BLM, INT vs. MAB

[Suiram]

Nothing SephYuyX wrote should be taken as actual equalities, aside from probably 1 skill = 1 Macc as best I understand it. The others are guidelines he is putting forward, some of which are more reasonable than others, but he doesn't make that clear.

[SephYuyX]

Whats not equal?

1 INT is the same as 1 or .5 MAcc/Skill
2 INT is the same as 1 MAB
1 MACC is the same as 1 Skill
1 Skill is the same as 1 MACC (.9)
3-4% Magic Critical Hit is the same as 1 INT worth of DMG.

All tested and proven.

[Suiram]

1 INT = .75 Macc on average only if it equals 1 Macc half the time and .5 Macc half the time, which is not a reasonable assumption to make.

2 INT has never equalled 1 MAB, like I said that's a guideline you're putting forward, take it for what you will.

1 Macc = 1 Skill I gave you (although you sneak in a ".9" there for some reason, which even if that were true would contradict the line immediately above it)

3~4% Magic Critical Hit = 1 INT is also a guideline, except worse because it doesn't capture the fact that 3~4% Magic Crit is 5.5%~6.5% increase in damage 3~4% of the time while 1 INT is a 0.25%~0.4% increase all of the time.

So like I said before, the only one you can reasonably state as an actual equality is 1 skill = 1 Macc. All tested and proven.

[SephYuyX]

K. Ill take Kanican/Kaeko's word over yours.

[Suiram]

I'm not even sure how to respond to that lol

[SephYuyX]

Those are exact lines he's posted in various posts on his LJ, so im not quite sure what youre making out of them.

[Suiram]

Sorry, I'm still having a difficult time putting together a response. What I mean to say is something like this: If Kaeko wrote what you wrote, and said they were not guidelines but in fact actual equivalences, he was wrong. You'll have to find the blog entries for me that say that, because there's a lot of it I have never read like the BLM guides.

But, knowing Kaeko, he probably did say they were guidelines, and if he didn't it was in an uncharacteristic lapse of meticulousness that he'll likely correct here should he read this thread. But even if he were to mistakenly believe that the above are actually equal (which again, I seriously, seriously doubt), then this would be a debate for he and I to have, and not you, since he generally knows what he's talking about, I generally know what I'm talking about, and you are clearly someone just regurgitating something you read on somebody else's blog.

[Kaeko]

"1 MACC = 1 Skill"

Tested using Elemental Skill by Lodeguy, proven with 95% confidence that the 0.9 to 1 thing does not apply. Do not use the 0.9 rule the melees use for their skills. Also, he tested with Elemental Skill and we simply extrapolate this to other skill sets, which may or may not be accurate. I originally posted the 1 to 1 or 1 to 0.9 could be true in my post; however, a later post by Robonosto proved the 1 to 0.9 to be false; hence we should say 1 MACC = 1 Skill and vice versa until contradictory information is presented.

Also, I'm not sure what this statement means: "1 MACC is the same as 1 Skill, 1 Skill is the same as 1 MACC (.9)" other than the fact that they both cannot be true ;o

"2 INT = 1 MAB"

Guideline. MAB has diminishing returns. For example, your BLM will likely gain less from +1 MAB than your SCH due to higher MAB traits. You need to plug into the formula to truly know what is best, but w/o the calculator, you can GENERALLY use this as a guideline.

1 INT = .75 MACC/Skill (.75 is an average it is 1 or .5 depending on INT difference)

Most mobs have INT in the range of ~60-70. An Ebony Pudding, for example, has 90 INT. Because of this, adding 1 INT is more than likely going to only give you 0.5 skill (or less) on most mobs. There are very few exceptions such as BLM type mobs or Wyrms. It is not something you should just blindly assume to be 0.75.

In fact, Robonosto even pokes fun at people who say 1 INT =~0.75 or similar things to this:

The unbearable triteness of preening: One more time

Also, Lodeguy only establishes the single tier at dINT=10. He does not exclude the possibility of other tiers at different dINT values, for example, there could be another tier at say dINT=50 where 1 INT = 0.25 MACC (random, unfounded example).

If there is any part of my site, or any part of a post I've ever made that was not clear or suggests an oversimplification, please tell me or link me. I'm pretty sure I was fairly clear on the SCH post everyone cites for MACC info though. I try to be as specific as possible with these "guidelines". Things are rarely as simple as people want them to be.

Also, Suiram basically co-wrote the page on BG Wiki on Magic Damage. If we were purely to compare knowledge of the subject and reputation in this area, I would defer to Suiram, not the other way around.

EDITS:
Pudding INT
BG Page

[Sifer]

K. Ill take Kanican/Kaeko's word over yours.

and I'll take suiram's word over yours.

2 of the lines you wrote contradict each other and the 2 int= 1 mab is as bad as the nin that tanked wyrms in mnd gear cos "2 MND= 1 MDB".

edit: beaten by kaeko ;o

[Argettio]

Owns those that missquote him/his blog

All of what he said^

The 2 INT = 1 MAB guidline also varies with the Tier of spell, it works very well on T4 against merit mobs. But for T3 or AMII it changes. The lower the based damage, the greater the INT effect becomes in percentage terms.

FF XI - Magic Damage Calculator is very good, it is based on the testing that forms the 'Calculating magic damge' page on otherwiki and is incredibly accurate (within the limitations of rounding errors etc).

[SephYuyX]

;o Id figure that quip would bring Kaeko out.

/shrug. I and others have posted those exact lines before over the past year, even when comparing Genie vs RR, and no one has said they were wrong or corrected them even when I asked.

So, nice to be clear now. It's basically correct, just worded poorly, obviously trying to shrink it down to being as simple as possible.

[Spekkio]

there's no possible hard and fast rule on int vs mab, nor is there with macc vs int.
mab's effect on damage varies with the target's MDB. if the target has 100 mdb, the gain from adding 10 mab is going to be significantly less than a mob with 0 mdb or even a weakened MDB caused by some effect (at least unless i screwed the math up.)

meanwhile int's effect depends heavily on the target. testing on wild rabbits should almost invariably result in a win for MAB as your int should be capped on them naturally. IE stacking more int is about as effective as panning for gold in your shower.

there's no "int is better than MAB" or vice versa rule out there, it depends on the target. for example, if nuking tiamat, ouryu, etc. i'm going to lean more on my int, ele skill, and macc gear (eg witch sash over sorc belt, ele torque over prudence/uggalephi, af+1 hands over morg hands, etc.) meanwhile if i'm nuking cupcake mobs like dynamis mobs, i haul out the heavy mab, with a side of int for some major damage. if you want to better understand it, use the calculator posted above with yigit hands vs zenith hands and see how the damage changes as you change the target's int/mdb and your int/mab. you'll learn a lot.

edit: and yes, i've written up much more in depth diatribes about why rules 1 and 2 are wrong but i'm at least trying to phrase this in a way the OP will be more comfortable with, avoiding terms like dint, inflection point, etc.

[Argettio]

/shrug. I and others have posted those exact lines before over the past year, even when comparing Genie vs RR, and no one has said they were wrong or corrected them even when I asked.

No, I and other argued with those people who said INT = 0.75 skill when comparing Genie vs RR (I forget who exactly said it)

And we agreed in princile with the 2 int = 1 mab approximation, but went on to far more in depth analysis comparing them for a range of INT and MAB values on T3, T4 and AMII spells.

[Suiram]

Anyway, on the INT vs. MAB question, I have been thinking recently of a way to write up a method for making choices for that which mirrors essentially what I have done for my own gear choices, but for the less-math-inclined, which I alluded to the haste/acc post recently. It should serve as a better guideline than 2 INT = 1 MAB, if people want to use it. But I have to go out for a few hours so I'll write it up when I get home.

[mcb]

grabed a few pugil
weak to thunder so thund 4
int = 1279 WIN
mab = 1293

how is this win when top dmg was win for others?

[Dominious]

Thunder IV has higher base damage than the others, so MAB is more effective for that than it is for lower tier IVs, it's easy to see if you compare something like errant body vs yigit body on all of the tier IVs. For my SCH errant body wins on the lower 3 because INT increases the lower base damage, and yigit wins on the higher 3 because the MAB. The INT would add the same amount to the base damage of both, but in proportion it's smaller compared to the base damage of thunder and blizzard than it is to stone.

[Absolutely Virtue]

Is it possible we could try to make the rules of thumb a bit more direct? I fool around with my gear at events when we fight mobs we don't normally fight, swapping pieces out (usually when I'm pissed about resists). But I have to confess I don't care enough to record the data, and I never really learn anything in the long run. I love FFXI, and I love to play BLM, but not enough to show the dedication that you guys show.

Is there any interest in a chart that says for "X" mob, aim for the following ("A" MAB, "B" INT, "C" skill, for example) targets? Puddings are an obvious good choice, but highly resistant mobs could be another. The idea being you're looking for goals with your stats, not the gear you use to get there.

You can't hold everyone's hand and play the game for them, but I think these kinds of heuristics would be of great benefit to people who do try to play the game properly, but aren't inclined to have the magic damage calculator and FFXI client open at the same time. I'm thinking of a compromise between nuking in full AF and Kaeko/Suiram.

[Suiram]

Okay, so as I was saying before, this is a simplified version of what I (and I think other mathematically-inclined people) do when they are trying to decide between INT or MAB in a particular situation. It's simplified in that, in order to make it less cumbersome for everyone, it makes a few assumptions.

I assume that a person is using T4 or AM2. The calculation for T3 would be different since the rate of damage increase as a function of INT is different. If someone really needs to know T3 and can't figure it out, I can adjust it and make another thing for that, too.

I assume that you have more INT than your target. This is reasonable, I think, because the handful of situations where you may not, I think other considerations will take over priority, like INT/Skill/Macc since those mobs tend to be more resistant.

I assume that you're just trying to decide between two pieces, one which has more INT, and the other which has more MAB.


This relies upon the fact that the relative damage increase from adding MAB depends only on the current MAB of the caster, and that while the relative damage increase from adding INT depends on more factors--the caster's INT, the target's INT, and the base damage constant of the spell in question--there is a reasonable range of possible base damages you can expect to see for any given spell.

So here goes: First, total up your MAB. To start, you get MAB from job trait (+32 for 75BLM, +24 for 75RDM and 37BLM, +20 for 37RDM, or from BLU job trait). Then, if you're a BLM, you get 2 for every potency merit in a particular element. Then add on the stuff from gear that says "Magic Atk. Bonus". Include the calculation here the MAB from the piece with less MAB but more INT that you're trying to compare, if it has any (like if you want to compare Yigit hands vs. Zenith for instance). Call this total MAB, "LOWMAB".

You can ignore the bonus from HQ staves, MB bonus gear, day/weather bonus, all that stuff. For this, you just want the total MAB for this comparison.

Now, repeat the same calculation above but with the MAB from the armor with higher MAB and lower INT. Remember to deduct off the piece you are replacing if both pieces have some MAB on them, like in the hands example above you'd subtract 2 from taking off the Yigit hands before adding 5 for the Zenith Mitts. Call this number, "HIGHMAB".

Finally, take the amount of INT on the High INT/Low MAB piece and subtract off the INT--if any--from the Low INT/High MAB piece, and call this number, "INTCHANGE".

Now, take the values you got for LOWMAB, HIGHMAB, and INTCHANGE, and replace them in the URL below:

http://www.wolframalpha.com/input/?i=(HIGHMAB%2B100)*x+%3D+(LOWMAB%2B100)*(x%2B(2*INTCHANGE))+solve+for+x

The website will give you a value for X. That value is the highest Base Damage your spell can have before MAB starts beating INT for damage. So refer to this chart below (dINT is how much more INT you have than your target):

          dINT 0     dINT 50    dINT 100
Bliz4       506        606        706
Thun4       541        641        741
AM2         710        810        910

I made the chart up to dINT 100, but really that's not realistic for most things you'd care about, but you get the idea. So, in the easy cases, if you get a number less than 500, then the MAB always wins. If you get a number bigger than, let's say, 820, the INT always wins. If you get a number like 600, then the MAB is better for AM2, but you might have to try both pieces for Tier 4 to see which wins if you don't happen to know the target's exact INT.


I just realized this post is megahuge, sorry. TL;DR: Add up your MAB with the low MAB piece, call it LOWMAB. Add up your MAB with the high MAB piece, call it HIGHMAB. Take the difference in INT, call it INTCHANGE. Plug it all into the URL, and refer to the chart to see which piece is better. I'll do an example in a minute.

[Suiram]

Okay, let's use SephYuyX as an example. This pic is from his blog from last September.

http://img255.imageshack.us/img255/3998/blmcapemm7.jpg

Now, suppose he was fighting something he didn't want to use AF2+1 hat for, and wanted to know if AF1+1 was better or worse than Yigit hat. First we total up his MAB. Assuming he has 5/5 potency merits in the spell being used, his total MAB before head is 32+10+8+7+5+6+5+10 = 83. Since AF1+1 hat doesn't add any MAB, LOWMAB = 83. Since Yigit hat is MAB+2, HIGHINT=85. Since AF1+1 has INT+5, and Yigit doesn't have any INT, INTCHANGE=5. So we plug that into the URL and get:

- Wolfram|Alpha

x=915

So since it's unreasonable that he'll have 915 base damage on something that he cares about, (would have to be casting AM2 on something he has over 103 INT on), the AF1+1 wins.

[Lahan]

grabed a few pugil
weak to thunder so thund 4
int = 1279 WIN
mab = 1293

how is this win when top dmg was win for others?

I obviously screwed that one up sorry.