Hydaelyn Role-Players

Parry Dissertation
I used to rate Parry very highly for its mitigation properties until I sat down at endgame and did a little parsing and a little paying attention during fights. And when I did this, I became very disenchanted with the stat. So first, let's look at what the Parry stat actually does for you:

1) Increases the Parry Rate with main hand weapon.
2) Increases the Block Rate with off hand shield.

What are the limitations of the Parry?

1) Cannot proc against Magical Damage.
2) Procs against Physical Damage are entirely RNG based for all intents and purposes.

There is currently no concrete data pinpointing a statistical point value at Lv50 for how much a single point of Parry affects Parry Rate. However, preliminary testing by Hulan was published on B.L.I.T.Z.B.A.L.L.'s site back in Beta suggested that 1 Parry = 0.076 increase in Parry Rate. Later testing by Reddit user Kestiel seemed to corroborate this number. Meaning a hypothetical increase of +200 Parry stat would only increase your Parry Rate by +15.2%, statistically.

I emphasize statistically because you can ask ANY crafter or ANY gatherer how reliable a 20-25% HQ rate is in reality, and they will laugh at you with tears in their eyes. That is FFXIV RNG and you are playing that roulette with your life.

So we are forced to work with hypothetical, and even that is quite grim, for WAR especially. Let's say you have two i130 WARs, one VIT with Parry focus, one STR without Parry focus.

>VIT WAR has 12000 HP, 486 STR, and 600 Parry, giving them a 25% (19.7% from Parry stat + 5% base) Parry Rate.
>STR WAR has 9800 HP, 631 STR, and 400 Parry, giving them a 10% (4.5% from Parry stat + 5% base) Parry Rate.

Statistically, the Parry WAR has the greater chance at frequently Parrying incoming damage.
Realistically, the difference is negligible and completely random.

Both i130 Tanks big pull 10 mobs in a hypothetical dungeon. Because there is 0 difference in DEF between the two tanks, both tanks take 500 DMG per attack unmitigated. Let's say things go as intended, the VIT WAR statistically Parries 2 of the 10 incoming attacks. While the STR WAR Parries 1 of the 10 incoming attacks.

>VIT WAR takes 8x500 dmg + 2x370 dmg (-26%) from the 2 parried hits = 4,740 dmg.
>STR WAR takes 9x500 dmg + 1x345 dmg (-31%) from the 1 parried hit = 4,845 dmg.

Statistically, Parry WAR comes out on top, barely. However, RNG can be a bitch, and hypothetically both tanks could Parry all or none of these 10 incoming attacks if (un)lucky enough and any number in between. The STR WAR could Parry 6 of those attacks while the VIT WAR only Parries 1, in which case, the STR WAR's mitigation becomes far superior.

Then there's fights where a boss or mob group only uses Magical Damage, in which case Parry is completely useless. And if you are that Parry WAR in a Magical Damage fight... congrats, you've wasted nearly the entirety of your stat allocations on a stat that does nothing for you, hypothetically and statistically.

----

So this is all great for WARs, but what about PLDs? You said Parry stat also affects Block Rate? Correct! It does, but at the same 0.076 increase weighted against the Shield's actual Block Rate stat. When a PLD takes incoming damage, the damage is actually weighed against the Shield first. RNGesus /randoms to see if your Shield will Block the attack, if it does not, it then rolls /random again to see if your Sword will Parry the attack, before the PLD actually takes the damage against its DEF rating and Mitigation from Shield Oath/Rampart etc.

So why then is Parry useless for PLD if it affects Parry Rate and Block Rate?? Because, a shield with high Block Rate will be just as if not more effective at increasing your Block Chance as dumping 200 Parry will, and because a high Block Rate trumps Parry Rate you're effectively still only plugging your Parry stat into your Sword, which you will rarely ever actually get to use, because your Shield procs first. Top a high Block Rate shield with Bulwark and you'll never actually see your sword parry. Your shield is doing all the work on its own. So why on earth would you want to dump all of your secondary stat allocations into Parry, when your shield will actually do the work all on its own?

So if you want to Block more frequently as PLD, get a higher IL shield. The only time Parry would hypothetically be recommended is for a PLD with a Noct Hoplon, which boasts a higher Block Mitigation stat than a Block Rate stat, so Parry could be used to supplement this lower proc rating. However, most endgame shields are Kite Shields that boast equally high Block Mitigation and Block Rate numbers.

So again, a Parry PLD will statistically have a slightly higher Block/Parry Rate versus a non-Parry PLD of the same Item Level, but Bulwark effectively makes this difference negligible since both tanks can hit near-100% Block Rate with Bulwark and crushing their Parry rating down to 0%.


So all of that was simply the math portion of the lecture! Lemme quickly wrap this up by talking from a Healer's perspective. I could care less what your Parry rating is, because 1) I can't see it actually working - I have NO idea when you've parried and when you've eaten a full blow. 2) I'm still going to hit you with Cure II or Lustrate regardless of whether you Parry or not. Because I'm a good healer and I have that shit prepped for when you're taking damage. Your heal is inbound before that damage is even calculated. So, let me reiterate my earlier statement, all your Parry did was keep you from being one-shot. If you were not one-shot, my heal is coming 0.5 seconds later regardless of whether you took 5000 damage or 4700 damage. My lustrate will still hit you for 2300~ in both cases lol.

Is it really nice and thoughtful that my tank is Parrying a few hits on big pulls in dungeons versus trash? Sure, but conversely, you could have also taken all that Parry stat you focused and put it into Determination or Critical Hit Rate on your WAR/PLD and help with damage, which 1) increases kill time and decreases the duration of time I have to curebomb you, 2) increases your overall enmity generation meaning I won't get murdered by mobs because I curebombed you, and 3) increase the amount of self-healing you are able to do to supplement the incoming damage.

Let's take the example above of the 2 i130 WARs in hypothetical dungeon with 10 mobs on them. Out of the 5,000 damage (10 x 500 dmg) they would've taken unmitigated, their Parrying only actually reduced incoming damage by 155(STR WAR) and 260(VIT WAR) points.

The Pro of the VIT WAR is that after taking 5000 damage, he still has 7000 (58%) HP.
The Pro of the STR WAR is that after popping Bloodbath, he can hit Overpower and heal back 1,000-1,500 HP in one move, bringing his HP up to 6000 (61%) HP after the fact.

Could the VIT WAR also Bloodbath+Overpower/Steel Cyclone/Inner Beast the same as the STR WAR? Sure he could, he just won't receive as good an HP return, which is fine, cuz he has his Healer's Curebomb inbound.

But overall, the VIT/Parry WAR is an inefficient way of optimizing the role, as you lose out on damage, enmity, and self-healing on the supposed basis that you gain more mitigation/survivability by stacking Parry and Vitality, when in fact, the difference in mitigation is mathematically negligible. Does a Parry/VIT WAR work? YES. Is it the most efficient way to play WAR? No.

/end lecture. Questions?