AvatarVecna
2019-01-29, 11:10 AM
"Tanking" as a strategy is a very odd proposal in D&D-style games: the ideal tank is a character with absolutely no damage output who is basically impossible to take out of the fight, and yet who gets hardcore focused by enemies despite everything I just said, and while the first point isn't difficult, the second point is much trickier to pull off. Let's start with an example encounter:
Say for the sake of argument we have a four-man party consisting of a Cleric, a Fighter, a Rogue, and a Wizard, all of whom (for some reason) are specialized in melee DPS (the wizard using spell slots for some kind of cool Arcane Strike or something to justify his DPR below). They come across a brutish monster like a giant or something that is also focused on melee DPS, and the giant loses initiative. Here are the Hit Points (HP) and Damage Per Round (DPR) of everybody involved:
Cleric: 12 HP, 1 DPR
Fighter: 20 HP, 3 DPR
Rogue: 8 HP, 5 DPR
Wizard: 4 HP, 2 DPR
Giant: 52 HP, 4 DPR
It takes 11 rounds for the Giant to kill the party regardless of the order in which he attacks them and whether they're fighting back or not. However, if the Giant isn't attacking party members, it takes the party only 5 rounds to kill the giant - the giant's job is to slow that down via targeting enough that he can survive and win. And because of how D&D works, where people's DPR isn't taken out of the accounting until they're at 0 HP, so the giant can't just switch targets halfway through, he needs to commit to killing one until it's dead then move on to the next one. In what order should the giant attack targets to win the fight? There are 24 different ways to go about it - which of the four to target first, which of the three remaining to target second, and which of the remaining two to target third, and what general approach works out the best?
As it turns out (https://docs.google.com/spreadsheets/d/1VkiFDHZdywzZDZCkwk6Ghktnf-WVFT0OO5YLe2jMIFQ/edit?usp=sharing), there's only one order that leaves the giant alive at the end of the 11th round with four dead adventurers: Rogue, Wizard, Fighter, Cleric; the next best (Wizard, Rogue, Fighter, Cleric, almost identical) has the Giant at 1 HP at the end of the 10th round...with the cleric still alive, and still dangerous, and killing him before the giant gets his 11th turn that he would've ended the cleric with. But that's not the order in which they're the most dangerous (Rogue, Fighter, Wizard, Cleric) nor the order by which they're the easiest to take out (Wizard, Rogue, Cleric, Fighter), so why that order?
After playing around with the math a bit in a few other tabs (since deleted), the conclusion I've come to isn't all that surprising: the ideal order is calculate the ratio between DPR and HP, and go from highest to lowest - in essence, targeting people by the ratio between how dangerous they are and how easy their damage is to take out of the equation. The ideal order to target people in is "start with the glass cannons, and end with the tanks", which makes sense on that side of the screen but is frustrating for the player playing the tank (in this case, the ideal tank is the cleric, who is the worst person to target first and the best person to target last).
The above example (while simple) does a good job illustrating the tanking problem: the ideal tank doesn't get targeted unless something besides DPR/HP are being taken into consideration, but most things that would screw around with targeting outside of that move you away from the "tanking" fluff you're probably wanting. However, the above example also has a couple things going on tactically that make things easier for the giant: it can only take out the part in ideal order if it knows their stats and always has the option to attack any living adventurer. Damaging those assumptions is one way of going about tanking effectively.
One strategy for "tanking", to talk in terms of the simplified example from before, is to be both the most difficult to take down and to have the highest DPR/HP ratio - however, this almost certainly means that while your HP is incredible, your DPR is going to be far and away broken by the standards of the table, and you will be outperforming your allies on offense so much that any encounter appropriate for them gets wiped by you, while any encounter appropriate for you wipes them. The enemy's primary goal will be shutting down your offense whatever the cost, because if you get to attack, they lose; this will often involve conditions or BFC as much as pure DPR on their part, but regardless of how the enemy goes about helping you have fun as a tank, the rest of the party isn't gonna be having as much fun.
This is gonna be your Hulking Hurlers, your d2 Crusaders, your astrally-projecting Mailmen, and so on. Such characters are only viable tanks in parties they vastly outclass, and that's not good for group morale.
While the OPM is my least recommended tank for party enjoyment, I can't deny that it's certainly the most effective way to tank...which leaves the Mindbender as the worst method of tanking IMO, even if it's slightly more fun a tank to have in the party if you're not the one playing the tank. The mindbender, in essence, tanks by taking away the enemy's ability to pick their own targets, forcing the enemy to target the Mindbender. It doesn't matter that you're the worst person for them to attack, they have to attack you now, because you said so. The main problem with the Mindbender is that, because abilities that let you dictate enemy action are among the most effective tactics in the game, there's a lot of options available for resisting it, or even shutting it down completely: 99.9% of methods you might use for dictating an enemy's actions are fully shut down by five of the fifteen creature types, and that's without touching any other methods of shutting out mind-affecting effects in general or fear in specific. And more often than not, the method by which you're screwing with the enemy's action economy is limited in other ways just as part of how the mental attack works. But more important than the commonality of ways to get around this style of tanking is how unfair it seems on the other side of the screen; because you're dictating that enemies attack the worst person for them to attack, you're making yourself quite frustrating for a DM trying to provide challenges, and will start including enemies that can take you out more quickly, enemies that can shut down your Mindbender abiltiies, enemies that are immune to your Mindbender abilities, or some combination of the three.
This is your beguilers, your enchanter wizards, your Knights, and so on. This style of tanking is the closest to the "ideal tank" I discussed at the start, but is the most vulnerable to mechanical limitations, out-of-game metagaming, and DM frustration.
One method of tanking without giving up and being a caster is a very tricky way of tanking: a tank that pretends as hard as they can that they're a glass cannon, even though they very much aren't. To reference the earlier example, if the Cleric or the Fighter were capable of pretending to be the Rogue or the Wizard, this could result in informed/intelligent enemies targeting the Cleric or Fighter by mistake. While such builds can theoretically be effective tanks, in practice they only really work extremely short-term because of gradually-increasing information on the party in-game between fights (or even between rounds in the same fight). In-character, a PC that looks dangerous and frail might initially get swarmed, but when they don't go down quickly and don't really pour out the damage, even kinda dumb enemies will realize they're not a threat and move on to more important targets; at best, you'll maybe waste two rounds a fight with this, assuming that no enemies learn of your tricky ways ahead of time. But speaking of that, we need to look at this out-of-character: after a few levels of this schtick every combat, your DM is going to possibly-not-even-intentionally convince himself that the enemy wouldn't target the Oscar Nominee for some reason, and your schtick stops working as often and your party members are getting attacked more and now you're getting a D minus on your Tanking Report Card where you used to be a straight A student.
This is gonna be your Dwarven Defenders, your VoP monks, your illusionists, and so on. Such characters are theoretically viable tanks but become exponentially less effective the more information the enemy has to work with.
One method of tanking is rather...antithetical to the normal ideas about what a "tank" looks like, is the Enabler. This "tank" focuses on improving the party's ability to fight; every action this person takes benefits the party, makes the party harder to take down or more dangerous to the enemy. The Enabler sees that the most effective way for the enemy to behave is to target based on the DPR/HP ratio, and spends their turns screwing with the ratio by increasing either side (or both) on whomever they please. They act as a source of chaos to any enemy calculations, and every extra turn they get to act is another turn they'll get to screw with the math. It could be reasonably argued that constantly making your allies more tanky/dangerous increases your own effective DPR every round, making you the best target by the assumed math; it could also be reasonably argued that taking you out, even if you're possibly not the best target by the math, makes the math a whole lot simpler to perform round-to-round. This method is probably the second-most effective but is also not very...classically tanky.
This is your buffers, your healers, your summoners, your Apostle Of Peaces, and so on. This particular style of "tanking" is typically the domain of casters who tend to be closer to glass cannons than tanks by the math, which might make this style of tanking less ideal for you.
This next tanking method looks at the given example encounter and concludes that the best way to tank is to cut off the giant's access to targets; it doesn't matter that the cleric is the worst person to target if they're the only person to target. Typically, this build focuses on locking down targets, either by locking down a significant area to keep a horde busy, or going one-on-one with the big baddie to keep them occupied, usually in some manner related to the AoO subsystem. There's some builds that will overlap with the next method, which is even better.
This method is ruled by reach monster builds, Jack B Quick style builds, builds focused around trips or bull rushes, and so on.
The final tanking method is a simple idea that lends itself to interesting tactics: if the problem with the default tank is that high defense/low offense means not getting targeted (or tricking/forcing people into targeting you), and high defense/super-high offense means ruining everybody else's fun, what if you play somebody with low defense/low offense who can spend resource to make either of those a force to be reckoned with. That is to say, what if you play a character who could be a glass cannon or a tank from moment to moment? The goal of this is a character who is difficult to deal with when they're defending, but if you ignore them will turn into your biggest problem. These characters tank by making themselves your problem, because if you neglect them you will instantly regret it; they might not be the best person to target when they're on defense, but you can't afford not to attack them.
This is the purview of higher-op monks, defense-focused rogues, paladins, and uberchargers - especially Shock Trooper uberchargers.
The knight's problem isn't that it's unaware of these methods of tanking, the knight's problem is that it's trying to do all of them.
Knight is a mounted ubercharger, the easiest kind of ubercharger to shut down and the hardest to maintain past the low levels: mounted charging damage will have a tough time keeping up once it has to start competing with iteratives, most any mount you're using is going to be unusable in more cramped spaces and most easily tripped up or prevented from charging even in more open areas, and to top it off, you don't have any feature granting you a super-mount so past like level 6, even with Mounted Combat, even with Improved Shield Ally, your horses and whatnot will be dying like flies. Sure, the paladin also has a lot of these problems, but it also has smiting for non-mounted ubercharging, and it at least gets a super-mount...but even in a game where you always get to ubercharge, your damage isn't good enough to be a One Punch Man contender. Even in an ideal situation, you aren't good enough at this to ruin people's fun, which means that one of the other members of the party probably has a higher DPR/HP ratio than you.
Knight's got an okay-threatened area with its ideal style and a few class features synergizing with that, but doesn't have the feats to expand its threatened area or take make good use of AoOs beyond a piddly bit of damage. Doesn't have enough feats to be a reach monster, doesn't have enough feats to be Jack B Quick, doesn't have enough feats for fun ToB addons. The knight designers know that a common tactic used against "armored warrior types" is mind-screw stuff, so it gets a good Will save in exchange for giving up a good Fort save...but every main "tank" class handles this problem better: Barbarian gets rage bonuses to will saves, Paladin gets Cha to saves, Monks just have all good saves by default (and an extra bonus against enchantment effects), and Fighters have an ACF letting them use an immediate action to exchange BAB for Will save bonus.
The knight gets one ability for taking control of enemy actions, and it's basically an ability they can realistically be used once a fight, gives a Will save that will lag behind equal-level caster saves from the second you get it, doesn't affect creatures with too low Int, doesn't affect creatures that are too much weaker than you (meaning it's only getting used on creatures that might have a good chance to make the save), can't be used more than once on the same creature per day, and ends on any creature that gets attacked by somebody other than you. Are there upsides to this ability, of course! It takes a swift action, affects a 100 ft radius, is Ex so it'll work in antimagic environments, and isn't explicitly mind-affecting. But the upsides don't balance out the downsides enough to make this a solid tanking ability on its own.
And all of this is kinda clear in the example builds of Person_Man's knight handbook: all of the builds end up focusing on one of the knight's tanking attempts and focuses on making it good (even to the point of basically not being Knight builds but rather other builds with Knight tacked on), but you can't use everything; if you try, you either fail, or you become the One Punch Man and ruin the fun.
TL;DR the knight is designed with an understanding of what can make a good tank, but tries to touch on a bunch of them and ends up being awful at all of them by default.
Say for the sake of argument we have a four-man party consisting of a Cleric, a Fighter, a Rogue, and a Wizard, all of whom (for some reason) are specialized in melee DPS (the wizard using spell slots for some kind of cool Arcane Strike or something to justify his DPR below). They come across a brutish monster like a giant or something that is also focused on melee DPS, and the giant loses initiative. Here are the Hit Points (HP) and Damage Per Round (DPR) of everybody involved:
Cleric: 12 HP, 1 DPR
Fighter: 20 HP, 3 DPR
Rogue: 8 HP, 5 DPR
Wizard: 4 HP, 2 DPR
Giant: 52 HP, 4 DPR
It takes 11 rounds for the Giant to kill the party regardless of the order in which he attacks them and whether they're fighting back or not. However, if the Giant isn't attacking party members, it takes the party only 5 rounds to kill the giant - the giant's job is to slow that down via targeting enough that he can survive and win. And because of how D&D works, where people's DPR isn't taken out of the accounting until they're at 0 HP, so the giant can't just switch targets halfway through, he needs to commit to killing one until it's dead then move on to the next one. In what order should the giant attack targets to win the fight? There are 24 different ways to go about it - which of the four to target first, which of the three remaining to target second, and which of the remaining two to target third, and what general approach works out the best?
As it turns out (https://docs.google.com/spreadsheets/d/1VkiFDHZdywzZDZCkwk6Ghktnf-WVFT0OO5YLe2jMIFQ/edit?usp=sharing), there's only one order that leaves the giant alive at the end of the 11th round with four dead adventurers: Rogue, Wizard, Fighter, Cleric; the next best (Wizard, Rogue, Fighter, Cleric, almost identical) has the Giant at 1 HP at the end of the 10th round...with the cleric still alive, and still dangerous, and killing him before the giant gets his 11th turn that he would've ended the cleric with. But that's not the order in which they're the most dangerous (Rogue, Fighter, Wizard, Cleric) nor the order by which they're the easiest to take out (Wizard, Rogue, Cleric, Fighter), so why that order?
After playing around with the math a bit in a few other tabs (since deleted), the conclusion I've come to isn't all that surprising: the ideal order is calculate the ratio between DPR and HP, and go from highest to lowest - in essence, targeting people by the ratio between how dangerous they are and how easy their damage is to take out of the equation. The ideal order to target people in is "start with the glass cannons, and end with the tanks", which makes sense on that side of the screen but is frustrating for the player playing the tank (in this case, the ideal tank is the cleric, who is the worst person to target first and the best person to target last).
The above example (while simple) does a good job illustrating the tanking problem: the ideal tank doesn't get targeted unless something besides DPR/HP are being taken into consideration, but most things that would screw around with targeting outside of that move you away from the "tanking" fluff you're probably wanting. However, the above example also has a couple things going on tactically that make things easier for the giant: it can only take out the part in ideal order if it knows their stats and always has the option to attack any living adventurer. Damaging those assumptions is one way of going about tanking effectively.
One strategy for "tanking", to talk in terms of the simplified example from before, is to be both the most difficult to take down and to have the highest DPR/HP ratio - however, this almost certainly means that while your HP is incredible, your DPR is going to be far and away broken by the standards of the table, and you will be outperforming your allies on offense so much that any encounter appropriate for them gets wiped by you, while any encounter appropriate for you wipes them. The enemy's primary goal will be shutting down your offense whatever the cost, because if you get to attack, they lose; this will often involve conditions or BFC as much as pure DPR on their part, but regardless of how the enemy goes about helping you have fun as a tank, the rest of the party isn't gonna be having as much fun.
This is gonna be your Hulking Hurlers, your d2 Crusaders, your astrally-projecting Mailmen, and so on. Such characters are only viable tanks in parties they vastly outclass, and that's not good for group morale.
While the OPM is my least recommended tank for party enjoyment, I can't deny that it's certainly the most effective way to tank...which leaves the Mindbender as the worst method of tanking IMO, even if it's slightly more fun a tank to have in the party if you're not the one playing the tank. The mindbender, in essence, tanks by taking away the enemy's ability to pick their own targets, forcing the enemy to target the Mindbender. It doesn't matter that you're the worst person for them to attack, they have to attack you now, because you said so. The main problem with the Mindbender is that, because abilities that let you dictate enemy action are among the most effective tactics in the game, there's a lot of options available for resisting it, or even shutting it down completely: 99.9% of methods you might use for dictating an enemy's actions are fully shut down by five of the fifteen creature types, and that's without touching any other methods of shutting out mind-affecting effects in general or fear in specific. And more often than not, the method by which you're screwing with the enemy's action economy is limited in other ways just as part of how the mental attack works. But more important than the commonality of ways to get around this style of tanking is how unfair it seems on the other side of the screen; because you're dictating that enemies attack the worst person for them to attack, you're making yourself quite frustrating for a DM trying to provide challenges, and will start including enemies that can take you out more quickly, enemies that can shut down your Mindbender abiltiies, enemies that are immune to your Mindbender abilities, or some combination of the three.
This is your beguilers, your enchanter wizards, your Knights, and so on. This style of tanking is the closest to the "ideal tank" I discussed at the start, but is the most vulnerable to mechanical limitations, out-of-game metagaming, and DM frustration.
One method of tanking without giving up and being a caster is a very tricky way of tanking: a tank that pretends as hard as they can that they're a glass cannon, even though they very much aren't. To reference the earlier example, if the Cleric or the Fighter were capable of pretending to be the Rogue or the Wizard, this could result in informed/intelligent enemies targeting the Cleric or Fighter by mistake. While such builds can theoretically be effective tanks, in practice they only really work extremely short-term because of gradually-increasing information on the party in-game between fights (or even between rounds in the same fight). In-character, a PC that looks dangerous and frail might initially get swarmed, but when they don't go down quickly and don't really pour out the damage, even kinda dumb enemies will realize they're not a threat and move on to more important targets; at best, you'll maybe waste two rounds a fight with this, assuming that no enemies learn of your tricky ways ahead of time. But speaking of that, we need to look at this out-of-character: after a few levels of this schtick every combat, your DM is going to possibly-not-even-intentionally convince himself that the enemy wouldn't target the Oscar Nominee for some reason, and your schtick stops working as often and your party members are getting attacked more and now you're getting a D minus on your Tanking Report Card where you used to be a straight A student.
This is gonna be your Dwarven Defenders, your VoP monks, your illusionists, and so on. Such characters are theoretically viable tanks but become exponentially less effective the more information the enemy has to work with.
One method of tanking is rather...antithetical to the normal ideas about what a "tank" looks like, is the Enabler. This "tank" focuses on improving the party's ability to fight; every action this person takes benefits the party, makes the party harder to take down or more dangerous to the enemy. The Enabler sees that the most effective way for the enemy to behave is to target based on the DPR/HP ratio, and spends their turns screwing with the ratio by increasing either side (or both) on whomever they please. They act as a source of chaos to any enemy calculations, and every extra turn they get to act is another turn they'll get to screw with the math. It could be reasonably argued that constantly making your allies more tanky/dangerous increases your own effective DPR every round, making you the best target by the assumed math; it could also be reasonably argued that taking you out, even if you're possibly not the best target by the math, makes the math a whole lot simpler to perform round-to-round. This method is probably the second-most effective but is also not very...classically tanky.
This is your buffers, your healers, your summoners, your Apostle Of Peaces, and so on. This particular style of "tanking" is typically the domain of casters who tend to be closer to glass cannons than tanks by the math, which might make this style of tanking less ideal for you.
This next tanking method looks at the given example encounter and concludes that the best way to tank is to cut off the giant's access to targets; it doesn't matter that the cleric is the worst person to target if they're the only person to target. Typically, this build focuses on locking down targets, either by locking down a significant area to keep a horde busy, or going one-on-one with the big baddie to keep them occupied, usually in some manner related to the AoO subsystem. There's some builds that will overlap with the next method, which is even better.
This method is ruled by reach monster builds, Jack B Quick style builds, builds focused around trips or bull rushes, and so on.
The final tanking method is a simple idea that lends itself to interesting tactics: if the problem with the default tank is that high defense/low offense means not getting targeted (or tricking/forcing people into targeting you), and high defense/super-high offense means ruining everybody else's fun, what if you play somebody with low defense/low offense who can spend resource to make either of those a force to be reckoned with. That is to say, what if you play a character who could be a glass cannon or a tank from moment to moment? The goal of this is a character who is difficult to deal with when they're defending, but if you ignore them will turn into your biggest problem. These characters tank by making themselves your problem, because if you neglect them you will instantly regret it; they might not be the best person to target when they're on defense, but you can't afford not to attack them.
This is the purview of higher-op monks, defense-focused rogues, paladins, and uberchargers - especially Shock Trooper uberchargers.
The knight's problem isn't that it's unaware of these methods of tanking, the knight's problem is that it's trying to do all of them.
Knight is a mounted ubercharger, the easiest kind of ubercharger to shut down and the hardest to maintain past the low levels: mounted charging damage will have a tough time keeping up once it has to start competing with iteratives, most any mount you're using is going to be unusable in more cramped spaces and most easily tripped up or prevented from charging even in more open areas, and to top it off, you don't have any feature granting you a super-mount so past like level 6, even with Mounted Combat, even with Improved Shield Ally, your horses and whatnot will be dying like flies. Sure, the paladin also has a lot of these problems, but it also has smiting for non-mounted ubercharging, and it at least gets a super-mount...but even in a game where you always get to ubercharge, your damage isn't good enough to be a One Punch Man contender. Even in an ideal situation, you aren't good enough at this to ruin people's fun, which means that one of the other members of the party probably has a higher DPR/HP ratio than you.
Knight's got an okay-threatened area with its ideal style and a few class features synergizing with that, but doesn't have the feats to expand its threatened area or take make good use of AoOs beyond a piddly bit of damage. Doesn't have enough feats to be a reach monster, doesn't have enough feats to be Jack B Quick, doesn't have enough feats for fun ToB addons. The knight designers know that a common tactic used against "armored warrior types" is mind-screw stuff, so it gets a good Will save in exchange for giving up a good Fort save...but every main "tank" class handles this problem better: Barbarian gets rage bonuses to will saves, Paladin gets Cha to saves, Monks just have all good saves by default (and an extra bonus against enchantment effects), and Fighters have an ACF letting them use an immediate action to exchange BAB for Will save bonus.
The knight gets one ability for taking control of enemy actions, and it's basically an ability they can realistically be used once a fight, gives a Will save that will lag behind equal-level caster saves from the second you get it, doesn't affect creatures with too low Int, doesn't affect creatures that are too much weaker than you (meaning it's only getting used on creatures that might have a good chance to make the save), can't be used more than once on the same creature per day, and ends on any creature that gets attacked by somebody other than you. Are there upsides to this ability, of course! It takes a swift action, affects a 100 ft radius, is Ex so it'll work in antimagic environments, and isn't explicitly mind-affecting. But the upsides don't balance out the downsides enough to make this a solid tanking ability on its own.
And all of this is kinda clear in the example builds of Person_Man's knight handbook: all of the builds end up focusing on one of the knight's tanking attempts and focuses on making it good (even to the point of basically not being Knight builds but rather other builds with Knight tacked on), but you can't use everything; if you try, you either fail, or you become the One Punch Man and ruin the fun.
TL;DR the knight is designed with an understanding of what can make a good tank, but tries to touch on a bunch of them and ends up being awful at all of them by default.