than it is. For most, simply having a bard or druid with healing word available sounds like enough. But in reality it's more complicated because of the breadth of bad things that can happen to PCs. To be really effective, a party's support should have enough healing to keep the party going through the full 6-8 encounters, have the ability to notice and remove as many status effects as possible (including curses, diseases, blindness, deafness, restraints, possession, death, and so on), and keep at least one way to buff the party's ability to do damage. On top of that, support characters should be able to solve one or more common utility problems such as the need for food or the need to identify treasure.

A party might be designed to be great at fighting, but if they can't overcome complications like being low on hit points, catching diseases, being cursed, or dying, then a lot of common adventuring scenarios will basically force them to flee prematurely or else face greatly reduced effectiveness. That is why support is so important, and why I think that it deserves more attention than it gets right now.


That's an interesting way of thinking. I suspect the PC himself would have a different perspective on the value of not getting ripped/torn/shredded for those extra 13 HP of damage--but vanilla 5E doesn't model pain so you're right, in a sense it doesn't matter. (One fun variant rule: track cumulative HP loss over time. That's how much scarring you have. How do you think your players will react to an ogre who has survived well over 1200 HP of damage so far during his lifetime? They'll probably assume that this ogre is much tougher than the normal ogre, and perhaps they'll be right. NPCs can react similarly to the PCs--probably won't want a PC to marry the princess, but might hire him for a tough job. You can likewise say that any damage taken within the past week is a measure of the current physical pain the PC is experiencing, which helps explain why peasants don't just jump off buildings and such even though they can survive 1d6 damage from a 15' fall without problems.)

I think that a fighting person's reaction to the prospect of injury would change greatly if he was certain that someone could instantly cure him for free within seconds of being injured. In the real world, getting hit with a weapon means you might bleed out, might have to spend weeks and face financial losses being cared for, face reduced functionality with the injured area, may even lose a body part entirely, and so on. That makes injury IRL a much more dangerous proposition than hit point loss in fifth edition D&D is. I think 5e would benefit from a better and more nuanced permanent injury system than the DMG variant; players wouldn't be so cavalier (or at least would quickly learn not to be) about losing hit points, if losing too many meant they might lose a finger or have an ear ripped off.

[Sigreid]

Short answer is yes. So long as your defense is high enough to allow you to make use of your offense. Designing to do 300hp per round does you no good if you die before your turn.

[MaxWilson]

I think that a fighting person's reaction to the prospect of injury would change greatly if he was certain that someone could instantly cure him for free within seconds of being injured. In the real world, getting hit with a weapon means you might bleed out, might have to spend weeks and face financial losses being cared for, face reduced functionality with the injured area, may even lose a body part entirely, and so on. That makes injury IRL a much more dangerous proposition than hit point loss in fifth edition D&D is. I think 5e would benefit from a better and more nuanced permanent injury system than the DMG variant; players wouldn't be so cavalier (or at least would quickly learn not to be) about losing hit points, if losing too many meant they might lose a finger or have an ear ripped off.

That's the interesting thing: 5E doesn't say anything about pain, and it doesn't say that Cure Wounds removes the pain--it just restores the HP. There are fantasy stories in genre in which healing magic heals wounds but leaves behind phantom pain; there's nothing in the PHB to contradict that possibility.

It's an interesting variant, anyway.

=======================================


To add to this...

Lets say you have enough DPR to kill 1 monster per round. There are 5 monsters.

You will receive:
- 5 "monster damage" the 1st turn, then you will kill 1
- 4 "monster damage" the 2nd turn, then you will kill 1
- 3 "monster damage" the 3rd turn, then you will kill 1
- 2 "monster damage" the 4th turn, then you will kill 1
- 1 "monster damage" the 5th turn, then you will kill the last one.

Notice that you received 1 + 2 + 3 + 4 + 5 "monster damage". Total damage received is proportional to SQUARED {sum is x*(x+1)/2}. This means +10% offense has the same impact as +21% defense.

So no, defense has to look better because it wouldn't balance otherwise.

This math is wrong. Total damage received is proportional to the number of enemies squared, not anything else; it's linear in the offensive output of the monsters, and linear in the offensive output of the PCs. If you kill them twice as fast but take double damage because of lower AC, it cancels out. You'll take 5*2 + 3*2 + 1*2 = 18 units of damage by being offensive instead of (1+2+3+4+5) = 15 units by being balanced. +100% offense and +100% defense have exactly the same value.

In math terms, the damage dealt by monsters over the course of a fight-to-the-death combat will approximate (monster DPR * number of monsters) * (monster HP * number of monsters / PC DPR) / 2. Refactor and that's (number of monsters)^2 * (monster DPR * monster HP) / (PC DPR). Notice that number of monsters is the only quadratic ("SQUARED") term there--both PC offense and monster offense have linear impact, and PC defense is just the reciprocal of monster offense.

In practice though it's much easier to get a +100% defense boost than a +100% offense boost. If you have a good AC, Blur is +100% defense or more (i.e. it halves the damage you're taking, or maybe cuts it even lower) whereas Haste is typically only a +50% offense boost, or +33% if you're a GWM/PM type with a bonus action attack already. If you use Haste for defense instead of offense, for the mobility, it's more like a 100% defense boost (you can trivially cut your damage taken by 50% merely by Disengaging every other round; other, more effective tactics are possible such as Hiding within heavy obscurement to become untargetable, if you have a decent Stealth).

[Slipperychicken]

That's the interesting thing: 5E doesn't say anything about pain, and it doesn't say that Cure Wounds removes the pain--it just restores the HP. There are fantasy stories in genre in which healing magic heals wounds but leaves behind phantom pain; there's nothing in the PHB to contradict that possibility.

I like it. I've sometimes thought about having the healing itself hurt, kind of like getting stitches or something. Perhaps some kind of pain through the whole area of the injury, as tissues are strained to mend faster than they should.

[Giant2005]

Monster HP are not infinite however, and a monster that is dead or disabled does not hit back. It's been my experience so far that defeating enemies as fast as possible so they get fewer turns is a strong defensive strategy.

No monster hit points aren't infinite - that test is only a metric by which you can determine the ratio of outgoing damage:incoming damage, and the defensive builds win in that metric.
If something is strong enough to pose enough of a challenge that how optimized your character is is actually relevant, then the defensively built character has a greater chance of killing the enemy before it kills them.

[Casualoblivion]

No monster hit points aren't infinite - that test is only a metric by which you can determine the ratio of outgoing damage:incoming damage, and the defensive builds win in that metric.
If something is strong enough to pose enough of a challenge that how optimized your character is is actually relevant, then the defensively built character has a greater chance of killing the enemy before it kills them.

I have no idea what your point is.

Fights aren't one on one. It's party vs monsters. If you're max defense, the monsters ignore you and beat on everyone else. What exactly does a max defense character do? To what purpose does it serve in a party vs monsters situation? Unless you are assuming everyone is max defense.

How does your point apply to the individual classes, like for example the Wizard?

You are also shifting goalposts. The standard of the encounter guidelines listed in the DMG is not the same as what is necessary to challenge an optimized group of PCs.

[Giant2005]

I have no idea what your point is.

Fights aren't one on one. It's party vs monsters. If you're max defense, the monsters ignore you and beat on everyone else. What exactly does a max defense character do? To what purpose does it serve in a party vs monsters situation? Unless you are assuming everyone is max defense.

How does your point apply to the individual classes, like for example the Wizard?

You are also shifting goalposts. The standard of the encounter guidelines listed in the DMG is not the same as what is necessary to challenge an optimized group of PCs.

Firstly, I'm not shifting the goalposts - shifting the goalposts is impossible if there are no defined goalposts in the first place.
Secondly, a defensively strong character isn't absent all damage potential, in fact, their damage potential is only marginally lower than an equivalent person that is going balls to the wall offense. If they are so defensively strong that the enemy decides they can't win and changes their target, then that is great - not only have you absorbed the 1 or more of their attacks required for the enemy to make that determination (shifting the odds of victory further in your team's favor), but you are also getting an opportunity attack out of the transaction which is more than likely enough to push your damage higher than what a balls-to-the-wall offensive character would have been able to achieve.

[MaxWilson]

I have no idea what your point is.

Fights aren't one on one. It's party vs monsters. If you're max defense, the monsters ignore you and beat on everyone else. What exactly does a max defense character do? To what purpose does it serve in a party vs monsters situation? Unless you are assuming everyone is max defense.

He grapples, or holds a chokepoint, or carries out opportunity attacks while everyone else flees (and then executes his own exit strategy afterward, e.g. Expeditious Retreat or Dimension Door). Have you ever tried to execute a defensive strategy when none of your units can stand toe-to-toe with the enemy? No? Sometimes it's possible, in open terrain, but especially in close quarters it is much, much simpler and better to have at least one melee-capable tank that can get in their face and still be pretty much invulnerable to their attacks. (It helps if the melee-capable tank is a switch hitter who can also engage in ranged combat, e.g. a paladin/warlock or paladin/sorc.)


How does your point apply to the individual classes, like for example the Wizard?

You are also shifting goalposts. The standard of the encounter guidelines listed in the DMG is not the same as what is necessary to challenge an optimized group of PCs.

I thought we'd settled this. I've said over and over that if you stick to easy, DMG-Medium/Hard encounters, optimization doesn't matter because you can pretty much do anything and still win. But if you roll up an Nth level party on kobold.com and try to run them through, say, six randomly-generated encounters for a party with three times the actual number of PCs (e.g. for four 11th level PCs, set kobold.com to twelve 11th level PCs and hit "Random Encounter", then do it five more times) you will find that the purely offensive-oriented builds will fail to accomplish their objective more often than builds with a good balance of offense and excellent defense. 5E is constructed so that you don't have to give up much offense, after all, to get excellent defense. A Sharpshooter EK can have amazing damage output and still afford Expeditious Retreat and Shield to keep him safe; a Moon Druid can chug away all day on a single helping of Mage Armor; many spells like Conjure Animals do double duty as offense and defense simultaneously; an Evoker can maintain Wall of Force on some enemies while he (and the rest of the party) plinks away at the remaining enemies with cantrips for respectable damage.

Did you start this thread because you were looking for insight, or did you just want to confirm your bias with people telling you things you already agreed with? If the former, then I encourage you to actually run the aforementioned kobold.com random gauntlet with two separate parties. You'll find that boosting defense by +100% or so is simpler and much more effective than trying to "optimize" your offensive DPR for an extra +25% damage output. That's just the way 5E is designed.

[famousringo]

False choice.

The important thing to optimize is control. A tank is of little use if the enemy ignores her, and your DPR won't be impressive if the enemy force surrounds and focus-fires your squishy butt.

Whether you use chokepoints, mobility, range, stealth, spells... ideally all of the above... control the fight well enough, and you don't need to worry about how strong offense and defense are.

Sun Tzu knew his business.

[Rummy]

I too prefer Defense, although I thing having both is preferable. One note, at very low levels, a Vuman rocking Heavy Armor Mastery turns many fights trivial.

[Socratov]

False choice.

The important thing to optimize is control. A tank is of little use if the enemy ignores her, and your DPR won't be impressive if the enemy force surrounds and focus-fires your squishy butt.

Whether you use chokepoints, mobility, range, stealth, spells... ideally all of the above... control the fight well enough, and you don't need to worry about how strong offense and defense are.

Sun Tzu knew his business.

Haha, that last bit reminded me of this... (https://www.youtube.com/watch?v=h42d0WHRSck)


But what you offer is a false choice as well. Ideally you need all three to be successful:


Offense you need beucase if you can't kill it it will certainly at some point kill you.
Defense you need because it is only logical for something that if you are trying to kill it it will reciprocate the kind gesture and try killing you. The better you get at not being killed, the longer you have to kill the enemy. Simple.
Control you need because it minimizes the resources you need spend on points 1 and 2 to kill the enemy/overcome the obstacle while succesfully following the BeeGee's orders (https://www.youtube.com/watch?v=I_izvAbhExY)


A praty focusing on only 1 aspect won't be successful. A party focusing on 2 aspets of the list can be successful and a party focusing on all three aspects on the list will be successful.

[dev6500]

It seems like everyone in this thread is truly missing the point of character optimization. This is not team optimization or group effectiveness optimization. This is optimization of an individual. Unless you are somehow a selfless team player, then optimization will always follow the rule of optimize until you have the highest defense of someone in your group.

As a hardcore optimizer, there can be serious pitfalls of certain types of optimization. Offensive optimization runs the risk of causing a DM to target you more proactively. When a DM runs a game, they tend to notice the player who deals the most damage/kills the most monsters/generally wrecks the NPCs and normally responds by upping the difficulty of the encounter, targeting the offensively optimized, or both. When the DM begins doing one or both, the truly offensively optimized character will suffer the most because they are likely to have a weaker defense since they focus so purely on offense. Thus you will likely die.

A defensively focused character tends to be harder to hit and affect by disabling abilities a DM might choose to use against them. A DM might notice that one player is harder to hit or disable with a save or suck ability but the DM is likely to respond in 1 of 2 ways. Either they ignore your character because you are harder to hit or they will increase the difficulty of the encounters and increasingly target you. If they choose the first option, then your character is even safer and is unlikely to be wrecked by an encounter and spend multiple turns unconscious like many a party wizard has. If they choose the second option, well your character is still in the best situation of any character in the party because he has a higher defense than the rest of the party and in order for the DM to take you down they may end up forcing a TPK. Thus most times, DMs just avoid attacking defensive characters because in order to successfully hamper a defensive character they often create encounters that squish the other players.

In conclusion, a defensively focused player is the optimal choice for humans who play dnd because humans are selfish and do not like their dnd characters dying or spending rounds unconscious or otherwise disabled. Finally, a defensive character is likely to deal the most damage/kill or finish the most enemies because they are most likely to spend all the turns in a combat actively attacking enemies while offensively focused characters are more likely to spend combat unconscious at points.

[GoodbyeSoberDay]

I like Easy Lee's Fighter/Rogue as an example of an effective character with low traditional defenses, but I'd rephrase his argument. Offense and defense are both quite important, but the most important thing is the ability to choose your engagements effectively. And that's all about perception, stealth, intel, and special abilities.

Think about the drawbacks of over-focusing on offense or defense. If you're a glass cannon, and your enemies know it, you can get focused down before you even get to lay down the smack. If you're a high AC tank taking the dodge action, enemies can ignore you in favor of juicier opponents, and you're not doing much of anything. But if you set up the fight in the right way, the glass cannon gets off a lot of damage before the enemy even blinks, and the tank sets up a position where it's hard to attack anyone but him.

[Cybren]

False choice.

The important thing to optimize is control. A tank is of little use if the enemy ignores her, and your DPR won't be impressive if the enemy force surrounds and focus-fires your squishy butt.

Whether you use chokepoints, mobility, range, stealth, spells... ideally all of the above... control the fight well enough, and you don't need to worry about how strong offense and defense are.

Sun Tzu knew his business.

1) Sun Tzu is incredibly overrated and Spring & Autumn era chinese statecraft/military strategy isn't as broadly applicable as people claim it is
2) A lot of those things are natively available without needing to be optimized for, and are more predicated on having experience in how to maximize using the abilities your character already has. I'd also call it arbitrary to classify mobility, range, and stealth as "control" rather than just concepts completely separate from offense _or_ defense. Having someone with 15HP and 3AC in a choke point isn't very productive, after all, and having superior mobility and range isn't useful if you aren't capable of dealing relevant damage to the enemy in the time-scale you need to (the ridiculous phantom steed/tarrasque example is one, since it presumes the tarrasque has nothing better to do and there's no cost to leading it on a miles-long rampage)

[krugaan]

1) Sun Tzu is incredibly overrated and Spring & Autumn era chinese statecraft/military strategy isn't as broadly applicable as people claim it is
2) A lot of those things are natively available without needing to be optimized for, and are more predicated on having experience in how to maximize using the abilities your character already has. I'd also call it arbitrary to classify mobility, range, and stealth as "control" rather than just concepts completely separate from offense _or_ defense. Having someone with 15HP and 3AC in a choke point isn't very productive, after all, and having superior mobility and range isn't useful if you aren't capable of dealing relevant damage to the enemy in the time-scale you need to (the ridiculous phantom steed/tarrasque example is one, since it presumes the tarrasque has nothing better to do and there's no cost to leading it on a miles-long rampage)

Gonna add to this, generally speaking control is a way of making your offense and defense more efficient, not the be-all and end-all. If anything, it looks something like this:

Offense opposes Defense.
Control opposes Action Economy.

Of course, offense and defense both influence action economy (hard to take actions when you're dead).

[xanderh]

I'm playing an Eldritch Knight optimised for defense, and it's been impossible for my DM to disable him. He started at level 5 and we're at 11 now, and he hasn't been dropped to 0 once so far.
We fought an adult blue dragon at level 10. My two party mates got dropped after two turns due to breath attacks, but I defeated it on my own after they got downed.

And I find that being terrifying is a good way of being the primary target for the enemies. Shout at them, and be particularly brutal in the way you kill enemies, and just make sure that they feel like they can't just ignore you and focus on squishies.

[JellyPooga]

Building a character for Offence rather than Defence is almost always superior if only because Defence has one major weakness...NO defence is perfect. No matter how high your AC, your foes, of which there are many, always have a chance of an auto-hit on a natural 20. The GM has a whole book full of monsters, let alone the NPC's, that he can throw at you to circumnavigate any specific resistances or strengths your defence has. No matter how fast you run or how well you hide, death will always find your character in the end.

If you've got a good offence, on the other hand, you can take more fools with you into the afterlife, than if you built for defence.

[Addaran]

What's important is not one or the other but the balance between both. You check all the options and compare how much difference they do. +1% defence isn't worth +50% offence. Just like +1% offence isn't worth +50% defence.


Building a character for Offence rather than Defence is almost always superior if only because Defence has one major weakness...NO defence is perfect. No matter how high your AC, your foes, of which there are many, always have a chance of an auto-hit on a natural 20. The GM has a whole book full of monsters, let alone the NPC's, that he can throw at you to circumnavigate any specific resistances or strengths your defence has. No matter how fast you run or how well you hide, death will always find your character in the end.

If you've got a good offence, on the other hand, you can take more fools with you into the afterlife, than if you built for defence.

The same can be said about offence, it's never perfect. The DM will always be able to choose a monster or NPC that will be immune to your damage type(s), with better range/mobility or perfect crowd control. =P

[JellyPooga]

The same can be said about offence, it's never perfect. The DM will always be able to choose a monster or NPC that will be immune to your damage type(s), with better range/mobility or perfect crowd control. =P

Offensive options are, as a rule, more versatile though.

A GWM Fighter who normally uses a Greatsword is just as effective with a Maul in hand if faced with an opponent immune to Slashing damage, for example.

Offence is easy, Defence is not. That's why huge expense and time is poured into Defence; fancy chain and plate armour, castles, walls, etc., whilst any bunch of guys with sticks can be an army and be a pretty good one too, even if their enemies have a decent defence. Likewise, all you need is big arms and a bigger stick to be a decent warrior. To be a decent defensive duelist (not the Feat), on the other hand, takes years of training just to get to a point where one lucky blow can still take you out.

[Cybren]

Offensive options are, as a rule, more versatile though.

A GWM Fighter who normally uses a Greatsword is just as effective with a Maul in hand if faced with an opponent immune to Slashing damage, for example.

Offence is easy, Defence is not. That's why huge expense and time is poured into Defence; fancy chain and plate armour, castles, walls, etc., whilst any bunch of guys with sticks can be an army and be a pretty good one too, even if their enemies have a decent defence. Likewise, all you need is big arms and a bigger stick to be a decent warrior. To be a decent defensive duelist (not the Feat), on the other hand, takes years of training just to get to a point where one lucky blow can still take you out.

I'm not sure what that final paragraph is in reference to? Historically, armies didn't use armor because they couldn't afford them, not because of some strategic conceit to offense over defense. After all, in antiquity and the Middle Ages when soldiers were expected to outfit themselves, they went with the best they could afford. The Middle Ages in Europe are litterally defined by defense over offense, too, with fixed defenses in the form of castles and manors and prolonged siege warfare being the most iconic aspect of combat in that era. In terms of dueling, I think you're being nonsensical. Look at the rules for fencing and right of way. That's not an arbitrary rule to punish people for no reason- it's a codified way of training people to not get stabbed

[JellyPooga]

I'm not sure what that final paragraph is in reference to? Historically, armies didn't use armor because they couldn't afford them, not because of some strategic conceit to offense over defense. After all, in antiquity and the Middle Ages when soldiers were expected to outfit themselves, they went with the best they could afford. The Middle Ages in Europe are litterally defined by defense over offense, too, with fixed defenses in the form of castles and manors and prolonged siege warfare being the most iconic aspect of combat in that era. In terms of dueling, I think you're being nonsensical. Look at the rules for fencing and right of way. That's not an arbitrary rule to punish people for no reason- it's a codified way of training people to not get stabbed

My point is this;

Offence is easy, Defence is not.

Historically, you're right and that's my point. Sharpening some sticks and calling them spears is an effective way to build a half-decent army if you've got enough dudes. Equipping those same guys with armour was expensive and largely ineffective, so wasn't done (as a rule). Whatever the latest defense that got developed, an effective offensive counter was short to follow. Make some chainmail? I'll use a spear instead of an axe. Make some plate armour? I'll use bodkin arrows for my bow. Build a castle? I'll just dig a hole under the wall and collapse it. See the pattern? Expensive and time-consuming defense countered by a cheap and effective offence. It's not a strategic conceit; it's plain facts. Defences are hard and expensive to make effectively, whether on a personal or wider scale and frequently come with a drawback of one kind or another (Want a shield? Can't use a bigger sword. Want plate armour? Sorry it's so heavy. Wear a helmet? Nice and dark in there, right? Build a castle? Not exactly mobile, is it?).

As for fencing and right of way, that's a perfect example of my point; a big guy with a big stick swings at your head. Without the codified training you talk about, which took years to develop and more time to teach, you have to be just as big or have a bigger "stick" to compete. Even with the training, one false step and you're a greasy stain on the floor against big-stick-guy. "Having a big stick" isn't exactly a great offence; there's much more effective tools and methods out there, but it takes an actively good defence to defend against it effectively.

[Cybren]

okay, and how is that relevant to gameplay?

[JellyPooga]

okay, and how is that relevant to gameplay?

Let's say you build a "Defence" character. What defences has he got? Any Tiefling and some Dragonborn have resistance to Fire; a great defence, right? Well, yeah, but only if the GM throws Fire damage at you; by no means a certainty. How about if you go for a huge AC? Natural 20's aside, there's no obligation for the GM to roll an attack roll against you...ever. Between AoE's, contested Athletics, mind-affecting spells and so on, your plate armoured knight might never find himself under direct attack. It's not likely, I'll admit, but it's there.

In short, any given Defence is easily circumnavigated. IF you build your character focused on that defence, every time it is circumnavigated, you'll be feeling cheated of your concept.

On the other side of the coin, if you build for Offence you'll almost always feel useful because, as I mentioned; it's typically more versatile.

[Cybren]

"these specific examples of very narrow, isolated defensive abilities are not as versatile as this hypothetical concept of 'Offense' "

???

[JellyPooga]

"these specific examples of very narrow, isolated defensive abilities are not as versatile as this hypothetical concept of 'Offense' "

???

High Strength and Great Weapon Master isn't limited to a single weapon; Greataxe, Greatsword, Maul, Pike...it's an effective offence with multiple "outputs" should you encounter something resistant or immune to one of them.

Contrast this to "having a high AC" which cannot be so manipulated to cater to the situation at hand.

It's a simple example, but I think it demonstrates my point.

[Socratov]

"these specific examples of very narrow, isolated defensive abilities are not as versatile as this hypothetical concept of 'Offense' "

???


High Strength and Great Weapon Master isn't limited to a single weapon; Greataxe, Greatsword, Maul, Pike...it's an effective offence with multiple "outputs" should you encounter something resistant or immune to one of them.

Contrast this to "having a high AC" which cannot be so manipulated to cater to the situation at hand.

It's a simple example, but I think it demonstrates my point.


easy: get High AC? it's circumvented with area spells and combat control, not to mention falling prone, being grappled, etc.

Get magic resistance, it does not apply to physical attacks.

In the game we know it's impossible to create a perfect defense since even if you raise your AC to a gazillion, it's still bypassed by natural 20's.

similarly it's impossible to create a perfect offense. However, since offensive options cost less resources then defense and are more easily obtained it's generally more applicable.

[Cybren]

Great, but now how does that deal with swarms of high AC, low HP enemies? Or enemies that have resistance to non-magical damage, or regenerates HP? Or, say, a Rust Monster? You're overly focused on damage type resistances

[Socratov]

Great, but now how does that deal with swarms of high AC, low HP enemies? Or enemies that have resistance to non-magical damage, or regenerates HP? Or, say, a Rust Monster? You're overly focused on damage type resistances

well, combat wise you have 3 main flavours:

physical
magical/elemental
conditions,

and is deliverd in 3 ways:

attack roll,
skillchallenge,
saveing throw

a decent character will have at least one form either collumn ticked, a halfway decent character will have at least 2 and a great character has all of them (that's wizards for you).

and that is not all.

imagine that the flavours we have are soft-ice, gelatto and popsicle.

the methods of delivery is coupe, paper cup and waffle

the flavours themselves have different options in them: soft-ice has various dips: nuts (piercing), chocolate sprinkles (slashing) and liquid chocolate (bludgeoning). Gellatt has chocolate(fire), mint (frost), lemon (acid), malaga (thunder), stracciatella (lightning) and vanilla (force/'just magic'). popsicles come in jsut about every fruitinfusion imaginable (you get my point that every condition will have a different fruity flavour.

Then you have the deliveries for which only waffles have real flavour and shape options (wide cone, narrow cone, cone with chocolate, flat between 2 waffles for icexream sandwich).

Now, you'll agree that for the ice-cream man it's easy to hand out mutliple types of delivery and multiple types of flavours. You will need a serious truck to provide every flavour and every delivery (and every subflavour), but to eat them all will give your tummy a bad ache (not to mention diabetes). No, for eating you might prefer gellato and finish it with a popsicle later.

Now consider that you only have so much room to take stuff to hand out and stuff to eat. You will get hungry and the stuff to hand out is strictly for handing out and the stuff for eating is strictly for eating. Now you need to predict how hungry you will get and how much money you can make by handing out. That results in a handout/eat balance, or in this case offense/defense/control balance in DnD.

[xanderh]

My point is this;


Historically, you're right and that's my point. Sharpening some sticks and calling them spears is an effective way to build a half-decent army if you've got enough dudes. Equipping those same guys with armour was expensive and largely ineffective, so wasn't done (as a rule). Whatever the latest defense that got developed, an effective offensive counter was short to follow. Make some chainmail? I'll use a spear instead of an axe. Make some plate armour? I'll use bodkin arrows for my bow. Build a castle? I'll just dig a hole under the wall and collapse it. See the pattern? Expensive and time-consuming defense countered by a cheap and effective offence. It's not a strategic conceit; it's plain facts. Defences are hard and expensive to make effectively, whether on a personal or wider scale and frequently come with a drawback of one kind or another (Want a shield? Can't use a bigger sword. Want plate armour? Sorry it's so heavy. Wear a helmet? Nice and dark in there, right? Build a castle? Not exactly mobile, is it?).

As for fencing and right of way, that's a perfect example of my point; a big guy with a big stick swings at your head. Without the codified training you talk about, which took years to develop and more time to teach, you have to be just as big or have a bigger "stick" to compete. Even with the training, one false step and you're a greasy stain on the floor against big-stick-guy. "Having a big stick" isn't exactly a great offence; there's much more effective tools and methods out there, but it takes an actively good defence to defend against it effectively.

Small digression, but you just showed a complete lack of knowledge of medieval armour and the counters.
Chainmail is not going to stop an axe. It's going to turn it into a large club, but axes were very effective against chainmail, since the chainmail doesn't do much to stop blunt force trauma, of which the axe does plenty.
And bodkin arrows were completely ineffective against plate armour. The massive volume of arrows fired at Agincourt meant that some of the arrows found the tiny gaps in the armour that were present, but a bodkin arrow fired from a warbow was never enough to penetrate plate armour. A good, direct, hit was enough to dent the armour, but not enough to actually penetrate. And even if it did, it still had to get through the padding underneath, and actually penetrate far enough to do any damage to the target. It simply didn't happen. There was one ranged weapon that worked effectively, and that was the arbalest, which was a crossbow designed to penetrate plate armour. It could fire once every 10-15 seconds, if the wielder was very skilled in its use. The melee weapons of choice were warhammers, flanged maces, and pollaxes.

And plate armour isn't as heavy as you make it out to be. An example of a very heavy configuration weighs 60 pounds (most configurations weighed between 40 and 60 pounds), distributed over the body. The lightest load of a modern soldier in the US army is a bit over 60 pounds, most of which is on the torso. A marine has a lightest load of about 90 pounds. It's easy to move in, and doesn't restrict movement to a major degree.

Armour was very effective. The major drawbacks were that it was hot to wear (while not helping against cold), and tired you out faster. Meanwhile, the benefits were that your chances of surviving were increased dramatically. If armour wasn't effective, nobody would wear it. People weren't idiots, if the armour didn't provide meaningful protection, professional soldiers wouldn't be spending money on it.

And your example of the GWM fighter is a pretty bad one. ALL of the options you gave him can be countered by a high AC. Meanwhile, it's possible to get a high AC and respectable DEX, CON and WIS saves without much trouble. The paladin gets charisma bonus to saves, and the fighter gets a lot of ASIs and Indomitable. Both can get heavy armour and a shield, and Shield Master can do wonders against DEX saves. A bear totem barbarian has resistance to ALL damage except psychic when raging. Getting good defenses isn't as hard as you make it out to be. If you're mundane, it can be very hard to get around some defenses, some of which are very easily obtained. Spellcasters have it easier here. They get plenty of different damage types and avenues of attack to deal with defenses, but that can be defeated with a single type of defense: magic resistance. And bear totem barbarians are still laughing at them, since they're most likely not dealing psychic damage.

The fact of the matter is, getting good defenses isn't nearly as hard as you try to make it sound. My eldritch knight has an amazing AC (22 with a cloak of protection, and shield to boost it), alright DEX saves (with shield master), good CON saves, Absorb Elements to help him withstand those avenues of attack, and is getting good WIS saves next level. I've managed to defend myself well against almost all avenues of attack against him, and the remaining weaknesses are uncommon attack vectors. If they are consistently attacked, it breaks the verisimilitude of the world, since they're supposed to be the less attacked defenses. It'll feel cheap, and no DM I've played with attacks the weaknesses I have consistently for that reason.

[MrFahrenheit]

False choice.

The important thing to optimize is control. A tank is of little use if the enemy ignores her, and your DPR won't be impressive if the enemy force surrounds and focus-fires your squishy butt.

Whether you use chokepoints, mobility, range, stealth, spells... ideally all of the above... control the fight well enough, and you don't need to worry about how strong offense and defense are.

Sun Tzu knew his business.

This. It's why Otto's Irresistable Dance is a 6th level spell: your controller just gave the party tanks free AoOs on the BBEG.

[MaxWilson]

Building a character for Offence rather than Defence is almost always superior if only because Defence has one major weakness...NO defence is perfect. No matter how high your AC, your foes, of which there are many, always have a chance of an auto-hit on a natural 20.

AC is kind of a mediocre defense compared to other things like distance and stealth. Nevertheless, a 1/400 chance of a natural 20 shouldn't scare anyone. If you build for defense, of course you are going to impose disadvantage on your foes while you're at it. (Blur, Careful Web, Devil's Sight + Darkness, grapple + push prone, Protection From Evil, etc.)

That "major weakness" isn't.

[Socratov]

AC is kind of a mediocre defense compared to other things like distance and stealth. Nevertheless, a 1/400 chance of a natural 20 shouldn't scare anyone. If you build for defense, of course you are going to impose disadvantage on your foes while you're at it.

That "major weakness" isn't.

Well, it is compared to enough bonus to your saves that you literally can't fail them... (Like, +20 or something). AS is getting an arbitrarily high skill mod so you win every skilltest, ever (for example athlethics). However, attacking against Ac will always have a chance of success.

[MaxWilson]

Well, it is compared to enough bonus to your saves that you literally can't fail them... (Like, +20 or something). AS is getting an arbitrarily high skill mod so you win every skilltest, ever (for example athlethics). However, attacking against Ac will always have a chance of success.

Sorry, I'm not following. It is ____ compared to getting enough bonus that you can't fail saves?

I said AC is mediocre. I think you might be saying that AC is mediocre as a defense compared to never failing a save. Would you mind clarifying?

[Socratov]

Well, it is a major flaw compared to enough bonus to your saves that you literally can't fail them... (Like, +20 or something). AS is getting an arbitrarily high skill mod so you win every skilltest, ever (for example athlethics). However, attacking against Ac will always have a chance of success.
changes mine to clarify

Sorry, I'm not following. It is ____ compared to getting enough bonus that you can't fail saves?

I said AC is mediocre. I think you might be saying that AC is mediocre as a defence compared to never failing a save. Would you mind clarifying?

As a way to defend yourself from stuff it is a lousy defence. AC will never fully and completely defend you from attacks, no matter how high you get it. One will always be able to overcome it through rolling a nat. 20.

AS far as defensive measures go it's pretty much flawed form the get-go.

Saves can no longer critical hit or fail. Crank them up high enough and they will completely, fully and perfectly defend you.

For skills exactly the same: as long as you have your modifier high enough you can auto succeed in defending from said targeting attack. With AC you will never be able to defend. The only thing that defends fully against stuff targeting your AC is Wall of Force (or a similar indestructible wall effect between you and the attacker.

[MaxWilson]

changes mine to clarify

As a way to defend yourself from stuff it is a lousy defence. AC will never fully and completely defend you from attacks, no matter how high you get it. One will always be able to overcome it through rolling a nat. 20.

AS far as defensive measures go it's pretty much flawed form the get-go.

Saves can no longer critical hit or fail. Crank them up high enough and they will completely, fully and perfectly defend you.

For skills exactly the same: as long as you have your modifier high enough you can auto succeed in defending from said targeting attack. With AC you will never be able to defend. The only thing that defends fully against stuff targeting your AC is Wall of Force (or a similar indestructible wall effect between you and the attacker.

AC is somewhat flawed, but the claim that I was responding to was that defense is flawed. AC and defense are not synonymous; a proper defense is a multilayered defense in depth. E.g. EVEN IF the enemy beats your Stealth +19 and spots where you're hiding inside the dark cavern and manages to get within melee range of you despite your high movement (maybe there are a lot of enemies) and makes you fail your Con save against Banshee wail, you still don't die because your buddy cast Death Ward on your earlier, and besides someone can just pop you right back up with Healing Word anyway. Even on your unluckiest day even, you're still probably going to live long enough to kill the enemy before he kills you. Contrast this with a brittle offense-only party who will probably just walk up to the enemy and start hitting them with greatswords; all it takes is one nasty surprise ("that isn't a hobgoblin you're fighting, it's a Medusa in plate armor!") to change the fight in ways you didn't expect, and very possibly make you dead. Defense makes you resilient, and given the way 5E is set up, it doesn't even cost that much.

BTW I'm not aware of any good ways to get +20 to all of your saves on an ongoing basis. +11 or even +15 is attainable, but that's not high enough to be in auto-succeed territory; so comparing saves to AC on the basis of +20 seems a bit pointless. In practice, only weak stuff like drow poison ever becomes auto-resistible. That's not much different from how high-AC tanks (e.g. AC 21 Paladorc with Quickened Blur and Sanctuary up and Shield spell on tap) become functionally immune to attacks from weak enemies like hobgoblins and drow. The difference is one of degree, not kind.

[Easy_Lee]

As far as the DM being able to choose monsters resistant to your damage types, I don't think there are any creatures resistant to both force damage and magic weapons. Blade pact warlocks get strong at-will options for both of those.

[Fable Wright]

So, let me look at the opinions here to try and straighten things out.

MaxWilson is stating that the best defense is by no-selling enemy attacks entirely. Either by means of using stealth to avoid getting attacked, staying out of range of enemy attacks, rendering attacks unlikely to succeed with counterspelling or grapple+prone where crowd control isn't working, and essentially just shutting down enemy actions while you mop up. AC and saves are just one portion of this tactic; the important thing is forcing the enemy to engage your strengths and not letting them engage your weaknesses. Optimizing a proactive defense, he claims, is the most mechanically sound strategy in the game, and optimizing a few reactive defenses (such as a Paladin's AC against Prone enemies) to enhance this strategy will be of the largest marginal benefit to the team. After all, he reasons, when it's the Paladin's turn to use his form of proactive defense (grapple+prone-ing to keep melee enemies away from the caster and unable to damage the tank), those benefits have a larger marginal increase in effectiveness than a 5% offensive boost against all opponents. His +2 AC means that he's essentially incapacitated the enemy with two attacks with his strategy, which is a quicker incapacitation rate than can be easily achieved with damage in 5th edition.

Socratov, I think, is stating that optimizing your reactive defenses, especially AC, is a waste of resources. After all, an enemy can attack your saves, armor class, or through passive Aura effects. When looking at your enemy's possible attacks, they can target your AC, your saving throws, or your HP directly, and they can do so through many means. Optimizing one character to solo against all possible enemies, he reasons, is an exercise in futility. One does not have a high armor class, resistance to all damage, and proficiency in all saving throws. And even if one did, the enemy would just ignore the tank and work on the other party members, and the tank would have spent so many resources on defense that they would be unable to effectively sway the course of the fight. Therefore, investing in a few offensive options that are impossible to completely counter, such as attacks or spells like Wall of Force, will keep you relevant in the fight longer.

xanderh is claiming that a moderate investment of character resources is enough to adequately cover all bases, and that a perfect defense isn't necessary to achieve good results. An Eldritch Knight, for example, can have Resistance to elemental damage types when it matters and high AC when it matters, and can easily obtain a very good set of saves by investing a character resource they're flush with (that is, feats). By investing perhaps a third of his character's resources in defense, he was able to at least double, quite possibly triple or more the amount of time he spent in combat with a hostile that defeated his offensive-based companions in short order, resulting in a much higher contribution to the fight than the teammates who burned out early. Thus, he reasons, while you can't have a perfect defense, your investments there will contribute far more than offensive investments when facing an overwhelming foe.

From these arguments and others in the thread, what I can infer is that in low-teamwork casual games, the decision isn't overly important and offense is far more enjoyable to many, due to the fact that it's always relevant. When challenge scales up, however investments in defense become increasingly important, as monsters' damage capacity in hordes or several levels above the PC greatly outstrips the PCs' offensive scaling. Partially mitigating this is of paramount importance, and the resources to do so are very dependent on class features. A Fighter defends with feats and heavy armor; a Barbarian defends with Rage, Uncanny Reflexes, and a high Constitution; a Paladin defends with healing and saves; spellcasters defend with low-level spells and performing area control on enemies; Rangers and Rogues defend by relying on mobility and evasion. One cannot defend against everything, and shouldn't: instead, with teamwork and area control, defensive fighting styles can be turned into a form of offense when you're capable of funneling enemies to match against PCs most resilient against their tactics.

On the whole, it seems like teamwork > defense > offense, though focusing on one to the exclusion of all others is a very poor plan. It seems to me that, if your PCs' resources are your budget, your offense is equivalent to your free spending; fun to throw around, and something you love to optimize. When life is good, this is what you want. Defense is your health insurance—a drag to pay, but something that you'll be incredibly thankful for when you need it. Teamwork is your finance management skill, making sure that everything goes smoothly in good times and bad; if you have this, you'll probably be fine no matter what role you end up in.

Of course, I'm probably horribly mistaken and ignoring many relevant points of data, but that's my takeaway from the thread right now. Your mileage may vary.

[MaxWilson]

I think that's a fine synopsis as far as I'm concerned. The only statement I'm uncomfortable with is "Optimizing a proactive defense, he claims, is the most mechanically sound strategy in the game." I don't think I would defend that claim as stated. After all, "proactive defense" is ultimately a contradiction in terms; if you're going to be excessively proactive, it will necessarily involve some kind of offense. Waiting until your enemy takes off his armor and goes to bed and then ambushing him with a dozen Mage Armored Invisible Stalkers (without you observing the fight only through a scrying device) is mechanically quite effective, but it's not clear whether it ought to count as "proactive defense"...

I'm comfortable with your other characterizations of my position though, especially the idea that you're trying to force the enemy to engage with your strengths.

[xanderh]

Yeah, you pretty much nailed my points

[Ganders]

I personally lean toward defense, even on caster characters. I'll spend feats on armor proficiency if necessary. But that's a personal issue.

However, there's something that seems to have been left out of the discussion so far. Double-defense is not actually as good as double-offense in many situations.

Consider character A: high offense, kills a mob in 2 rounds, but takes 20 damage during those two rounds.
against character B: high defense, kills a mob in 4 rounds, but takes 20 damage during those four rounds.

These two are NOT equal! Character A can finish off the last mob and then use a healing kit on his ally before he dies, character B might be too late. Character A can finish off the last minion then chase after the fleeing BBEG, but character B might be too late to catch him. Character A can force a very difficult concentration check on the caster that just used Hold Person on the rest of the group, while character B can only force very easy concentration checks. And many other situations can be imagined.

Even if it were completely possible, doubling defense is actually NOT as good as doubling offense, even if theoretically the 'amount of damage done to opponent before you drop' or 'size of fight you can win' is the same either way.

[xanderh]

I personally lean toward defense, even on caster characters. I'll spend feats on armor proficiency if necessary. But that's a personal issue.

However, there's something that seems to have been left out of the discussion so far. Double-defense is not actually as good as double-offense in many situations.

Consider character A: high offense, kills a mob in 2 rounds, but takes 20 damage during those two rounds.
against character B: high defense, kills a mob in 4 rounds, but takes 20 damage during those four rounds.

These two are NOT equal! Character A can finish off the last mob and then use a healing kit on his ally before he dies, character B might be too late. Character A can finish off the last minion then chase after the fleeing BBEG, but character B might be too late to catch him. Character A can force a very difficult concentration check on the caster that just used Hold Person on the rest of the group, while character B can only force very easy concentration checks. And many other situations can be imagined.

Even if it were completely possible, doubling defense is actually NOT as good as doubling offense, even if theoretically the 'amount of damage done to opponent before you drop' or 'size of fight you can win' is the same either way.

Doubling defence is pretty trivial. Against cr 5 creatures, the difference between plate and plate+shield+defence fighting style is exactly a halving of their chances of hitting you, effectively halving their damage. The shield spell further improves this, making them go from a 6/20 chance to a 1/20 to hit you. Give them disadvantage, and it's 1/400,which is also trivial to get as an Eldritch Knight. And absorb elements is exactly equal to halving the damage taken from elemental damage. Add in some feats for save proficiencies, and you have more than doubled your defence against saving throws.

Meanwhile, doubling your damage dealing is going to be significantly harder. GWM isn't going to double it on its own. Increasing strength doesn't count in my opinion because that improves offense, defence (strength saves, defence against control), and controlling (grapple, shove). No matter what aspect you focus on, you're going to increase your strength. It's going to take a lot of effort to double your damage output with conscious choices where you get to choose and don't just get it because of class features. Meanwhile, the defence fighting style alone is going to improve your defence by 1/6th.

Comparing shield+longsword+defence fighting style to greatsword with GWS gives us the following:

Defensive:
Defence: 3/20
Offence: 4.5 plus strength

Offensive:
Defence: 6/20
Offence: 8.33 plus strength.

One could also go for a balance.
Defensive balanced (shield, longsword, dueling):
Defence: 4/20
Offense: 6.5 plus strength

Offensive balanced (greatsword, defence FS):
Defence: 5/20
Offense: 7 plus strength.

Comparing them, using the lowest numbers as the baseline gives us this (adding the percentage increases):

Baseline
Defense: 6/20
Offence: 4.5

Defensive:
Defense: +100 %
Offense: +0 %
Total: +100 %

Offensive:
Defense: +0 %
Offense: +85 %
Total: +85 %

Balanced defensive:
Defense: +66 %
Offense: +44 %
Total: +100 %

Balanced offensive:
Defense: +33 %
Offense: +55 %
Total: +88 %

In general, we can see that the sword&board defensive builds have a greater overall increase. And that's just from fighting style and gear. If we just consider the gear, we get a +66 % defense for the defensive build, versus a +55 % offense for the offensive build.
And once we start to consider other features, defense becomes even easier to boost. Plenty of feats that boost defense somehow, yet not many that boost offense. Of note are the -5/+10 feats, as well as polearm master. I'm going to hazard a guess that they're probably not going to double the damage output, especially at later levels as a fighter (one extra attack per turn is less impactful the more attacks you have, though a paladin is going to benefit from this with his nova). Combined, they might just achieve a doubling of damage output as a paladin against the right enemies. Meanwhile, as an Eldritch Knight, picking Shield and Absorb Elements is going to tremendously improve your defenses, as is Blur once you get that far. And if you combine them, you go from about a 3/20 chance to hit, to a 1/400 chance. That's a 5900 % increase in defense against attack rolls. In terms of saving throws, feats can grant you proficiency if you don't already have it. Shield Master can completely nullify the damage of a DEX saving throw if you succeed, which is pretty major. And if you fail the save, you've still got Absorb Elements which is going to increase your defense by 100 %. The fallback is a 100 % increase in defense, with the hope being a complete nullifying of that attack. Admittedly, you have to choose each turn whether to massively boost defense against attack rolls or saving throws, but you get to make that choice when you're actually hit by one of them. And having blur up, and spending the reaction for save defense is an effective balance.

[Socratov]

So, let me look at the opinions here to try and straighten things out.

MaxWilson is stating that the best defense is by no-selling enemy attacks entirely. Either by means of using stealth to avoid getting attacked, staying out of range of enemy attacks, rendering attacks unlikely to succeed with counterspelling or grapple+prone where crowd control isn't working, and essentially just shutting down enemy actions while you mop up. AC and saves are just one portion of this tactic; the important thing is forcing the enemy to engage your strengths and not letting them engage your weaknesses. Optimizing a proactive defense, he claims, is the most mechanically sound strategy in the game, and optimizing a few reactive defenses (such as a Paladin's AC against Prone enemies) to enhance this strategy will be of the largest marginal benefit to the team. After all, he reasons, when it's the Paladin's turn to use his form of proactive defense (grapple+prone-ing to keep melee enemies away from the caster and unable to damage the tank), those benefits have a larger marginal increase in effectiveness than a 5% offensive boost against all opponents. His +2 AC means that he's essentially incapacitated the enemy with two attacks with his strategy, which is a quicker incapacitation rate than can be easily achieved with damage in 5th edition.

Socratov, I think, is stating that optimizing your reactive defenses, especially AC, is a waste of resources. After all, an enemy can attack your saves, armor class, or through passive Aura effects. When looking at your enemy's possible attacks, they can target your AC, your saving throws, or your HP directly, and they can do so through many means. Optimizing one character to solo against all possible enemies, he reasons, is an exercise in futility. One does not have a high armor class, resistance to all damage, and proficiency in all saving throws. And even if one did, the enemy would just ignore the tank and work on the other party members, and the tank would have spent so many resources on defense that they would be unable to effectively sway the course of the fight. Therefore, investing in a few offensive options that are impossible to completely counter, such as attacks or spells like Wall of Force, will keep you relevant in the fight longer.

xanderh is claiming that a moderate investment of character resources is enough to adequately cover all bases, and that a perfect defense isn't necessary to achieve good results. An Eldritch Knight, for example, can have Resistance to elemental damage types when it matters and high AC when it matters, and can easily obtain a very good set of saves by investing a character resource they're flush with (that is, feats). By investing perhaps a third of his character's resources in defense, he was able to at least double, quite possibly triple or more the amount of time he spent in combat with a hostile that defeated his offensive-based companions in short order, resulting in a much higher contribution to the fight than the teammates who burned out early. Thus, he reasons, while you can't have a perfect defense, your investments there will contribute far more than offensive investments when facing an overwhelming foe.

From these arguments and others in the thread, what I can infer is that in low-teamwork casual games, the decision isn't overly important and offense is far more enjoyable to many, due to the fact that it's always relevant. When challenge scales up, however investments in defense become increasingly important, as monsters' damage capacity in hordes or several levels above the PC greatly outstrips the PCs' offensive scaling. Partially mitigating this is of paramount importance, and the resources to do so are very dependent on class features. A Fighter defends with feats and heavy armor; a Barbarian defends with Rage, Uncanny Reflexes, and a high Constitution; a Paladin defends with healing and saves; spellcasters defend with low-level spells and performing area control on enemies; Rangers and Rogues defend by relying on mobility and evasion. One cannot defend against everything, and shouldn't: instead, with teamwork and area control, defensive fighting styles can be turned into a form of offense when you're capable of funneling enemies to match against PCs most resilient against their tactics.

On the whole, it seems like teamwork > defense > offense, though focusing on one to the exclusion of all others is a very poor plan. It seems to me that, if your PCs' resources are your budget, your offense is equivalent to your free spending; fun to throw around, and something you love to optimize. When life is good, this is what you want. Defense is your health insurance—a drag to pay, but something that you'll be incredibly thankful for when you need it. Teamwork is your finance management skill, making sure that everything goes smoothly in good times and bad; if you have this, you'll probably be fine no matter what role you end up in.

Of course, I'm probably horribly mistaken and ignoring many relevant points of data, but that's my takeaway from the thread right now. Your mileage may vary.
(Emphasis mine)

Yup, you captured what I said pretty well.

However, in your own analysis there is still one flaw: Offense might not be insurance, but it sure as hell covers mandatory purchases like groceries and gas for you car, etc.

In a fight you can prolong your life with defense, but not indefinitely. At some point your defence will falter or start degrading. I can say from first hand in-game experience that you want to make sure the other one snuffs it as it were before you reach that point. then there is the binary nature of defense and the continuous nature of offense. Depending on your preferred method of offense it can be better or worse, but have an effect regardless (a cantrip at lvl 4 comes to mind: it does not deal very much damage, but it deals damage regardless). At a certain point it's not going to matter much in terms of cost to increase your offense by X points damage or X points expected value to hit.

For defence, it either works or it doesn't. So, by its discrete nature you can't grade it on an continuous scale (like percentages), unless you can, with any certainty, predict what your expected value of encountered offense will be. it is, however very useful for establishing baselines. For that you can make an assumption that if you are going to face certain levels of challenges, you can at least defend to an expected value of X attacks of that sort.


You could even weigh these against classfeatures like ASI increase and increases in proficiency. these points can be an indicator of having to upgrade capabilities in either category. From here on it's only evident to find the economically optimal solution.

I have also seen that at low levels defense is easier to obtain then offense, but at high levels offense will have an easier time to trump defence as defence tapers off in terms of effectiveness/effort. .

[Giant2005]

I think that in general, people seem to be overestimating the offensive differences between an offensive and defensive build.
A defensive build isn't going to neglect their offensive potential entirely - in fact that isn't actually something that is very easy to do in 5e. A defensive build is still going to have both level 5 and 11 damage upgrades.
A defensive build will have about 80-90% of an offensive build's offensive potential, while having 200-300% of its defensive potential.

[Socratov]

I think that in general, people seem to be overestimating the offensive differences between an offensive and defensive build.
A defensive build isn't going to neglect their offensive potential entirely - in fact that isn't actually something that is very easy to do in 5e. A defensive build is still going to have both level 5 and 11 damage upgrades.
A defensive build will have about 80-90% of an offensive build's offensive potential, while having 200-300% of its defensive potential.

I think 80~90% is an overestimation. I also think that an offensive build won't neccessarily neglect defence.

for example, a sorcerer could easily reach 16 AC at lvl 1 through mage armour/draconic ancestry and a decent dex. As a non front line combatant that is fine. Con proficiency and a decent mod does the rest for concentratin on spells and hte nastier effects.

later on this will become harder as 16 Ac won't cut it anymore, that's when you use mirror image and blink to further increase your defensive potential. This is meanwhile a class that is fairly combat oriented thourhg spells like Chromatic Orb and the like. It can also control pretty well (especially when using Heighten at lvl 3) with sleep/Tasha's hideous laughter and its ilk. Besides, offense has it's defensive effects as well: a dead/controlled enemy won't have the capabilities to use his offense to target your defence.

anecdotally that's how I, as a lvl 2 druid, won a fight wihtout receiving damage: I entangled 3 out of 4 enemies so my mates could burst them down one by one (LMoP, combat versus 2 goblins, a wolf and a bugbear). So, by casting 1 offensive spell I pretty much defended my party from 3 attacks per turn for about 3 turns. that is 1 offensive action defending against 9 attacks. Assuming encountering an enemy that wil on average roll to hit of 15 (14, 14, 16 for a +2 dex, +2 dex +4 str on +2 proficiency enemies, assuming an average roll of 10), that's pretty much giving my party a 15 AC, regardless of their defensive potential.

This kind of effect should be taken into account when calculating offensive/defensive potentials. The capability to inflict pro-active defensive measures on enemies can ahve a great effect on defence in general for the party. You could even make the same argument for healing by calculating the expected value for damage, having prevented the damage by pro-active 'healing' (which incidentally makes for a great House-like character concept)

[Giant2005]

I think 80~90% is an overestimation. I also think that an offensive build won't neccessarily neglect defence.

for example, a sorcerer could easily reach 16 AC at lvl 1 through mage armour/draconic ancestry and a decent dex. As a non front line combatant that is fine. Con proficiency and a decent mod does the rest for concentratin on spells and hte nastier effects.

later on this will become harder as 16 Ac won't cut it anymore, that's when you use mirror image and blink to further increase your defensive potential. This is meanwhile a class that is fairly combat oriented thourhg spells like Chromatic Orb and the like. It can also control pretty well (especially when using Heighten at lvl 3) with sleep/Tasha's hideous laughter and its ilk. Besides, offense has it's defensive effects as well: a dead/controlled enemy won't have the capabilities to use his offense to target your defence.

anecdotally that's how I, as a lvl 2 druid, won a fight wihtout receiving damage: I entangled 3 out of 4 enemies so my mates could burst them down one by one (LMoP, combat versus 2 goblins, a wolf and a bugbear). So, by casting 1 offensive spell I pretty much defended my party from 3 attacks per turn for about 3 turns. that is 1 offensive action defending against 9 attacks. Assuming encountering an enemy that wil on average roll to hit of 15 (14, 14, 16 for a +2 dex, +2 dex +4 str on +2 proficiency enemies, assuming an average roll of 10), that's pretty much giving my party a 15 AC, regardless of their defensive potential.

This kind of effect should be taken into account when calculating offensive/defensive potentials. The capability to inflict pro-active defensive measures on enemies can ahve a great effect on defence in general for the party. You could even make the same argument for healing by calculating the expected value for damage, having prevented the damage by pro-active 'healing' (which incidentally makes for a great House-like character concept)

I think we might have different ideas of what offense and defense means, and what exactly it means to build for either one.
To me, building defensively doesn't mean devoting more than half of your resources to defense (although it sometimes can, in impractical builds designed purely for theory). Building defensively is taking a single level of Fighter or Cleric on that Sorcerer, in order to increase his AC to 20 or 21. It is a fairly minor price to pay for significantly higher defenses.
Your story about Entangle just further illustrates this being a difference in definition. You called it an offensive spell, but I would call it a defensive spell. It is a spell that inflicts no damage - its only purpose is to be used exactly how you used it: preventing incoming damage.
As for things like Mirror Image, I don't really rate them too highly. A character that gains their defense passively is going to fare better than one that achieves arguably similar defense actively. That turn wasted casting active defenses means that the offensively built character doesn't have much hope in meeting the offensive potential of the character that was built for passive defenses. Most fights only last for around 3 rounds - if you spend one of them fixing your terrible passive defense, then your offense is always going to pale compared to the guy that had enough passive defenses to be able to use all 3 rounds offensively. That will remain true no matter how much you build for offense - that offensive focus is not anywhere near enough to make up the difference that turn will bring.
However, there are defensive spells that render that point moot. Some exceptional builds combine powerful defense and offense via spells like Armor of Agathys. Casting AoA to shoe up some weak defense isn't a wasted action like Mirror Image, because it has enough offensive potential built in to it to make up for the loss of the action. Although, I expect that while I call it a defensive spell, you probably call it an offensive one due to our opinions really only differing by a matter of perspective and scale.

[Socratov]

I think we might have different ideas of what offense and defense means, and what exactly it means to build for either one.
To me, building defensively doesn't mean devoting more than half of your resources to defense (although it sometimes can, in impractical builds designed purely for theory). Building defensively is taking a single level of Fighter or Cleric on that Sorcerer, in order to increase his AC to 20 or 21. It is a fairly minor price to pay for significantly higher defenses.
Your story about Entangle just further illustrates this being a difference in definition. You called it an offensive spell, but I would call it a defensive spell. It is a spell that inflicts no damage - its only purpose is to be used exactly how you used it: preventing incoming damage.
As for things like Mirror Image, I don't really rate them too highly. A character that gains their defense passively is going to fare better than one that achieves arguably similar defense actively. That turn wasted casting active defenses means that the offensively built character doesn't have much hope in meeting the offensive potential of the character that was built for passive defenses. Most fights only last for around 3 rounds - if you spend one of them fixing your terrible passive defense, then your offense is always going to pale compared to the guy that had enough passive defenses to be able to use all 3 rounds offensively. That will remain true no matter how much you build for offense - that offensive focus is not anywhere near enough to make up the difference that turn will bring.
However, there are defensive spells that render that point moot. Some exceptional builds combine powerful defense and offense via spells like Armor of Agathys. Casting AoA to shoe up some weak defense isn't a wasted action like Mirror Image, because it has enough offensive potential built in to it to make up for the loss of the action. Although, I expect that while I call it a defensive spell, you probably call it an offensive one due to our opinions really only differing by a matter of perspective and scale.

good point. I deem an offensive action any action that does onto others, while primarily defensive options do onto the object to defend.

entangle is now an offesive spell. Otiluke's Resillient Sphere is a defensive spell. leomund's Tiny Hut is a defensive spell (but can give offensive opportunities) (http://www.giantitp.com/forums/showthread.php?489221-Leomund-s-tiny-fort)

[Fable Wright]

I think that's a fine synopsis as far as I'm concerned. The only statement I'm uncomfortable with is "Optimizing a proactive defense, he claims, is the most mechanically sound strategy in the game." I don't think I would defend that claim as stated. After all, "proactive defense" is ultimately a contradiction in terms; if you're going to be excessively proactive, it will necessarily involve some kind of offense. Waiting until your enemy takes off his armor and goes to bed and then ambushing him with a dozen Mage Armored Invisible Stalkers (without you observing the fight only through a scrying device) is mechanically quite effective, but it's not clear whether it ought to count as "proactive defense"...

I'm comfortable with your other characterizations of my position though, especially the idea that you're trying to force the enemy to engage with your strengths.

Apologies. I used the term 'proactive defense' to try and distinguish it from optimizing 'reactive defenses', like AC and Resistance. A proactive defense would be forcing enemies to engage against your strengths and using battlefield control to prevent conflict from ever occurring, as opposed to passively waiting for them to attack you. My apologies for the confusing wording.


anecdotally that's how I, as a lvl 2 druid, won a fight wihtout receiving damage: I entangled 3 out of 4 enemies so my mates could burst them down one by one (LMoP, combat versus 2 goblins, a wolf and a bugbear). So, by casting 1 offensive spell I pretty much defended my party from 3 attacks per turn for about 3 turns. that is 1 offensive action defending against 9 attacks. Assuming encountering an enemy that wil on average roll to hit of 15 (14, 14, 16 for a +2 dex, +2 dex +4 str on +2 proficiency enemies, assuming an average roll of 10), that's pretty much giving my party a 15 AC, regardless of their defensive potential.

As mentioned... MaxWilson, Giant2005 and I are counting Entangle as a defensive (proactively defensive, in my case) spell. You used a spell to render 3 enemies unable to deal damage using one spell slot, a much more effective use of resources than using one spell to case Thunderclap and moderately inconvenience two enemies. In this case, we're in agreement: The (proactively) defensive option is by far the most effective use of resources here.

Likewise, Wall of Force, Resilient Sphere, and other battlefield control elements that keep enemies from dealing damage to you are qualified by us as defensive. We're in agreement, at least, that these at least are the things that are most worth investing in for combat.


However, in your own analysis there is still one flaw: Offense might not be insurance, but it sure as hell covers mandatory purchases like groceries and gas for you car, etc.

Indeed it does! I chose this analogy because it's clear that offense should not be neglected. However, all classes have some feature (Extra Attack, ASI to max attack stat, cantrips or blasting spells) that allow them to contribute to contribute minimally to cover mandatory purchases. Of course you're going to invest some in this. In fact, I only ever advocated for spending around 1/3 of character resources on defense; the other 2/3 are divided between utility and offense, perhaps going entirely into the latter.

[MaxWilson]

Indeed it does! I chose this analogy because it's clear that offense should not be neglected. However, all classes have some feature (Extra Attack, ASI to max attack stat, cantrips or blasting spells) that allow them to contribute to contribute minimally to cover mandatory purchases. Of course you're going to invest some in this. In fact, I only ever advocated for spending around 1/3 of character resources on defense; the other 2/3 are divided between utility and offense, perhaps going entirely into the latter.

Yep, that's a pretty good split. The things that raise my eyebrows are when I see people trying to go 100% offense in order to marginally raise their DPR by a small fraction; but a balanced approach tends to work out pretty well. For example, on an Eldritch Knight 20, you might see me recommend Sharpshooter (offense) and Dex 20 (offense and some defense) and Mobile and Resilient/Wisdom (Defense) and Lucky (both offense and defense and utility) and Find Familiar (utility and recon, which you might classify as proactive defense) and Magic Weapon (offense) and Shield and Absorb Elements (defense) and Animate Dead (mix of offense and defense and control). That will do only slightly less damage than a purely offensive build which trades out defensive capabilities for marginal-effective offensive enhancements like e.g. Magic Initiate (Hex) and Savage Attacks.

It just isn't that hard to get offense to the point of diminishing returns.

==============================

@Socratov, BTW, if you like Entangle, you will love Careful Web (Sorcerer 3). Now you can fight from inside the web so that even enemies who save against the spell on round one have to make fresh saves every single turn; while you and your party are immune because you automatically make your saving throws.

As a bonus, it provides light obscurement so a Skulker Rogue can Hide inside the web every turn if he wants. That may or may not be relevant in a particular party but I find Skulker a pretty good feat overall anyway, since it lets you hide from monsters with Darkvision in the dark, so it's not too unlikely that you might find yourself paired with a Skulker at some point and be able to use this aspect of the spell.

[Socratov]

[snipperino]

@Socratov, BTW, if you like Entangle, you will love Careful Web (Sorcerer 3). Now you can fight from inside the web so that even enemies who save against the spell on round one have to make fresh saves every single turn; while you and your party are immune because you automatically make your saving throws.

As a bonus, it provides light obscurement so a Skulker Rogue can Hide inside the web every turn if he wants. That may or may not be relevant in a particular party but I find Skulker a pretty good feat overall anyway, since it lets you hide from monsters with Darkvision in the dark, so it's not too unlikely that you might find yourself paired with a Skulker at some point and be able to use this aspect of the spell.

Oooh, good idea. I will definitely remember that.

On the other hand, I'm currently enjoying the finer aspects of Mold Earth and message. The former is fantastic is you have a minute's notice to prepare an earthen fortress. Or to secure a hobbit hole as shelter. The latter is fantastic to make sure you can privately talk with someone or even up to torture (they will hear it, no save, no ifs, no buts, no ignoring) but use of annoying messages, thoughts and mental sounds. The fun part, when cast subtle nobody knows you actually did it but the target. Great is you want to torture but you have a goody two shoes next to you.