Requesting Data for Parry RATE (not amount mitigated)

[stoxastic]

Quick eyeball test shows 1% parry rate per 35-40 parry stat assuming linearity. This lines up well with the data collected on crit stat, which increases crit rate AND crit damage by 1% for every ~40 crit stat. Crit rate baseline is 5% and crit damage baseline is 145% I believe. There was a reddit thread on this a couple weeks back.

So what would you rather have? +1% crit rate/dmg or +1% parry rate? I like crit more. Maybe if parry increased parry strength it would be a lot better.

[Hundred]

Based on My result I believe it's definitely less then 40 given 734 Parry Drk.

1001 Hits 43 swings, 221 parries.
221/958.
23.06% Parry rate.
Could a) divide the whole rate by 734.

B) subtract the 354 base % rate then divide by 380.

Either way that's ~31-34. Give or take a point.

I would wager base Parry rate for War/Drk is actually 11% or close to 11% and not 10%.

[Dervy]

Quick eyeball test shows 1% parry rate per 35-40 parry stat assuming linearity. This lines up well with the data collected on crit stat, which increases crit rate AND crit damage by 1% for every ~40 crit stat. Crit rate baseline is 5% and crit damage baseline is 145% I believe. There was a reddit thread on this a couple weeks back.

The Crit scales 0.00023310023% per point. That's for both Chance and Damage. And yup, 5% Base with 145% damage mod.

So, if it scales around... ~35 per point…

Someone try this and tell me how accurate it is:

((Parry-354)/(858*4))+0.05 (for pld)

[Phoenicia]

The Crit scales 0.00023310023% per point. That's for both Chance and Damage. And yup, 5% Base with 145% damage mod.

So, if it scales around... ~35 per point…

Someone try this and tell me how accurate it is:

((Parry-354)/(858*4))+0.05 (for pld)

I think you need to factor in your block rate to find actual parry percentage. So assuming a 20% block rate, your parry would be something like:

(((Parry-354)/(858*4))+0.05)*0.8

Also enemy crit chance should be factored in for an absolute per-enemy chance. Making the function more like:
(((Parry-354)/(858*4))+0.05)*0.8 *(100-enemy crit%)

[Docfiord_Fowling]

Glad to see a lot of people chiming in now. Anyone have some updated results? Based on these numbers, can we safely assume DRK naturally has a higher base parry rate?

[Aaliyahrose]

I hope no one is using Stoneskin or Clemency to keep themselves alive, you can't block or parry while casting.

True. However, PLD should be using Clemency on party members to keep others alive and (you could use it on yourself if you see yourself with dangerously low HP) possibly on yourself to help the healers during raid wide aoe. PLD is already beefy enough as it is.

[Jpec07]

I have never been so happy for a necropost in my life, I don't think.

I didn't know this thread was a thing. Apparently the 35 point : 1% Parry conversion is as of Heavensward. Feel free to help me shove my foot into my mouth on that particular data point.

That said, let me math some. It takes 350 Parry to mitigate 2% of physical damage (350pt*1%/35pt = 10% *20% = 2%), when we compare that to Crit gains, according to the coefficient a few posts above you're trading right around 8.16% Crit for that 2% mitigation. When we adjust total damage dealt, assuming flat base stats, it becomes 91.84 + 150%*8.16 = 104.08 = 4.08% damage boost. Now, this doesn't account for the exponential scaling provided by Crit, but regardless, let's work with that 4.08% figure a minute. Let's give tanks the benefit of the doubt, and say everyone contributes even damage for the duration of a fight. In a 4-man group, that translates to a 1.02% increase in group DPS. It is doubtful that killing it 1.02% faster will outweigh 2% physical mitigation provided by equivalent Parry. So let's ballpark some more coefficients here: we'll say only 70% of incoming damage is parry-able, and you're not being attacked for 20% of the fight, so 80%*70% = 56%. And let's also say that tanks can contribute only 15% of a small party's DPS (so 60% of what we had before). That brings effective parry mitigation to 1.12%, and group DPS boost from Crit to 0.72%. Unless killing the boss .72% faster will mitigate more damage in a small party than the 1.12% mitigation you get from Parry, Parry is still the best option for mitigation.

Now, thanks to the exponential nature of Crit scaling, there IS a threshold where Crit will cause a tank to do substantial enough additional DPS to mitigate more total damage than equivalent Parry. But I have a sneaking suspicion that they built it in such a way that top tier gear pre-4.0 will cause them to be equal. It will also take a calculus equation to determine the actual effectiveness of Crit, which I am too rusty at math to perform (it's...been a decade since AP Calc. I feel old. :|). It is also difficult to gauge the additional value of Crit for WAR and DRK, who employ self-heals based on damage dealt, without comparing amount healed to amount damage received - data which is inaccessible without a ToS violating parser (or hours of meticulous analysis).

It's also worth noting that fights with extended periods of no damage and/or high magic damage will degrade Parry's value, and similarly, fights with extended periods of time wherein the boss cannot be attacked will degrade the value of Crit. Parry also has very little value for someone in an off-tank spot. And it is further of note that Crit's effectiveness is cut by at least half in a Full Party, while Parry's effectiveness is untouched.

So for main tanks of either Small Parties or Full Parties it seems that Parry has much more value for mitigation than other stats. However, for off-tanks, it seems like it's still advisable to optimize DPS.

[Grux]

snip

Interesting insight as always, Jpec.
And we all want to support the underdog materia.

My opinion is that, for DRKs and WARs, the idea of that small % of getting a crit for self sustaining heals (esp DRKs - a crit DA SE is about give or take 10% heal) is a lot more attractive than that small % of parry. Especially so when DA SE is part of DRKs rotation.

DRKs have the self sustaining toolkit whereas PLDs are designed to be more of a turtle tank.

This is all in terms of mitigation and post-mitigation, not damage dealt.

Requesting for others to chip in, esp. Syz.

[HoodRat]

DRKs have the self sustaining toolkit whereas PLDs are designed to be more of a turtle tank.

Blocking takes priority over parry, so I'd imagine parry would be less effective on plds. If anything, drks would get the most usage out of parry because parries have a chance to reset the cooldown on low blow and let them use reprisal. Also, with tank swaps being a thing again, I'm not sure you can really have a mt stack parry and an ot avoid parry. Unless you're just doing casual content, in which case your secondaries probably won't even matter.

[WeekendSoja]

Didn't mean it as an insult. You have something against princesses?

God, I was going to read the whole thread out of boredom...but than this...great response! Thumbs up! Lol