Hello Fellow PLD

[Armando]

Re: Hello Fellow PLD

DDs make the kill. If I'm really looking for damage from a PLD, I'd replaced the PLD with a DD.

DDs can only kill as well as the PLD's hate holding will allow them to.

is that the longer you delay the cures, the faster you'll fall behind the DDs

Why? The enmity accrued is the same whether you do it at the start of a fight or 1 minute into it; as long as you cast it before losing hate, it's irrelevant when it was cast.

Not to mention there's a higher chance of Cure IV getting interrupted compared to Cure III, and an interrupted Cure produces almost no enmity; it's a complete waste of time.

Well, the choice of IV over III is subject to lag. With little lag, a good PLD shouldn't have problem with interrupts. Even so, you should only get interrupted around half the time; the other half, shield, parry, or that 5% chance to miss will bail you out.

[Solymir]

Re: Hello Fellow PLD

I don't know why evasion gets such a bad rap. I evade things on my Pld and War all the time. I'm fairly sure I'm not doing anything special, but people still write it off as useless for those two jobs. No I am not advocating stacking on the Eva gear, just tired of the lol crap.

[Armando]

Re: Hello Fellow PLD

I evade things on my Pld and War all the time.

That's because the mob's hit rate can't be higher than 95%, just like us.

That's why I say it's not really that PLDs evade, it's that the mob is forced to miss every so often.

[Malacite]

Re: Hello Fellow PLD

I don't know why evasion gets such a bad rap. I evade things on my Pld and War all the time.

Bullshit. WAR's evasion rating is the biggest detriment to it's soloing ability. I can take mobs on MNK and SAM that would wipe the floor with my WAR. 230 cap @ 75 means even EP will give you a rough time.

[Solymir]

Re: Hello Fellow PLD

Bullshit. WAR's evasion rating is the biggest detriment to it's soloing ability. I can take mobs on MNK and SAM that would wipe the floor with my WAR. 230 cap @ 75 means even EP will give you a rough time.

Whatever.

[ItazuraNhomango]

Re: Hello Fellow PLD

DDs can only kill as well as the PLD's hate holding will allow them to.

You're assuming a PLD or a tank is needed at all.

I've a friend who likes to make no-real-tank parties from Lv.30's and up, using two or three SAM/WAR's instead--basically it's one Provoke to pry the critter off the puller, and someone else provoke if the first person who used Provoke is dying. I don't like tank-less parties, but he made it to SAM75 this way, so there's no argument that it works.

DDs can do as much damage as the recovery sources allow--there's nothing wrong with taking a hit unless it creates down time or KO danger.

Why? The enmity accrued is the same whether you do it at the start of a fight or 1 minute into it; as long as you cast it before losing hate, it's irrelevant when it was cast.

The faster you can generate enmity, the sooner the DDs can let loose without fear. DDs WS'ing sooner will do more for kill speed than the PLD deferring Cure for more swing time would anyway.

Well, the choice of IV over III is subject to lag. With little lag, a good PLD shouldn't have problem with interrupts. Even so, you should only get interrupted around half the time; the other half, shield, parry, or that 5% chance to miss will bail you out.

I don't think most PLDs leveling up have their shield skill capped; I know I didn't until months after hitting Lv.75. (Part of that was because I was on RDM more often than PLD, though.) Parry is likely even further behind, unless the player previously leveled NIN.

Plus, there's a bit of skill involved; if you're able to time Cure IV well, that's great, but I never did seem to be able to do it consistently. Observing PLDs in exp parties, I'd say I'm not the only with with that problem.

* * *

I never did get around to testing PLD's evasion vs. merit level critters; is there testing which already shows we're at the floor without eva- gear? (5%)

* * *

Sorry YM, missed this from before; think you were railing against my statement "Want more enmity? Cure more."

NOBODY IS ARGUING AGAINST THIS.

What those sane among us are saying is that both proper damage output and curing are among elements required of a good Paladin.

Well, I was alarmed by this:

Curing yourself is highly inefficient anyway, in terms of holding hate.

(From his posts, seems like he believes he's a good PLD.) Malacite isn't the only person with that misconception, either; I've definitely met a few of those in-game, including one incredibly stubborn Tarutaru RDM who absolutely refuses to leave any room on my HP bar.

So, one more time: As long as the MP recovery permits, curing is more efficient for enmity generation than swinging a sword if there's HP to cure. It's important for up and coming PLDs to understand this.

[Malacite]

Re: Hello Fellow PLD

So, one more time: As long as the MP recovery permits, curing is more efficient for enmity generation than swinging a sword if there's HP to cure.

Oh rly?

Kanican - Enmity Table (Damage Dealt)


Kanican - Enmity Table (Curing)


Sure doesn't look it to me.

[Nuriko]

Re: Hello Fellow PLD

You missed one thing, I think ... those are on a per-HP basis. You can expect not quite 200 HP of curing from a Cure III, and I seriously doubt that a single normal swing of your sword will do QUITE that much damage. Weaponskills can, of course, but they're hardly an every-swing kind of thing...

[ItazuraNhomango]

Re: Hello Fellow PLD

Oh rly?

Kanican - Enmity Table (Damage Dealt)


Kanican - Enmity Table (Curing)


Sure doesn't look it to me.

I already explained this, but let's retread in greater details:

Save the Queen: Lv.65, Dmg: 40, Dly: 240, TP/hit: 6.4%
Think we can agree this is a decent sword for PLD at Lv.65, right?

The 240 delay means 4 seconds between each attack around, for 1.6% TP/sec. (A PLD would get more TP from the damage taken, but he'd get that whichever weapon is used, so it matters not for this purpose.)

Assuming an average of 250 damage Vorpal Blade (which is on the high end) and disregarding the DoT damage for a moment, that works out to be 2.5 damage/TP. Then, that 1 second of swing time works out to be exactly 4 damage on the WS if the PLD does not hold TP above 100%.

250 damage/100% TP = 2.5 damage/1% TP
1.6% TP/sec
2.5 * 1.6 = 4 (damage/sec)

Factoring /WAR's Double Attack at 10% proc rate, it comes 4.4 dmg from WS per second of sword swinging as a generous estimate (including a 100% hit rate hidden assumption).

4 damage/sec * 1.10 = 4.4 damage/sec (Can call this "WS damage DoT" if so wish)


From before:

As a Lv.65 PLD fighting a Lv.72 target, a 190 HP Cure III is worth 931 VE before enmity modification, taking 2.5 seconds.

This is how the 931 VE figure came about:

VE Modifier (Lv.65): 4.90

VE = VE Modifier * HP Cured
VE = 4.90 * 190
VE = 931

Also from before:

You would need to do 193 damage in that same 2.5 seconds just to match the enmity output for a single Cure III--or, have to do even more damage, if the cure was cast with enmity+ gear on.

Working the VE for damage output formula:
VE Modifier (Lv.72 target): 4.80

VE = VE Modifier * Damage Dealt
Damage Dealt = VE / VE Modifier
Damage Dealt = 931 / 4.80
Damage Dealt = 193.9583333 (to be equivalent to VE from a 190 HP Cure III)

The amount of WS output you lose for 2.5 seconds spent on a Cure III:
2.5 sec * 4.4 damage/sec = 11 damage

Now, factoring the (generous) estimate on how much 2.5 seconds contributes to WS output, that means in that 2.5 second, your sword chucking need to do 193 - 11 = 182 damage just to match the VE generation from a Cure III.

Unless the PLD has (additional) enmity+ gear when the Cure lands--in that case, you'd need to sword chuck even more damage than 182 in that same 2.5 seconds.

Can a PLD output 182 damage in 2.5 seconds from sword swings? Maybe at Lv.75 with Joyeuse and Justice Sword both proc'ing a PLD can do that, but not going to happen on average. Definitely not going to happen at Lv.65.

Swinging a sword is a LOW ENMITY activity when comparing to curing 190HP in 2.5 seconds (in the Lv.65 PLD vs. Lv.72 target scenario). Or, use Provoke, use Flash, then swing the sword only when can't afford the MP to cure or if there's no HP to cure.

[Armando]

Re: Hello Fellow PLD

You're assuming a PLD or a tank is needed at all.

Obviously; this discussion is about tanking.

Or, use Provoke, use Flash, then swing the sword only when can't afford the MP to cure or if there's no HP to cure.

The sword will swing any time you're idle so I honestly don't understand your point.

Comparing DoT to Curing in this way is illogical since you do not Cure constantly. You can't do 190 damage in 2.5 secs through normal melee but you can do that much in the time you spend not curing. If you're going to divide WS damage by the time spent between WS, it's only fair that you do the same with Cures, which is a reasonable demand because your MP is restricted by your Refresh, in the same way your WS is restricted by TP; after factoring in Flash and Reprisal, you can only Cure so many times per minute before your MP starts to dip. Damage is what sustains your hate between your limited spikes; if you don't have enough of it you'll be forced to Cure more than your MP can provide, and you'll fare worse when your MP runs out on top of that.

15% Haste (spell) = 37 sec Flash cooldown and 2:33 Reprisal cooldown. Assuming these spells get used as much as possible, combined they consume 50 MP per minute on average. Assuming only Refresh, that's 4 MP/tick on the PLD, which is 80 MP a minute. A PLD can only spend 30 MP on Cures per minute without draining their MP. That's only ~130 HP per minute with Hospitaler's factored in. A DD PLD will easily average more than 130 damage (I would say 200) on his Vorpals and WS at least once a minute. This is setting aside the fact that the DD PLD will also do more than 130 damage just from normal swings in one minute. Using very conservative average PDIF (0.5) and hit rate (75%) and ignoring crits and DA, with a 10 DPS weapon you'd get 225 damage per minute.

I understand that curing can't be replaced as a means of spiking hate, especially when it gives you the flexibility of choosing to do so at practically any time, but I don't see how that makes it better or more important than damage.

[Malacite]

Re: Hello Fellow PLD

A DD PLD will easily average more than 130 damage (I would say 200) on his Vorpals and WS at least once a minute.

I've popped almost 600 VB in EXP with a BRD.

[ItazuraNhomango]

Re: Hello Fellow PLD

I've popped almost 600 VB in EXP with a BRD.

For this discussion, Let's stick to the average numbers reported by the parser.

"How high of damage can Vorpal Blader get" is an entirely different kind of topic.

* * *

Obviously; this discussion is about tanking.

I thought the discussion was about what a PLD should do (well, originally, how should he gear) to earn his slot in the party; tanking is only an mean to an end.

15% Haste (spell) = 37 sec Flash cooldown and 2:33 Reprisal cooldown. Assuming these spells get used as much as possible, combined they consume 50 MP per minute on average. Assuming only Refresh, that's 4 MP/tick on the PLD, which is 80 MP a minute. A PLD can only spend 30 MP on Cures per minute without draining their MP. That's only ~130 HP per minute with Hospitaler's factored in.

Not that I doubt your math ability, but sometimes, actual in-game data to inform the theory crafting is helpful.

Here are five exp parties I have parses for with PLD's in them (the last 5 parties I found, anyway). With the exception of the last one in the list, they were all pickup groups. (The PLD in that party is a friend of mine; the rest were flag up people.)

2010.02.20 (SCH58)
PLD Cure 36775
Fight Time: 3:52
# Fights 167
HP/min 158.51
HP/fight 220.21
PLD Tyiara
DNC Pipsqeek
Note: Beyond Aspir Samba, I don't remember how Pipsqeek performed in this party, but in the one or two times I had Pipsqeek in Dynamis, s/he was quite diligent about curing.


2010.01.03 (SCH60-61)
PLD Cure 24532
Fight Time: 1:54
# Fights 43
HP/min 215.19
HP/fight 570.51
PLD Semper
RDM Anthenia


2009.11.27 (SCH68-69)
PLD Cure 25715
Fight Time 1:56
# Fights 91
HP/min 221.68
HP/fight 282.58
PLD Way
BRD Guland


2009-11-21_001 (SCH69-70)
PLD Cure 21119
Fight Time 2:01
# Fights 87
HP/min 174.54
HP/fight 242.75
PLD Slyth
Notes: No MP recovery support. IIRC, Slyth used a Dark Staff.


2009-08-16_001 (SAM61)
PLD Cure 39815
Fight Time 1:33
# Fights 50
HP/min 428.12
HP/fight 796.30
PLD Ayumiko
RDM Ninjintsu
Notes: Think we had a problem with this RDM not curing and maybe other issues. Ayumiko always carried a Dark Staff, though I can't recall how often it was used for this party; it's been too long.


The numbers are all over the place, but looks like with MP recovery support, the average PLD cures over 200HP per minute of fighting should be considered normal.

I don't like Level Sync my PLD, unless a friend desperately needs a tank. Moreover, I didn't save most of those parses, but here's one I found:

2009-07-20_001 (PLD57)
PLD Cure 9389
Fight Time 0:32
# Fights 26
HP/min 293.41
HP/fight 361.12
PLD Itazura
RDM Eclipes
Notes: Think I turned on the parser way after the party started.



A bit out of order, but:

Comparing DoT to Curing in this way is illogical since you do not Cure constantly. You can't do 190 damage in 2.5 secs through normal melee but you can do that much in the time you spend not curing. If you're going to divide WS damage by the time spent between WS, it's only fair that you do the same with Cures, which is a reasonable demand because your MP is restricted by your Refresh, in the same way your WS is restricted by TP

Fair enough. I'd like clarify one thing, though; I was operating on the greedy algorithm: At each point in time, I ask "What is the most profitable (i.e. most enmity generating) next move?" For the condition I set, MP available and 200+ HP missing, Cure III looks like the best option to me. (Aside from Provoke and Flash, that is.)

Which, then answers this:

I understand that curing can't be replaced as a means of spiking hate, especially when it gives you the flexibility of choosing to do so at practically any time, but I don't see how that makes it better or more important than damage.

"Because with Cure III, I can have more enmity sooner."

[Yellow Mage]

Re: Hello Fellow PLD

Not that I doubt your math ability, but sometimes, actual in-game data to inform the theory crafting is helpful.

Yeah . . .

Note: Beyond Aspir Samba, I don't remember how Pipsqeek performed in this party, but in the one or two times I had Pipsqeek in Dynamis, s/he was quite diligent about curing.

Notes: No MP recovery support. IIRC, Slyth used a Dark Staff.

Notes: Think we had a problem with this RDM not curing and maybe other issues. Ayumiko always carried a Dark Staff, though I can't recall how often it was used for this party; it's been too long.

Notes: Think I turned on the parser way after the party started.

. . . helpful.

I'd like clarify one thing, though; I was operating on the greedy algorithm:

Usually when I see the greedy algorithm used, it's to explain how bad the greedy algorithm is.

[ItazuraNhomango]

Re: Hello Fellow PLD

Did you even look at the numbers, YM? lol. Even the shortest parse I listed had 26 fights. If you have actual criticism instead of a vague, meant-to-be-sarcastic comment, say it, and I'll do my best to address your concern.

There's nothing wrong with Armando's math, except that it underestimates the amount of curing PLDs were able to do in the parties I've had.

One explanation is that there were "down times"; taking the "2009.11.27 (SCH68-69)" for example (because it has the largest number of fights in the parse), the "Total Fight Time" was 1:56:44 (which I truncated to 1:56 earlier), while the "Party Duration" was 2:31:53 (which I didn't publish earlier).

Now, this is an Heraldic Imp party, at a camp which is characterized by short fight time and long pull time (esp. when there are competitors). Going over the "Fights" data just now, I don't see any especially large gaps between the fights, so the pulling was never stopped. i.e. As a normal operation at this camp, the PLD would get much more MP recovery than the Armando's scenario.

There's nothing nefarious about it; this is how an actual party works. Even in parties w/out a MP recovery job (like the "2009-11-21_001 (SCH69-70)" example), you can see the PLD handily exceed Armando's estimate (which assumes a party with Refresh) on the maximum curing doable by a PLD.

So, the problem is not with the math, but with not accounting for the way people actually play in the parties--such as getting a bit of time between pulls to recharge MP and snatching a tick of hMP here and there with Dark Staff on. Or, putting it another way, the math is OK (I came up with 32MP/min leftover with full utilization of Flash and Reprisal myself), but the modeling is off.

* * *

Armando states a 4 MP/tick MP recharge rate with a RDM; i.e. Auto Refresh, and Refresh. With Sanction Refresh and Parade Gorget, though, it's possible to hit up to 6 MP/tick, but it's difficult to maintain the HP for the Parade Gorget. While the Sanction's 1 MP/tick also isn't available at high MP, it is pretty easy to trigger for a heavy MP user.


That said, I've rarely had a RDM who's perfect with keeping up Refresh, so maybe a 4 MP/tick average MP flow isn't too stingy of an assumption.

* * *

Usually when I see the greedy algorithm used, it's to explain how bad the greedy algorithm is.

Well, if you're looking for the optimal solution for a graph problem, probably not the best thing to use.

The problem here is "How do I stay ahead of of the DDs' TE as much as possible, given that I can't fully predict how much they will increase their TE in the next step?"

Give me your algorithm.

[Yellow Mage]

Re: Hello Fellow PLD

Did you even look at the numbers, YM? lol. Even the shortest parse I listed had 26 fights. If you have actual criticism instead of a vague, meant-to-be-sarcastic comment, say it, and I'll do my best to address your concern.

Here's my criticism - you appear to have gotten to the point where you're doing little more for your arguments than hiding behind copy-pasta'd walls of anecdotal numbers which I (and I'm sure most other forum-goers) don't want to even bother trying to crunch. We're not all Armando over here, you know.

The problem here is "How do I stay ahead of of the DDs' TE as much as possible, given that I can't fully predict how much they will increase their TE in the next step?"

Give me your algorithm.

Well, since this discussion is obviously about tanking, I would hope it would be safe to assume that the DD's aren't going to go absolutely "balls to the wall" without knowing for certain that it would seal the mob's fate within an amount of time too short for it to cause any harm. In other words, you have it backwards: in a party with a tank, it is the DD's obligation to stay under the tank's hate line at all times (lest they learn that a dead DD does no damage), not the tank's obligation to stay above the DD's. The tank's goal is simply to keep that line up as consistently high as possible, and I can guarantee you that they will have a heck of an easier time accomplishing that if they are not gimping their damage output.

Heck, even in the worst case scenario of the Souleater-crazy Dark Knight, you could always just save your Cover ability for when they decide to pop it (Cover's recast is 5 minutes, versus Souleater's 6 minutes). Odds are the mob is going to die within the next 15 seconds, anyway, simply from the extra DPS that suddenly isn't very problematic anymore. They'll really only be problematic if they want to be uncooperative, in which case they literally have a death wish. Not to mention, hey, free HP to cure for spike hate, amirite?

Curing is an important, even integral, part of the Paladin job - only a nutcase would deny that. What too many people don't realize, however, is the Paladin's own base damage output is at least as important. Even if nothing else, it at least gives all your other many hate-generating Abilities something consistent and significant to build upon.