Why the -PDT on amano samsFollow

Hopefully the third time is the charm.

I pointed out he doesn't have the same EVA as a THF, and comparing naked to geared/meritted jobs is non-sense. A fully geared, meritted and fed SAM can get more Acc with a polearm than a naked, hungry DRG. And that matters why? It doesn't.

So? "Often" "regularly" are completely meaningless. Comparing to a naked THF is completely meaningless. For example, the best -dmg% we can get in hands slot is -2% physical, if you can put something there that raises SAM's evasion more than 2% it's better, period.
Parry has a really low cap, it's almost always better to go eva unless you're trying to skill up parry.

Often means just barely above floor. Regularly means a little less than that. Comparing to a naked THF is NOT completely meaningless, because SAM can't reach values much higher. A SAM fully geared is akin to a naked THF on the evasion front. Anything above EXP, your evasion is going to be floored. It takes a sizeable amount of EVA gear to raise JUST beat the floor. You would have to sacrifice the best parts of your -PDT set to do that. Thus you won't find a hands piece that will raise your evasion more than 2% without sacrificing much more effective -PDT.

Also, with -PDT being a cumulative % decrease in damage taken, much like haste is to delay, as you increase your -PDT, further increases become more potent until you reach the cap. That's to say, if you have -48% PDT already before you get the hands piece, that -2% PDT is actually going to be more than a 2% reduction in damage taken.

You can absolutely break the floor on evasion for a lot of stuff as SAM, but doing so isn't as worthwhile as -PDT gear.

*edit* you seemed to misunderstand my whole "playing off my job's strengths" comment. See, parrying skill doesn't have a convenient trait to enhance it. 269 is the max before merits. However, there's just as much parrying gear available to SAM in the same slots that I would warrant putting evasion gear. With the exception of a subjob that offers evasion bonus, you'd see better results from parrying gear w/ capped parrying *especially if merited* than evasion gear while going full -PDT. You won't be hitting the parrying cap regardless.

Edited, Feb 18th 2010 10:05am by Banggugyangu

*editted*
Nvm, its too early, im going to bed -_-

Edited, Feb 18th 2010 6:08pm by hitoseijuro

Yea but who have -50 pdt?

Let's keep it on Nq gear, arhat head -4, arhat body -6 but osode isnt better? 14 more base def plus vit and agi. Ring combo another -7, feet/legs combo another -5, that's 22 assuming you use arhat over osode, where you get the other -28? even with the Hq you're not getting closer.

In my case, i have jelly/patronus, switching the remaining to eva/parry i get 22 eva, 10 parry skill, 5 eva skill in exchange of -15 pdt, so wich is better now?

And I noted that he never said he did. He simply said that he had as much EVA skill as naked THF.


He said, "I can get enough evasion skill for it to make a difference, as much as a naked THF, even." You said, "You don't have the same EVA as a THF." Do you not see the non sequitur in your response?

It matters is a naked, hungry DRG could hit the targets that you are trying to hit with your "fully-geared, meritted and fed SAM." His point was just that he could get evasion fairly high, high enough to make it useful.

Tell me... why is it not safe to assume that a colibri's accuracy would not be a similar number to its evasion? The only listed jobs for colibri family are Red Mage, White Mage, and Black Mage. None of these jobs have ANY offensive combat skills that are lower than their evasion skill. This means that if the mob's DEX is anywhere NEAR its AGI, then the accuracy should be a similar number.

I love how you live in a magical world where people can have an evasion rate that ignores the floor completely....
No... What I did was show a more accurate representation as per the evasion floor that is hard-coded into the game. Your foolish example ignored it. Mine did not.


You know how haste has an increased rate of return the more you stack on? -PDT follows the same rule...
Without sacrificing your weapon, it's simple to put together a -31% PDT set. If you are lucky enough to acquire a Defending Ring, you can achieve -41%. If you need to sacrifice your weapon, you have 2 options: Iron Ram Lance *decent for a defensive weapon* adds -10%. Earth/Terra's staves *obviously only if you absolutely NEED to* add 20%. It's possible to reach the cap with easy-to-acquire gear when your damage output isn't priority. It's possible to reach the cap and still be able to deal a moderate amount of damage simultaneously when you add a single piece of hard to get gear. Outside of both of those situations, you can hit 31% easy with any weapon, and 41% easy with a small sacrifice.
Hi, moron... Obviously I was talking about hachiryu feet vs askar feet. For that matter, I would pick hydra feet anyway, unless I was in the lucky few to get a defending ring, where even w/ iron ram lance I would be 1% above the cap.
"enough to best me"? You obviously don't understand how chance works. The goal of a defensive set is to reduce the avg damage you take as much as possible.
Usukane hands do it better.
End of story.
[/quote]
The goal of a defensive set is to make surviving more of a possibility. The effectiveness of each set will change depending on the situation you're in. If you're soloing, then by all means, do whatever is going to reduce the total amount of damage you take over the whole course of the fight. *arguably unless you're /DNC for the same reason as the next situation* If you're in a situation with a limited healer, however, you're best off with decreasing the amount of damage you'll take with each hit. If you get hit a few times less per hundred, but you get hit for twice as much or almost twice as much, then your healer may not be able to keep up. /DNC healers suffer from this because they're limited to ~200 HP per 6 seconds. If you're in a situation with strong healers *RDM, SCH, WHM, DNC*, then it's a toss up, honestly. The piece with the largest difference between evasion and -PDT is a very negligible difference. I doubt a WHM is going to complain that their SAM tank is using usukane over darksteel mittens and taking 5-10 more damage per hit. I doubt a WHM is going to complain that their SAM tank is using darksteel mittens over usukane and getting hit 5 more times out of every 100.

btw...

This is what I've been referring to the whole time. Generally, the primary purpose of this type of set is when you unintentionally pull hate and don't intend to hang on to it too long *let's say 1-10 attacks*. In this situation, you don't have the gigantic sample size that you would love to use to support your example. In fact, let's use SAM's wonderful evasion skill in this situation.

The mob's accuracy is 320. It will deal 100 damage with its attack:
You pull hate and before the attack connects, you switch to 15% physical damage reduction. You hold hate for 3 attacks and take 255 damage.

You pull hate and before the attack connects, you switch to ADD 15% evasion to the floor. You hold hate for 3 attacks. Odds of getting hit by all three hits: 51.2%. Odds of evading a single attack: 45%. Odds of evading 2 attacks: 4%. Odds of evading all 3 attacks: .8%
Probability states that you will get hit 3 times for a total of 300 damage.

This is a moot point for anything above EXP mobs, where you are most likely never going to even break the eva floor.

For sam's that only carry around tp/ws gear i say is quite expensive, and most of them are like that. I have tp/ws/penta/ranged/eva/parry and spending money on something wich i consider a downgrade or in best case similar is such a waste.

The 22 eva and 5 skill was meant in 3 slots.

But, you're missing one thing, o-sh*t gear is probably used for something like 2-3 seconds before utsu/eye is up again, so you get 1-2 hits.

Now i prefer the chance of a miss or 2 instead of getting hit for 20% less, saving 20-50hp on a hard hitting mob is not such a big deal.

You're either ignorant of probability and the way it works, or ignoring nearly everything I've said in this whole thread...

Evasion rates are based on probabilities. If you have 70% evasion, that means that when that mob attacks you, there's a 30% chance that he'll actually hit. over time, this will average out to 30% of the time the mob attacks, he hits you. Note about probabilities: The smaller a sample size becomes, the more skewed it becomes toward the larger probable outcome. I.E.. if you have 100 marbles in a bag, each with a successive number, and you pull out a single marble above 13, probability states that you will do it. Each time you increase the sample size, the chance of pulling out a marble above #13 remains the same, but the chance of having a result which is different increases.

Like I said... At best, you're weighing 13.5% evasion vs -15% PDT. This means that you will have at best 18.5% evasion rate with those pieces of gear. This means that you have an 81.5% chance of being hit. In the 2-3 attack sample size, probability states you won't evade. In fact, it would take a sample size of 6 attacks or more for probability to state that you SHOULD evade at least one. But... like I also said... You're not gaining 13.5% evasion from those pieces of gear. A more reasonable estimate is somewhere around 5-6% evasion. I would much rather take my chances with floored evasion rates and take 15% less damage, than to have a very mediocre evasion increase and take the full damage.

Ouch epic fail. Every evade attempt is independent. The game doesn't care whether you evaded the last 10 attacks in a row, the 11th attempt still has the same probability to evade.

Both evasion and -PDT have their place. Basically you have two variables to consider: mob hit rate and damage per hit. The total damage you take is the product of these numbers. You modulate hit rate with your evasion and you modulate damage per hit with -PDT. Your ultimate goal is to minimize the product of these numbers. Do this and the probability works in your favor. Who cares about sample size? Are you only going to take 1 or 2 hits total over your whole FFXI playing career?

Just to show you that you should go back and check your stats. You define the probability of getting hit as 81.5%, and claim it takes 6 attacks to expect an evade. Actually, you should look at it like this: What is the probability of getting hit 6 times in a row with no evasion? 0.815^6 = 0.29. Only 29%. In fact, your odds of getting hit 4 times in a row are only 0.815^4 = 44%. In other words, more often than not you will get an evade within 4 hits if you have an 18.5% evade rate.



The bolded statement makes absolutely no sense. Results don't get skewed either way. OK, say I have a 30% evasion rate, I could take 1000 5-hit samples. Some of these samples I will evade 1 time or less (20% rate or less), some of these samples I will evade 2 times or more (40% rate or more). Average them all out and it will come to 30%. You should err above your average just as often as you err below.