Parry...

[LoganNinefingers]

Um no to everything you just said in the beginning of your post, nothing in this game has diminishing returns. Every single test ever run confirms this. The exception being Speed, and ACC for different reasons.
Crit being the only stat that might have an argument for diminishing returns, even though I still have not heard a good one.

By process of elimination, what you've effectively said here is "of the other secondary stats that matter to non-mages, determination doesn't have diminishing returns, but the others do/might under certain conditions", so why the immediate "Um no" response?

The vast majority of testing I've seen into any of the less understood stats has been carried out by people who either

A) have very little working knowledge of statistics;
B) are lazy; or
C) both of the above,

as, in general, they don't quote all relevant data (usually just something along the lines of "15 points Crit Rate Stat gives 1% actual crit rate increase, parser said so"), and often only use two data points (with which it is impossible to infer anything other than an often misleading linear trend).
As such, a blanket statement like yours cannot be made with any certainty at this point in time.

The fact that both Menae and ZDamned (amongst others) have recently asked for clarification on whether there are diminishing returns or a "soft cap" for Parry shows that this issue definitely hasn't been concluded once and for all, and further speculation and testing is required.

So please, stop making these sweeping declarations before any real supporting evidence is in place; for now it just spreads ignorance around the player-base, and needs to be kept in check until we have the data to say for sure one way or the other.

[Casper]

Snip

A large quantity of testing had been done very early into the game (see this). It showed conclusively that we had a formula of 10% parry rate flat + 1% per additional 13 parry rate above base value. I have tested myself that this formula is still valid now that we have access to much higher gear, with more stats into them, etc.

At this point this formula has been showed to work on all kind of level, all kind of gear, on all kind of monsters. There are still people to show up and say "but my healers can't tell the difference LOL". If there is any kind of soft cap, we have either not reached it yet, or just recently (which would be hard to prove until we have more gear). At this point, anyway, it is fucking CERTAIN that going zero to 250 additional parry is going to make your parry rate jump from 10% to almost 30%. Despite this, you have people still arguing that det would provide more benefits. At this point, this is a lost cause.

[SwordCoheir]

A large quantity of testing had been done very early into the game (see this).

Yes using P1-3 Beta data is definitely a reliable source...

I can already throw some of this data in the trash because I went over the value of Block Strength and STR's effect on Block Mitigation just a few weeks ago.

Here: http://forum.square-enix.com/ffxiv/t...85#post1998985

and Here: http://forum.square-enix.com/ffxiv/t...58#post2000858

And yes while I am aware the discussion is about parry, if that data has altered so much in your notes it shouldn't be used as a definitive point of reference.

[ZDamned]

(And I'm not trying to pick on you - honestly I'm not! - but 1 is a soft cap, not 3. A soft cap is where a stat suddenly drops in value, but not to zero. 3 is a hard cap, where the stat literally doesn't do anything for you after a certain point. Diminishing returns are when a stat slowly and steadily loses value as it increases. REGARDLESS OF TERMINOLOGY, those are great questions and I hope to learn the answers too.)

No worries! I was under the impression that a Hard cap is the highest attainable from available gear, and a Soft Cap is the highest usable amount of a stat. I always thought Diminishing Returns were in a different category, but a lead in to what I had understood was the Soft Cap. So, thanks for correcting me!

Here is the thing: yesterday, i92 gear oriented parry, on T7, i was doing 125 to 130 in shield oath on the boss. Doing your best to not forget any SW/CoS when in dps gear, then slacking in tank gear is going to skew your results. If anything, the numbers you provided just proved that you should most definitely go in full parry because the dps you will gain by going with no parry and det and crit instead is going to be negligible (6 to 7% more likely, a bit more if you go with str accessories). Those numbers are on par with what you said and what i said, i.e foregoing all parry is going to make your dps in T7 go from 125 to 136.

Ok, Looks like there needs to be a consistent standard set on how much damage you can deal with, and without Determination/Crit, because that number you pulled makes no sense to me, unless your parser works differently than mine does.
If we ignore secondary stats all together:
I was using a Wave Buckler to maximize Shield Swipes. 3xSS are greater potency than 1 RoH combo, so I only use RoH when it's about to fall off, or when SB has connected, otherwise SS is higher priority.
Your lodestone shows you using the Noct Hoplon, If you were using that shield, you should generate way less opportunities to use Shield Swipe.
So that alone should give me a huge potency edge by the 3 minute mark in how much damage was dealt. (Unless you got super lucky, and blocked the same number of times as I did with a Buckler)
Which is the next question I want to ask you. How much total damage did you deal in that fight? I'll get my number and post it soon. (I'm at work now)
Also, did you Main Tank, or Off Tank? and Did you clear the boss?
On a Training Dummy, I get a clear 30~ DPS boost in my DPS gear over my Parry gear, and that is without Blocks adding more potency, so my in combat DPS is higher while standing still on a boss like T7.
So your stated number ALSO bothers me because I have more strength than your lodestone profile(If that is the gear you were in) by at least 10 points.
Your DPS number is either flat out wrong, or your parser is badly configured. Because stats can't enable you do pull those numbers.

[Casper]

No worries! I was under the impression that a Hard cap is the highest attainable from available gear, and a Soft Cap is the highest usable amount of a stat. I always thought Diminishing Returns were in a different category, but a lead in to what I had understood was the Soft Cap. So, thanks for correcting me!



Ok, Looks like there needs to be a consistent standard set on how much damage you can deal with, and without Determination/Crit, because that number you pulled makes no sense to me, unless your parser works differently than mine does.
If we ignore secondary stats all together:
I was using a Wave Buckler to maximize Shield Swipes. 3xSS are greater potency than 1 RoH combo, so I only use RoH when it's about to fall off, or when SB has connected, otherwise SS is higher priority.
Your lodestone shows you using the Noct Hoplon, If you were using that shield, you should generate way less opportunities to use Shield Swipe.
So that alone should give me a huge potency edge by the 3 minute mark in how much damage was dealt. (Unless you got super lucky, and blocked the same number of times as I did with a Buckler)
Which is the next question I want to ask you. How much total damage did you deal in that fight? I'll get my number and post it soon. (I'm at work now)
Also, did you Main Tank, or Off Tank? and Did you clear the boss?
On a Training Dummy, I get a clear 30~ DPS boost in my DPS gear over my Parry gear, and that is without Blocks adding more potency, so my in combat DPS is higher while standing still on a boss like T7.
So your stated number ALSO bothers me because I have more strength than your lodestone profile(If that is the gear you were in) by at least 10 points.
Your DPS number is either flat out wrong, or your parser is badly configured. Because stats can't enable you do pull those numbers.

Roh is 203.3 average potency, SS 210. Difference 3%. The truth is, this is among the standard deviation over a fight given the deviation of attacks by + to - 5% and the crit chances. Not saying you wouldn't gain a tiny bit here, but 3% difference, on 17% more attacks (due to block rate going from 21 to 38) is a 0.5% dps increase... given that you are doing around 130 dps, that's not even 1 DPS of difference. 10 str is 1.5% more damage. Total for both is about 2.5 DPS of difference.

Did you have a warrior ? Did he keep the buff to slashing on the boss ? If you go double pld, for example, that alone would be a much greater factor.

Your 30 dps boost, do you mean in dps stance ? Cause going from 100 to 130 or 170 to 200 is not the same. When you say dps gear, i assume you mean, literally, dps gear, so with str instead of vit ? Because yes, in that case, i could see a 30% damage increase. But you are not talking about sacrificing parry only here, you are sacrificing a lot of vit too to get there. 250 parry into full det and crit, plus 75 vit into str, sure, maybe we are getting closer to 25-30 dps increase, but you did not only lose almost 20% parry rate, you also lost 1k hp.

[ZDamned]

When I say DPS gear, I meant to say DPS Fending gear, and it is in Shield Oath that I gain 30 DPS going from Parry to Crit/Det. Sword Oath, I tend to hit the 185 DPS mark in the Drit/Det set.

My groups Warrior does keep Storms Eye up.

Ok everyone! Results are in! Here are my Parse logs from tonight's T7 Clear, and my first T7 Clear on Sunday.

Damage taken on Turn 7 169,119 in Parry gear. (564 Parry, ilvl 94)
Damage taken on Turn 7 173,978 in Crit/Det Gear. (449 Parry, ilvl 93)

Damage Dealt on Turn 7 64,879 in Parry gear. (564 Parry, ilvl 94)
Damage Dealt on Turn 7 73,719 in Crit/Det gear. (449 Parry, ilvl 93)

-EDIT- Fixt the Parry Numbers I mis read it initially, and had reported 479 in Crit/Det gear. SORRY! =P

I will GLADLY take 4859 damage to gain a level 1 limit break every 6 minutes.

[Casper]

When I say DPS gear, I meant to say DPS Fending gear, and it is in Shield Oath that I gain 30 DPS going from Parry to Crit/Det. Sword Oath, I tend to hit the 185 DPS mark in the Drit/Det set.

My groups Warrior does keep Storms Eye up.

Ok everyone! Results are in! Here are my Parse logs from tonight's T7 Clear, and my first T7 Clear on Sunday.

Damage taken on Turn 7 169,119 in Parry gear. (564 Parry, ilvl 94)
Damage taken on Turn 7 173,978 in Crit/Det Gear. (479 Parry, ilvl 93)

Damage Dealt on Turn 7 64,879 in Parry gear. (564 Parry, ilvl 94)
Damage Dealt on Turn 7 73,719 in Crit/Det gear. (479 Parry, ilvl 93)

I will GLADLY take 4859 damage to gain a level 1 limit break every 6 minutes.

Am i corrent in assuming that this is with dps fending gear too ? Which means there was also a 1k max hp difference ? Or was it just by optimizing parry or not optimizing it ?

[MythToken]

When I say DPS gear, I meant to say DPS Fending gear, and it is in Shield Oath that I gain 30 DPS going from Parry to Crit/Det. Sword Oath, I tend to hit the 185 DPS mark in the Drit/Det set.

My groups Warrior does keep Storms Eye up.

Ok everyone! Results are in! Here are my Parse logs from tonight's T7 Clear, and my first T7 Clear on Sunday.

Damage taken on Turn 7 169,119 in Parry gear. (564 Parry, ilvl 94)
Damage taken on Turn 7 173,978 in Crit/Det Gear. (449 Parry, ilvl 93)

Damage Dealt on Turn 7 64,879 in Parry gear. (564 Parry, ilvl 94)
Damage Dealt on Turn 7 73,719 in Crit/Det gear. (449 Parry, ilvl 93)

-EDIT- Fixt the Parry Numbers I mis read it initially, and had reported 479 in Crit/Det gear. SORRY! =P

I will GLADLY take 4859 damage to gain a level 1 limit break every 6 minutes.

How long was each fight? These numbers don't mean anything without a time.

[bokchoykn]

When I say DPS gear, I meant to say DPS Fending gear, and it is in Shield Oath that I gain 30 DPS going from Parry to Crit/Det. Sword Oath, I tend to hit the 185 DPS mark in the Drit/Det set.

My groups Warrior does keep Storms Eye up.

Ok everyone! Results are in! Here are my Parse logs from tonight's T7 Clear, and my first T7 Clear on Sunday.

Damage taken on Turn 7 169,119 in Parry gear. (564 Parry, ilvl 94)
Damage taken on Turn 7 173,978 in Crit/Det Gear. (449 Parry, ilvl 93)

Damage Dealt on Turn 7 64,879 in Parry gear. (564 Parry, ilvl 94)
Damage Dealt on Turn 7 73,719 in Crit/Det gear. (449 Parry, ilvl 93)

-EDIT- Fixt the Parry Numbers I mis read it initially, and had reported 479 in Crit/Det gear. SORRY! =P

I will GLADLY take 4859 damage to gain a level 1 limit break every 6 minutes.

While I agree with you that Parry is a pretty crappy stat and isn't as good as most tanks hype it out to be, your testing method is totally flawed and the results are unusable.

There are HUGE uncontrolled variables in your test that aren't accounted for. If your overall DPS increased (eg. a week's worth of upgrades for your DPS), this would skew your Damage Taken and Damage Dealt since the fight would be shorter (and not just by virtue of your own damage increasing).

Unless your teammates performance was completely and exactly equal in both attempts (impossible), using Total Damage Taken and Total Damage Dealt as metrics to measure the value of Parry vs Crit/Det is not statistically sound.

[Hulan]

A large quantity of testing had been done very early into the game (see this).

Sigh. I knew it was only a matter of time before someone brought this up. As a point of order, if you read through the section about blocking and parrying, you will notice that the only section that directly pertains to Parry as a stat effecting the Block and Parry rates - the rest pertains entirely with the the amount of damage blocked and the rate of change by adding more STR - was under the Addendum.

Post Beta phase 3 I was made aware of this data testing a lodestone user by the name of Hulan made public:

From earlier in this thread:

I suspect a large amount of the justification for Parry as the go-to secondary stat is based heavily on a small study I myself did back in late closed beta, which showed that Parry drastically improved parry and block rates on low level enemies. ... The study had some pretty serious issues to be honest, as it was not preformed on high level monsters, nor did it provide a comprehensive look at a continuous spectrum of parry values. As such, I would not be at all surprised if the results I found are not completely incorrect and Parry is an awful stat to focus on.

The 10% +1 per 10 you are referring to is the STR mod on Parry.