Food choices and Defender

[ItazuraNhomango]

Re: Food choices and Defender

Store TP +5 is actually +5%. Double Attack +2??? how do you add +2 to a number that is figured in %? I still haven't figured out that one. Subtle Blow is another trait that, even though gear offers it in static, un-labeled amounts (Subtle Blow +5), it actually gives a % decrease in TP to the opponent. Things like that - and really, common sense of what would serve best to "balance" a game - tells me that it has to be at least partially % based.

You make a good point; there are plenty of what looks like percentage calculations in this game. Still, I think the programmers would go pretty far to avoid float point operations. Also, the items you've called attention to have different properties, which allows for different means to reduce or avoid entirely those dreaded float point calculations.

Double Attack doesn't actually need a "base" to multiply off of, and can be implemented as an integer threshold. (Random roll; if under threshold, proc.)

Store TP only needs to be account for when calculating TP per hit, and that only changes on equipment change, job swaps, or gaining a new level of "Store TP". Otherwise, how much TP gain from successful attack is a constant, AFAIK. (i.e. Calculated once, then used many times until getting a "discard" event and recalculate.)

Subtle Blow calculation can use the same shortcut as Store TP.

However, the situation is trickier for "Enmity multiplier"... To start with, it doesn't have a nice, nearly fixed base number as "TP per hit", since amount of enmity gained depends at least in part on how much damage was done. (For this purpose, it's assumed that the exact damage to be done on a particular swing cannot be determined ahead of the time, hence no fixed base.) That means the tricks I would use for Double Attack and TP type calculations won't work.

I suppose it can be implemented as a large lookup table (like the engine control for cars), though I doubt it, since memory space is a premium on servers, and a cache miss is much more expensive than integer operations.

Maybe implemented as two step operation, integer multiply then divide? I guess it's possible, but integer addition/subtraction is still better than integer multiple and divide. (Can shortcut the division with shift operation using 2^^n as divisor, however.)

I dunno... w/out experimental data, I'm not ready to believe the programmers would use a multiply operation over straight addition and subtraction. (And, no, I've no clue on how to perform an experiment sensitive enough to tease out the difference between enmity+5 and enmity+15.) I would admit that after thinking it over, enmity multiplier wouldn't have as bad of a performance hit as I thought originally, if using integer multiply then divide.

* * *

I'm getting enough of both "I didn't use Defender full-time" and "I used Defender full-time" posts to make me think both methods are quite workable. Since Bark Spiders at Lv.50 is over hunting slightly, I guess I should step up things incrementally, starting with Defense Food + Defender, and change tactic and food according to the situation.

That just means inventory crammed full of "alternative" food and gears... So much for making out like bandits off spider webs... /sigh

Thanks everyone for the responses. ^_^

[Radel-King_of_Pirates]

Re: Food choices and Defender

That just means inventory crammed full of "alternative" food and gears... So much for making out like bandits off spider webs... /sigh

Yea I think everyone knows that feeling, most inventory used for spare gear, weaps, and different foods. And then only about 10 spaces for stuff to make gil on.

[Lmnop]

Re: Food choices and Defender

10? The most I've managed is 6, and I regretted leaving 2 pieces of equipment behind for that. ._. Oh, and I still don't have a gear swap for Weapon Skills. Life will be fun when that time comes...

[Maju]

Re: Food choices and Defender

Basically, I wanted to say that job level is not the determining factor on whether sushi will outperform meat, but it came out wrong.

This is completely wrong. When leveling low level jobs, meat always destroys sushi. Sushi simply does not grant a noticeable increase in accuracy until high levels while meat will even double your damage per hit at low levels.

This was painfully obvious when leveling my nin as I often had sushi eating nins in my group and I'd easily melee for double their DoT.

[Armando]

Re: Food choices and Defender

Hm. Those are some very interesting points on Enmity. I think there's a way we could get some sort of answer, though. Spells do static damage. If we were to do tests with one person doing more damage than the other, but casting less often, and with different ammounts of +Enmity, a trend would eventually have to show. Of course, to keep the test clean we'd have to get a third person to initially claim the mob (I have strong reason to believe the first person to aggro a mob gets more hate,) and some way of keeping the mob from hitting the testers for damage that doesn't involve Utsusemi (since that lowers hate.)

[ItazuraNhomango]

Re: Food choices and Defender

This is completely wrong. When leveling low level jobs, meat always destroys sushi. Sushi simply does not grant a noticeable increase in accuracy until high levels while meat will even double your damage per hit at low levels.

This was painfully obvious when leveling my nin as I often had sushi eating nins in my group and I'd easily melee for double their DoT.

*shrug* I out damaged other NIN's back in dual tanking levels no matter what food I used. (That includes Dorado Sushi, Rice Dumpling, Dhalmel Pie, Navarin, and even the very icky Jack-o'-Lantern.) You're probably just a good player with good gears.


At level 30, I've always thought appropriate meat food will out do sushi. (This is an important level since I do a lot of Lv.30 capped Promyvion runs for my LS.)
What woke me up was a WAR/NIN (no merit anything) in my LS using Centurion Sword x2 and Sole Sushi out damaging my WAR/NIN on Great Axe and Rice Dumpling. Now, I thought I was out damaging him, especially with my impressive Sturmwind numbers--but the parser said otherwise.

Later, I did a run Promyvion-Holla run with two other WAR/NIN's while using WAR/THF myself. Even discarding the WS damages, my melee damage was a higher than both of them. One was using sushi, but I had better gears. The other had good gear, wielded axe and Centurion Sword, but was using rice dumplings. I had good gear, Centurion Axe, and sole sushi.

Since the weapons and subjobs are different, it's no "proof", but a good indicator that sole sushi is probably at least as good as rice dumplings. Sushi certainly didn't get "destroyed", and added lot's of Sturmwinds which pushed up my damage numbers even higher.

* * *

Do not trust what you feel--run parsers. People tend to find evidence for what they believe or feel is right, instead of what is real. (If I had gone with my feelings, I'd still think meat food is always wonderful at low levels.)

Run parsers, and you just may see very strange but true things--like a Lv.33 PUP out damaging a nicely geared WAR/NIN with 10% higher accuracy and a BLM addicted to melon juices and pies in an exp party. (Or, RDM/NIN out damaging WAR/NIN on 1st floor of Promvyion when both are using Centurion Sword x2, for that matter.)

[Maju]

Re: Food choices and Defender

Do not trust what you feel--run parsers. People tend to find evidence for what they believe or feel is right, instead of what is real. (If I had gone with my feelings, I'd still think meat food is always wonderful at low levels.)

I did run parsers and I was always way, way ahead of the other NINs and I often beat other melee as well, though I wish I still had the parser logs somewhere so I could check the exact numbers. I always filter other people's misses so I never got any parses of their accuracy, but it seemed like it was no different from my own even though I was eating meat.

It actually caused some problems occasionally as the other NIN I would dual tank with would suddenly be unable to voke the mob back from me because I outdamaged him so badly.

[ItazuraNhomango]

Re: Food choices and Defender

Hmm. Like I said, my experience on NIN was "always out DD'ed other NIN, and did that on every food I tried." Pretty much same as yours, except I know I tried different food. (Not sole/squid, but did use dorado sushi a lot.)

Did you try sushi and received bad results? If you only used meat, I'm more inclined to believe you're a good player with decent gear instead of meat helping you much much more than sushi in the Lv.30's.

Anyway, the evidence I have tells me sushi works well for my WAR in Promvyion-Holla and Promvyion-Mea, and those are level capped to 30. I've tried: Meat Mithkabob, Dharmel Pie, Rice Dumpling, Sole Sushi; they all work, just I seem to get best result from sole sushi.

This doesn't make me a diehard sushi fanboy, though--I've seen other people use sushi for Promvyion-Dem, and was not impressed (nor amused). For that boss mob, meat is the way to go for melee's.

Sushi's are situational food, like every food.

I guess we can just repeat our own experiences until we're both blue in the face or something, then call it a draw. ^_-

[Celeal]

Re: Food choices and Defender

IMHO, the food choice for Promvyion runs is not necessary the same with traditional exp. points party.

At lot of those mobs in Promvyion have buffs like evasion-boost, blink, etc. At level 30 cap no one has Dispel or Finale too.

The difficultly of the Promvyion mobs in the 3rd floor or the boss is harder than those popular mobs in exp. party anyway. Those IT mobs in Promvyion is different from those mandies and goblins in the Jungles. Usually traditional parties hunt weaker mobs like carbs, crawlers, etc.

Food choice should be determined by the situation.

[Maju]

Re: Food choices and Defender

Did you try sushi and received bad results? If you only used meat, I'm more inclined to believe you're a good player with decent gear instead of meat helping you much much more than sushi in the Lv.30's.

I did eat a few sushis when I ran out of dhalmel steaks and was always disapointed by it. I didn't hit noticeably more often and my damage per hit dropped by roughly 50%. I've only eaten perhaps 2 or 3 sole sushi on NIN, but the difference really seemed quite massive.

It always made me squirm a little inside seeing the other NIN whiff just as much as me and barely break 10 damage per hit when I'm easily dealing ~20 and nearly 30 per hit with Berserk and then doing Retsu for 200 damage when theirs can't even get close to 100.

You're right in that I did indeed always outgear the other NINs but the stat differences weren't exactly massive as this was all low level stuff.