How would you fix Warrior?

[Hachiko]

A shield would be far bettr.
No wasted healing.

A shield would be far better. But keep in mind it would also be pro-active. If I knew a big hit was coming in I could put up the shield, and then mitigate the damage essentially before it happens.

This would be super nice in terms of balance, but it might go against their design goals (maybe they like WAR dipping dangerously low every time they take a big hit?) I'm not sure.

[CurlyBruce]

A shield would be far better. But keep in mind it would also be pro-active. If I knew a big hit was coming in I could put up the shield, and then mitigate the damage essentially before it happens.

This would be super nice in terms of balance, but it might go against their design goals (maybe they like WAR dipping dangerously low every time they take a big hit?) I'm not sure.

Make it so the Shield also applies a 50-75% healing boost to the next cure spell cast on the WAR. That way you can either buff up with Shield before a big hit to absorb it and reduce damage, or use it after to let the mages top you off easier and give you a bit of buffer to survive the next few auto-attacks. If you lose the shield early you lose the healing buff, and alternatively the shield goes away once a Cure spell is cast.

[Gamemako]

This would be super nice in terms of balance, but it might go against their design goals (maybe they like WAR dipping dangerously low every time they take a big hit?) I'm not sure.

I would have to agree. A PLD acts before damage is delivered; a WAR acts after damage is delivered. A shield would make WAR proactive.

[Hachiko]

I would have to agree. A PLD acts before damage is delivered; a WAR acts after damage is delivered. A shield would make WAR proactive.

Well the idea that WAR is all reactive is kind of not accurate in the first place. Our only really "reactive" mitigation is Inner Beast

Featherfoot = Proactive
Foresight = Proactive
Convalescence = Proactive
Thrill of Battle = Proactive (but can be used as reactive)

[Gamemako]

Featherfoot = Proactive
Foresight = Proactive
Convalescence = Proactive
Thrill of Battle = Proactive (but can be used as reactive)

I highly disagree. Only Foresight is innate and proactive. Featherfoot is cross-class and proactive. Thrill of Battle is either, because it always recovers 20% of your HP. Convalesce is reactive; the ability won't do you any good until you've already been hit. It is also cross-class. IB is reactive, Bloodbath is reactive, Wrath's healing bonus is reactive, Storm's Path is reactive (and sucks), and cross-class Mantra and Second Wind are reactive. So what you really have are:

1-2 innate proactive (Foresight, Thrill of Battle)
2 cross-class proactive (Awareness, Featherfoot)
5-6 innate reactive (Defiance, Inner Beast, Bloodbath, Mercy Stroke, Storm's Path, maybe Thrill of Battle), plus 3 more which enhance reactive (Berserk, Maim, Storm's Eye)
3 cross-class reactive (Second Wind, Convalesce, Mantra)

Innately, reactive outnumbers proactive 5-1 if you exclude ToB or 6-2 if you include it in both. In total, you have at 3-4 proactive and 8-9 reactive, plus 3 more which support reactive. This is clearly intended to be a reactive tank, not a proactive one. Compare to PLD:

6 innate proactive (Rampart, Sentinel, Bulwark, Shield Oath, Hallowed Ground, Awareness)
1-2 cross-class proactive (Foresight, Stoneskin if you're really smoking something)
1 innate reactive (Convalesce)
2-3 cross-class reactive (Mercy Stroke, Bloodbath, Stoneskin if someone hit you on the head really hard), plus 1 which can play a minor support to reactive (Fight or Flight)

It's clear enough that innate PLD is almost exclusively proactive while innate WAR is almost exclusively reactive. PLD's proactive abilities are also simply more potent than WAR's, and the opposite is true of reactive abilities for PLD. WAR gets more abilities which enhance the power of those reactive abilities, while PLD doesn't because PLD doesn't need them -- PLD is a proactive tank.

[Hachiko]

Bloodbath isn't really reactive. Since it's not burst mitigation it's just "over time" and it's a small amount over a long period of time, calling it either proactive or reactive is a little bit of a mischaracterization. Same with Storm's Path, it's not usable on demand, and its only contribution is in the form of an "over time". I know you're counting all healing as reactive, but usually healers consider HoT type skills as "proactive" heals, meaning you put them on the target once they take some damage or are about to take some damage, in the hope that after they take the damage the hot will heal some of it. Essentially, Bloodbath and Storm's Path are the same way, except they are so minimal counting them as proactive seems kind of silly.

Convalescence is functionally a pro-active cooldown. You want it up before you take the damage so any pre-cast heals will get the buff from it. The effect doesn't hit you until after you take the damage, but it's something you should be pre-buffing with. I don't think you should really be counting mercy stroke at all, either, because the heal is not of use in any sort of raiding situation and even solo it's hard enough to time it between attacks / auto attacks such that you actually get the kill with it and not an attack of some other sort. You are being pro-active with convalescence, even if the healers are being re-active with their heals.

Also stoneskin is pretty amazing for a PLD. There are a lot of uses on many fights (Garuda, Titan, ADS, etc.)

But if you look by bulk mitigation, Inner Beast is clearly intended and usually does provide more than any other skill, and that is reactive. So maybe it's the design intent, but it isn't really a functioning intent anyway, so it's hard to tell if that is really what they had in mind or if they even thought about it from this angle.

[Jediman]

problem goes away if they turn it into dps war instead.

[Gamemako]

Convalescence is functionally a pro-active cooldown.

I think we are not on the same page about what is meant by proactive versus reactive. Think of the tank cycle as this:

[damage taken] -> [damage healed] -> [damage taken] -> [damage healed] -> ...

Proactive abilities act on the [damage taken] cycle. They are enacted before an enemy hits you, or they have no effect. Reactive abilities act on the [damage healed] cycle. They have are enacted after an enemy hits you, or they have no effect. A few abilities, like Thrill of Battle or Stoneskin, are a bit of both: they will have an effect when enacted any time. Bloodbath, Mercy Stroke, Storm's Path, Inner Beast, Convalesce, Defiance's recovery stacks, Second Wind, and Mantra will not have an effect before you are hit because you can't heal what hasn't been hit. You may enact the ability beforehand, but that is not relevant -- the timing of the effect is the essential element, and in the case of +healing buffs, the effect is post-damage.

You certainly could consider it as a impact-by-rate, though, and PLD is easily 98% proactive while WAR is easily 95% reactive. I don't think it's all that hard to divine design intent from that sort of dichotomy.

//EDIT: Also, I want to quickly expound the implications of this and how they point to at least one obvious fix for WAR. Let's say WAR heals for 40% of it's max HP, which would be balanced with PLD's Sentinel. However, with identical eHP, PLD is flat-out better. Take this example.

->Scaryguy Readies Big Honkin' Hammer
->PLD uses Sentinel
->Scaryguy hits PLD for 60% of PLD's max HP!

PLD reduces the damage by 40% and survives easily. Now, WAR doesn't get that. Instead,

->Scaryguy Readies Big Honkin' Hammer
->WAR stares him down, eager to hit back!
->Scaryguy hits WAR for 100% of WAR's max HP!
->WAR dies.

Even though WAR could have healed it back, WAR is dead because its abilities are reactive rather than proactive. This is why WAR must have a considerable eHP advantage over PLD. At this time, PLD has identical eHP to WAR. This is just flat-out broken.

[Hachiko]

I suppose we just have different definitions on proactive and reactive. I Think of it in terms of ability useage, i.e. do you use it before or after you take damage. You think of it in terms of when the damage is mitigated, i.e. is it mitigated before or after it comes in.

This is why I think the intent is hard to divine. It's difficult to tell what the devs had in mind with this specific type of nuance, because so many of the WAR's available mitigation skills are meant to be used before the big damage comes in, even if the general idea around it is self healing.

I will just say in general that reactive abilities are also generally crappier because they aren't as reliable, and they are prone to having you die from burst, and they have a tendency to end up as overheal. Which is why moving Inner Beast to a pro-active shield would do wonders in terms of balance.

[Gamemako]

I will just say in general that reactive abilities are also generally crappier because they aren't as reliable, and they are prone to having you die from burst, and they have a tendency to end up as overheal. Which is why moving Inner Beast to a pro-active shield would do wonders in terms of balance.

As I was just editing in, the issue is more that you have to design around it. Self-heals can end up as overheal, but not all reactive abilities are self-heals. +healing bonuses are also reactive, and they're effectively identical. The problem that WAR has is that WAR is almost exclusively reactive in nature (i.e. effects are post-damage rather than concurrent with damage). This requires that WAR have a considerable eHP advantage over PLD as a rule. When you give WAR 25% more eHP over PLD, then it's not a problem at all. Instead, you can do this:

->Scaryguy Readies Big Honkin' Hammer
->WAR stares him down, eager to hit back!
->Scaryguy hits WAR for 75% of WAR's max HP!
->WAR retaliates and recovers 25% of WAR's HP!

WAR ends with 50% of its HP, and you're back to the fight as before. Same boss, same incoming damage, but now WAR walks away just fine. The only thing which matters in implementation is when the mitigation comes. This is why I just cannot for the life of me figure out why WAR doesn't have a large HP advantage over PLD. This is not a matter of scaling, either; this is true in every single case.