Blue Mage: Best thread ever

[Nightfyre]

Max increase of 18 is consistent with multipliers of damage ((skill*0.11)*2, 3.0 multiplier, 3.15 pDIF, 2*3*3=18). This suggests there is a skill tier at 337.

However, your numbers brought something to my attention. Using 336 skill:

D=36*2+3=75

fSTR capped at 22

WSC=156*.5*.85=66

ratio capped at 2, pDIF MAX = 3.15 (as a crit)

(75+22+66)*3.0*3.15=1540

When you achieved 1549. This suggests I'm missing 1 base damage somewhere in my calculations. What did I miss?

EDIT: Same deal for your 337 skill set, my number is 9 short of your max. It's consistent, so it's an error on my end somewhere. As for your small sample size, I'm thinking you rolled near enough to max pDIF on your high in both sample sets to achieve the highest possible value after truncation, so looks good! Thanks for gathering the data.

[Sho]

Max increase of 18 is consistent with multipliers of damage ((skill*0.11)*2, 3.0 multiplier, 3.15 pDIF, 2*3*3=18). This suggests there is a skill tier at 337.

However, your numbers brought something to my attention. Using 336 skill:

D=36*2+3=75

fSTR capped at 22

WSC=156*.5*.85=66

ratio capped at 2, pDIF MAX = 3.15 (as a crit)

(75+22+66)*3.0*3.15=1540

When you achieved 1549. This suggests I'm missing 1 base damage somewhere in my calculations. What did I miss?

I'm not so sure myself how that came about, checked back and it wasn't a mistake. Same conclusion for everything else, however. You didn't miss anything.

Edit: No prob! I'll compile a 50 Rabbit set after I obtain Scarf and Gloves.

[Nightfyre]

Hmm. I'll try to replicate the damage difference. If I get similar results I'll play around with some base damage values, see if I can figure anything out.

[Rena]

Playing around with ranged spells, fSTR2 seems to cap at 46 now, rather than 44, so it's looking likely fSTR might cap at 23 as well. Might have missed other possibilities, and more testing is needed (working on it, slowly), but for now, it would fit.

At capped fSTR2 (107/106 STR vs. tiny mandragora, did the same damage with feather storm at 100 STR), feather storm (107 STR, 77 AGI) was doing 615 damage, queasyshroom was doing 472 (106 STR, 66 INT), and pinecone bomb (106 STR, 77 AGI) was doing 952. Unless I missed something, working backwards through this gives a fSTR2 of 46 for all of them.

Edit: vertical cleave testing at uncapped fSTR so far agrees with the above theory. It's hard to disprove, seeing as it would require hitting the highest ~9 points of damage, which might just not happen due to bad luck, but for now I'm pretty confident saying it's just a fSTR cap increase. As for why that happens, I have a guess or two, going to test it as well as I can.

Edit #2: getting 44 fSTR from WHM79/BLU39 (585 damage at 105 STR, 67 AGI from feather storm), so it looks like level 80 = +1 to fSTR cap. Need more testing at lower levels to determine pattern, but an fSTR upgrade every 5 levels seems a likely possibility. Would love to do the testing myself, but unsure how much of it I can do myself given current resources (jobs, gear, sync targets).

Edit #3: I'm not 100% certain on the result of the following synced testing due to SE messing around with level correction, and lower level ones may not be completely accurate still, but I don't think it should differ enough to matter; synced to 63, assuming level correction is still 0.89 (has a decently large room for error, could be as high as 0.93 and as low as 0.87), I'm getting an fSTR2 of 38, which would be consistent with fSTR cap increasing by 1 every 5 levels. That's about as much testing as I can do at present I'm afraid, if anybody else wants to do some more that'd be great, but I'm pretty confident in this as is.

[Nightfyre]

fSTR might cap at 23 as well... fSTR upgrade every 5 levels seems a likely possibility.

This is what I suspected was occurring. If it actually occurs at 80 I won't be able to verify sadly (still 78 ; ;), but I'll knock out a few VCleaves on some mandys in the next hour or so to contribute some data towards whether it occurs at or before 80 if we're right.

EDIT: Wrapping up a 100 spell sample set, my highest was within 3 points of a 22 fSTR cap estimate. Evidence of a sort, but no proof. I was unfortunately unable to produce a minimum value that would disprove 23 fSTR at 78 either, as I only bottomed out at 1223 damage with 150-151 base damage, or a pDIF of 2.7177~ or 2.6998.

I probably won't pursue this line of inquiry (level 78) further; between hopefully hitting 80 soon and the minimal impact of the exact level at which the change occurs I'm ok with what data I've generated. Establishing that the cap has most likely changed on the other hand is of value since it's information of some use now and in the future, especially given Rena's evidence that it may continue to grow.

[Yugl]

I'm LV 80 BLU. What do you want?

[Rena]

Nothing to be tested at 80; what I personally think is needed is testing at 59/60/64/65/69/70/74 to see if it follows the pattern I proposed. We knew from testing by Releena a few years ago that fSTR2 cap at 75 was 44, and fSTR 22. Sho's testing showed fSTR likely capped at 23 at 80; further testing by me showed fSTR2 capped at 44 at 79, and 46 at 80 (and some limited vertical cleave testing at uncapped fSTR at 80 did nothing to disprove this by providing no damage higher than the expected range. Limited as I said, but it helps rule out any other possible unknown variables that could have caused the increase to Sho's numbers.), as well as 38 at 63.

The values I got fit a floor(level/5)+7 formula for fSTR (multiplied by 2 for fSTR2), which seems fairly simple and, in my opinion, likely. Testing at 59/60/64/65/69/70/74 would help since these are the expected levels where any change would be noticed. (Lower levels would ideally be tested too, but because of difficulties in reaching the cap due to limited STR options, I'm not sure how feasible that would be.) If anybody wants to test at those levels, that'd be great, but to be honest it's not very useful or practical information. We only really need to know the caps at current level caps, knowing the formula is more a curiosity than anything, as fSTR2 cap is very easy to find, and fSTR has so far always been half of that. I do intend to test it as levels increase past 80 though, which should be enough to prove this pattern if it is that simple.

[Marcicus]

Any tests on Cimicine Discharge? Feels like Fliamented Hold is more potent.

[Nightfyre]

I believe testing showed it to be 15%, which would indeed make Filamented Hold more potent.

[Draylo]

It isn't bad if you are doing crowd control on a large group of mobs for certain things. I had to tank a large group of mobs for certain wotg missions and its nice to slow all of them at once.

[midguardian]

But Filamented Hold is AOE, so why not use that?

[Darkmagi]

Conal is not AOE. Cimicine Discharge also has a really huge radius so if you're aiming for crowd control it's superior.

[Kegsay]

Apparently Symbios Gloves (Blue Magic Skill ones) is from Irrlicht in Abyssea-La Theine. It is a corpselight, so I would be curious to see if a BLU could solo it. Appears to not be pop-spawned.

[huckster]

Hey guys - now with higher level cap, new spells and items - it's time to revise the cannonball build.

I want to update my cannonball chart and I need your help! Please post your thoughts on how to epeen our cannonball shots - I know there's a lot to update (eg mirage items, enkidu gear etc)

Here's the 'old' version of the chart:
http://hitczykit.com/cannonball.jpg.

Post your changes - and I'll update it tonight.

Thanks!

[Draylo]

Yeah, not looking forward to grinding that out. There's a third option though...

Khanda: D:50 "Double Attack+7" delay 252

DAKhanda/Shamshir is about 2.5% faster than Shamshir/OATKhanda. Assuming it procs on both hands:

2.44 hits/round vs 2.64, 8.2% loss of WS frequency. However, for non-Sanguine Blade WS you're doing more damage via DA and 2 higher base damage. Worst-case scenario here (Vorpal) still yields a ~5% increase in per-WS damage for a total loss of 3.0-3.1% in WS phase. For Expiacion/Savage Blade-heavy playstyles the value loss could go a bit lower. Best case (Savage Blade) gets me a ~2.4% net loss. So we range from a 5.7% loss to dead even depending on WS of choice after factoring in lower delay.

TP phase difference in a scenario favoring OAT (positive fSTR) is (55*1.22+53*1.22)/(53*1.15+47*1.49)=0.6%, increasing to 1.5% at neutral fSTR. Adding in delay advantage gives 3.1-4%.

So DA Khanda is anywhere from ~0.5% behind to up to 2.4% ahead depending on your WS's of choice and fSTR by my napkin math. lolxhit implications due to delay change favor OAT ever so slightly but still, I'm liking this option if the DA affects both weapons...

I'll try and do a better workup over the weekend if nobody beats me to it but I'll be out of town so it won't get posted until Monday at the earliest.

Interested in your findings, I have been looking to do some magian weapons as much as I dread it... Was thinking of shamshir/OAT so far, but curious on the other choices (OA3 and DA)

[Sho]

Hey guys - now with higher level cap, new spells and items - it's time to revise the cannonball build.

I want to update my cannonball chart and I need your help! Please post your thoughts on how to epeen our cannonball shots - I know there's a lot to update (eg mirage items, enkidu gear etc)

Here's the 'old' version of the chart:
http://hitczykit.com/cannonball.jpg.

Post your changes - and I'll update it tonight.

Thanks!

For sword, Firmament +1 completely trumps Wing Sword (+1) and is now dual wield-able as you know. Neck now has Ire Torque (+1) which is +6(7) STR. Then there's Magus Jubbah+1 which many BLU should have by now... Vulcan's Ring (+6 STR) is not common, but will soon be a little bit commonplace to obtain. Legs maintain as Enkidu Subligar, and the best feet should be Rutters.

[Ddboner]

Smilodon Mantle+1?

[Aben]

1) Heafoc Mitts (+13 str) for hands
2) Teutates Subligar (+8 str) for legs,
3) Gnad helm was never added for head I don't think (nor a Moogle str +4, crit +2% or vit +4, but thats situational).
4) +3 str earrings

The first two assume you are using SA of course

As for duel wielding firmament swords, keep in mind Acheron shield adds a good amount to the pdif part of the cannonball equation, and may still be effective on high high def mobs. On the flip side, high def mobs may become less common soon enough.

Soulsaber +6 str/soulsaber +6 vit = NQ firmamentx2

[Clash Perez]

For sword, Firmament +1 completely trumps Wing Sword (+1) and is now dual wield-able as you know. Neck now has Ire Torque (+1) which is +6(7) STR. Then there's Magus Jubbah+1 which many BLU should have by now... Vulcan's Ring (+6 STR) is not common, but will soon be a little bit commonplace to obtain. Legs maintain as Enkidu Subligar, and the best feet should be Rutters.

Would dual-wielding Firmament +1's be worthwhile, considering the need to sacrifice 7 set points? I'm inclined to stick with the Vulcan/Brave combo, pending some mathmagicianing.

Regarding the chart, it's missing a number of things. Strigoi Ring is another option, as well as Vulcan's Staff and Brave Grip, Dusk Trousers, Anwig Salade if one so desires. Others?

Spell list is out-dated too, IMO; here is what I've been using recently:

Spoiler: show

Option A: Defense Bonus, STR+24, VIT+9
Bludgeon
Pinecone Bomb
Uppercut
Cannonball
Sub-zero Smash
Vertical Cleave
Blastbomb
Bomb Toss
Self-Destruct
Firespit
Heat Breath
Triumphant Roar
Cocoon
Plasma Charge
Saline Coat
Plenilune Embrace

If you're not getting Refresh and/or don't have a chance to rest for MP, or if you just can't be arsed to bother:

Option 2: Defense Bonus, Batt. Charge, Auto Refresh, STR+21, VIT+9
Pinecone Bomb
Uppercut
Cannonball
Sub-zero Smash
Vertical Cleave
Blastbomb
Bomb Toss
Self-Destruct
Firespit
Heat Breath
Triumphant Roar
Frightful Roar
Cocoon
Plasma Charge
Saline Coat
Battery Charge

These do assume full Assimilation merits, which may not always be the case.

[Sho]

Was looking at using then for lower end mobs. Of course one would still maintain sword/shield for higher def mobs. At least, those with Assimilation 4-5/5 have that option without much trouble.

Also happy to see that many BLU don't bother with Auto Refresh during Cannonball/VC these days.