Sunday, April 4, 2010

Questions About Magic Accuracy

Add this to the list of things I really should have just written a blog post about...

The original post from Alla, titled 1 CHR = 1 Magic Accuracy = 1 Skill?

And my response, beginning with a quote from the original post:
Lets assume we are against a target with high CHR (think HNM). So we are well below the CHR+10 mark. If this article is correct that means in this situation: 1 CHR = 1 skill right? In other words this article values CHR more then we currently do, am I right?

And the last time you met a BRD HNM was---?

You'll find that direct elemental resistance and innate song resistance matter more than relative stat levels in the HNM scene. No matter how good your debuffing setup is, you'll never land elegy on ouryu (ice threnody, however...). No matter how good your debuffing setup is, you'll have trouble landing elegy on ToAU kings without elemental seal.

Disclaimer Follows

WARNING: Asmoranomar's above analysis is essentially correct. Only keep reading if you'd like a more detailed conceptual comparison, and a bit more insight into why it may be worthwhile for you to pick Skill/Macc over CHR. My usual debuffing setup, for example, works with 89 CHR total (including merits).

Conceptual Analysis

So! Breaking down the logic of the initial analysis:

Tests show show that 1 magic accuracy = 0.5% magic hit rate below 50% hit rate, 1.0% increase after 50% magic hit rate.

it also shows 1 skill = 1 magic accuracy and for stats such as INT = 1 magic accuracy until total int is >targetINT+10. Each additional point of INT after that gives 0.5 magic accuracy. It is unknown if any amount of stats after that limit can give 0.5 magic accuracy. It is assumed the same goes for MND vs TargetMND and CHR vs TargetCHR.

We're talking about two different things here: "Magic Accuracy" and "Magic Hit Rate."

The goal of the test that wiki references was trying to model the relative relationship between Skill, Magic Accuracy, and STAT as stats that contribute to a general "Magic Accuracy" number.

"Magic Hit Rate" -- again citing otherwiki for kicks -- is apparently determined by a check between the caster's Magic Accuracy stat and the target's Magic Evasion stat. Magic Evasion is extremely important in any discussion about M.Acc.

Important points about Magic Evasion: It can vary by target and by spell. Magic Evasion is one reason you will never land paralyze on a UFO in sea, even though it's weak to ice-based magic. UFOs are arbitrarily immune (and/or have extremely high magic evasion) to the specific spell. Wiki uses imp resistance to sleep but not to gravity as another example.

So. What does this mean for BRD? Assuming the model is anywhere near true in a conceptual sense:

1. CHR = 1 magic accuracy until total CHR is >targetCHR+10. Each additional point of CHR after that gives 0.5 magic accuracy.

2. 1 magic accuracy (or 1 skill) = 0.5% magic hit rate below 50% hit rate, 1.0% increase after 50% magic hit rate.

Again we're not talking the same stats. Let's assume we have an HNM target with high CHR and a naturally high ME stat on Elegy.

In this situation 1 CHR = 1 Macc the whole time, because we never hit the >targetCHR+10 stat. HOWEVER, that still only contributes 0.5% magic hit rate overall -- because our general hit rate is still below 50%. CHR and Macc end up equalized by the pressure of high magic evasion. I have no idea if Khimaira has high CHR for example, but that doesn't matter because his magic evasion on elegy is so high! Elemental Seal directly increases caster Magic Accuracy by some amount (no one knows how much) -- but it makes Elegy pretty much unresisted, which is why bards /BLM at Khimaira.

Now consider pretty much anything non-HNM in the game, which is a relatively low-magic evasion target. Elegy lands on everything not immune. That means that MAcc and Skill are generally +1% magic hit rate ... while CHR is at best 1 M.acc and at worst .5Macc if/when you go beyond the >targetCHR+10 stat. Here's the kicker. That mob CHR stat can be tricky, and with level correction involved the Magic Evasion can become even more complicated.

A Chart of Sorts

Overall mob magic evasion to elegy and lullaby will tend to look something like this, in an extremely simplified conceptual way:

NOTE: The dash represents a range of values, as in "between .5% and 1% hit rate." It is not a minus sign.

High Evasion Targets (Immune): You're not landing it. Give up.

High Evasion Targets with High CHR (Not Immune): 1 Macc = .5% hit rate | CHR = 1 Macc (.5% hit rate)

*High Evasion Targets with Low CHR (Not Immune): 1 Macc = .5% hit rate | CHR = .5-1 Macc (.25-.5% hit rate)

Mid Evasion Targets with High CHR: 1 Macc = .5-1% hit rate | CHR = 1 Macc (.5-1% hit rate)

*Mid Evasion Targets with Low CHR: 1 Macc = .5-1% hit rate | CHR = .5-1 Macc (.25-5% or .5-1% hit rate)

Low Evasion Targets with High CHR: Macc = 1% hit rate | CHR = 1 Macc (1% hit rate)

Low Evasion Targets with Low CHR: Macc = 1% hit rate | CHR = .5-1 Macc ( .5-1% hit rate)

The starred values above are important. In cases where the mob CHR stat is low, player CHR might already hit the CHR > targetCHR+10 target. Any additional points of CHR in that situation are only worth half as much as one point of M.Acc/Skill.

The important thing to take away from the above table is that, other things being equal, CHR has a much more variable effect than either Skill/M.Acc. CHR is cheap and easy for bards to obtain because it shows signs of diminishing returns that Skill/MAcc don't.

My general conclusion, and what I do in practice, is to privilege Skill and Macc gear above CHR for debuffs (within reason) ... and Skill above everything for certain buffs (ex. Marches). CHR and M.Acc are important only for debuffing, while Skill levels are important for both buffing and debuffing. Skill's universal application to buffing and debuffing is what makes bard magic skill merits so attractive to mainjob/diehard endgame bards.

The overall usefulness of skill is also why RDM tend to prioritize maxing elemental and enfeebling magic skills. RDM's debuffs (i.e. Slow, Paralyze) fall prey to variable amounts of potency, whereas Bard's Carnage Elegy is a static 50% slow effect when it lands. Stacking stats therefore becomes more important for RDM to stack than BRD.

Practical Gear Examples

To add some more examples to this assessment, using the gear you mentioned:

EX: Marduk body +12CHR vs. Shadow coat +10MACC
Assessment: At best equal to 12Macc, at worst equal to 6Macc vs. a static 10Macc. This isn't a good example. People get marduk body for refresh and fastcast on one piece of gear. Marduk is better than ErrantBody/Osode, but not as consistent a debuffing piece as Shadow/Valk coats.

EX: Oracle feet +5 wind skill vs. Shadow clogs CHR +5, Magic Accuracy +2
Assessment: One is a static +5 Magic Accuracy, the other is at best 7MACC and at worst 4.5Macc. Oracle's feet are really for march tiers.

EX: Jester cape +10CHR vs. astute cape +5 singing skill
Assessment: One is at best 10Macc and at worst 5Macc, the other is a static 5Macc.

EX: Balrahn's Ring +4MACC vs. Veela ring +6CHR vs. Omega Ring +3 Macc +3 CHR vs. Nereid Ring +3 Skill
Assessment: Static +4 Macc vs. +3-6Macc vs. 4.5Macc-6Macc vs 3MAcc. Omega is for debuffs, Nereid Ring is for march tiers. Veela ring is at worst equal to a nereid ring without the benefit to marches, and at best equal to what an omega ring would do in the same best-case scenario.

Why this is Important

Conclusion? MAcc pieces have more stable effects than CHR pieces. Really high-end bard equip is focused on M.Acc and skill.

Food for thought! No one knows how BRD's overall magic accuracy number and RDM's overall magic accuracy number compare. As far as Paralyze goes, for RDM that's based directly on Enfeebling Magic Skill + MND + M.Acc. RDM has an A+ in enfeebling magic, which caps at 276. With full merits, that number rises to a base of 292. A good endgame RDM can hit 320-340+ enfeebling skill.

Bard songs are based on the combination of two skills: Singing skill and Instrument skill. As C-rated skills, they cap at 225. 225+225=450, assuming no merits. You need full singing/wind skill merits and an additional +51 skill from gear to cap marches (barring tiers reachable with relic horn). A maxed march tier bard will hit 533 combined wind/singing skill.

It's entirely possible that Magic Accuracy and Magic Evasion work, conceptually, the same way for BRD and RDM. It's an entirely different thing to try to directly compare numbers and relative effect/efficiency between the two jobs. RDM has to worry about debuff potency (Slow) and BRD doesn't (Carnage Elegy). BRD has to worry about buff potency (March) and RDM doesn't (Haste).

Old Hat Suggestion for Finding Your Ideal Debuff Gear

Suggestion: Put on the gear you think of as your best "debuffing" set. Privilege Macc and Skill over CHR in every slot. See if you get resists in your regular events, and how many.

If you aren't getting resists, keep taking off CHR and putting on MP until you do. That's about what your everyday set should look like.

If you're getting resists with your best debuffing setup, odds are you need to acquire more skill/macc gear. Nereid Rings are cheap. Omega Ring is easily obtainable with Einherjar Ampoules. And Oracle's Pigaches are easy to get if you drop by someone's Dea run -- they're often floored now.


tl;dr version: M.Acc/Skill have more stable effects than CHR for BRD debuffs. Skill has the added benefit of being essential to capping songs and hitting higher march tiers.

For the Blog Folks

I'm terrible, and I should really stop making blog-length discussion posts on the Alla forums. Granted, not a week goes by when I don't see some variant on this same question. I've used Osode/ErrantBody to debuff and now I own a Valkyrie's Coat, so it could be that I'm biased ... BUT ... I love my coat! I love Macc and Skill gear for debuffs and I recommend getting the Hadean Abjuration body if you have the chance.

It took about 11-12 Odins before my group's first H.Body Abj dropped, and it was well worth the wait.

No comments:

Post a Comment