February 2017 Version Update

Wait huh. They adjusted evasion down again?

combat skill nerfs

Ahh they nerfed the evasion skill, which is probably how they nerfed the magic evasion since that's just a hidden C rated skill.

It's been doing that randomly for a while now.

Naw it's how they nerfed macc of ws defensive ones just got caught in the crossfire. The update notes says combat skills, magic skills, magic evasion.

No I mean mechanically, that's how they nerfed magic evasion on CL content, would also explain why HTBC's were left out (their just level 99 monsters with hyper inflated stats).

Magic Evasion is a "C" rank skill everything possess's. A Level 99 player has 373 magic evasion, a level 99 monster also has 373 magic evasion. Then we add on all the other stuff like elemental magic resists and dSTATS to get the final magic evasion value. SE recently lowered CL monsters evasion by nerfing their gains through AGI but left their obscene skill levels in tact. This time around they nerfed their skill level growth directly, which reduces practically everything they have. It's a lot more then just magic accuracy / evasion.

Magic Accuracy on TP moves and additional effects (determined by combat skill of main weapon)

Proc Rate on additional effects for TP moves and regular attacks (determined by combat skill of main weapon unless its 100%)

Magic Accuracy on Spells (determined by magic skill of spell)

Magic Evasion (determined by magic evasion skill)

Physical Attack (determined by combat skill of main weapon)

Physical Accuracy (determined by combat skill of main weapon)

Parrying Rate (determined by combat skill of main weapon)

Shield Block Rate (determined by combat skill of main weapon)

Guarding Rate (determined by combat skill of main weapon)

So pretty much an across the board reduction. Of course some of there won't matter much, while others the reduction isn't sufficient to counter the *fixed* glitch in Geomancy.

I know what you meant I'm saying they specifically mentioned mechanically how they nerfed macc and then specifically said meva. Could be lolSEtranslations or poor wording true but it's also likely they chose that wording on purpose and defensive combat skills were just one of many things caught in the crossfire as you yourself note. As far as HTBF and UNM not sure how that reasoning works... just because they are 99 mobs doesn't mean you can't adjust them the same. They were probably just doing it by zone. That seems to be a thing for them. Note the reason those mobs are 99 is because those zones are still level corrected and so they went with higher stats to mimic higher lvls to sort of make it so there was no lvl correct. Also note the listed legion zone which does contain old content stuff... also note they initially did the AoE adjustment there and the getting rid of level correction there before ambuscade existed. It's probably why ambuscade is in legion.

And the whole hidden always capped meva skill thing wasn't really proven or even really tested (partly because it's completely untestable since that idea is that it's something invisible and unchangeable for players) for but more of a way of giving a simplified way of thinking about how meva rose with lvl. But given it's assumed to be the same for every job and auto capping it sounds less like a skill and more like a general lvl factor though again can't really prove. I'm quite aware for how magic hit rate equations work though could add in resist traits and sdts and then that weird soft immunity they added with the immunobreak update. And also quite aware there was a lot of collateral dmg in this nerf which is along with their worded is a decent part of why it seems like this was just another consequence they didn't foresee with their quick and easy hot fix. Was one of the very curious things noticed as soon as the update notes went up. Though given what they had said I'd figured they'd just meant offensive but that has clearly been proven otherwise

Ok let me explain how they do monsters in games like this.

It would be a colossal waste of time for them to individually edit every single monster and notorious monster released after SOA. So the idea that they went through every single one of them and said "hmm I guess we should take away X from this guy" is dumb. Instead all they do is edit the template of that monster type and viola the change happens across the board to all of them instantly. There is no actual rank A/B/C/D/E/F in this game, it was merely something the JP's came up with and we just kinda adopted it. So when I say "Rank C" I mean the formula used to dynamically determine a monsters resistance is the same formula as used to determine a rank C skill for a player. We figured out base magic evasion a long *** time ago at level 75 against level 75 monsters, it's the same way that dINT's bonus to magic acc was determined. It's not impossible, it just's hella time consuming and requires a ton of data points and controlled circumstances.

So in this game every monster has a template that says "As monster gains level X also gain amount Y" of a skill. We differentiate it because "stat" usually refers to STR/INT/MND and those have special scaling based more heavily upon job while "skill" is based more heavily on level.

So all SE did was adjust the formula that determines monsters magic evasion skill so that it didn't rise as sharply past 101, which is why the old vs new gap would grow as the monster's level went up. This wouldn't effect HTBC's because those aren't level 100+ monsters and thus their scaling is at level 99 but with gigantic stat bonus's given to them instead. So a VD Leviathan isn't a level 135 NM but a level 99 NM with a huge bonus applied to it's Attack / Defense / Magic Evasion / Evasion / ect. SE's stat growth nerf would of never applied to those situations and SE never bothered to even check. It will likely be updated in the near future as people remind SE that their quick fix missed stuff.

In other words, SE doesn't manually set any stats in this game, instead they leave it to formula's that automatically generate those stats. Which is why they completely missed a huge *** glaring bug in geomancy for two years.

Didn't see anyone mention this about the new crafting shield quest. If the craft you are attempting has a subcraft it's not counting.

Wrong, I see that if you NQ not meeting condition its 0 points.

Meh was gunna huge wall of text but probably better to keep it comparatively brief.

First off stop explaining the most basic of stuff like I'm an idiot then adding the inferences onto it. I've read thru the entirety of Robonosto's blog of translations from Lodeguy's testing multiple times. Even tried looking thru Lodeguy's blog but google translated entire testings is pretty not fun.

Second and sort of other points as well. Never said they set each one manually anywhere. There is a world of space between this one adjusting 1 growth formula that filters down to most everything and adjusting each individual stat. Obviously everything is based on formulas that we colloquially refer to as Ranks. SE has even confirmed this in the past with some of their other adjustments notably the hp/mp racial adjustment. They've also noted that they can and do add to said base stats. But we have no real way of knowing how many of these are linked. Could be 1 like you suggest. It could be like a few like maybe how it was worded they have one for combat skills, magic skills and magic evasion (and of course others not adjusted like hp base stats etc). We'd have no real way of knowing the difference and neither is the great burden that manually editing all are.

3rd there was a lot of macc tests with various levels mob types and spells and gearing/stats done. Reading back over Robonosto's notes I can't find anything about where that C rank thing came from as it wasn't in the blogs at all. Closest I can figure it was some of the element neutral tests against a 75 steelshell that showed hit rates that look like they could correspond to something around their though hard to say for sure since they don't list mob int and newer dint testing suggests they might not have had dint completely right. However a single level doesn't show much in that regard and while there was testing on the next lvl up keeping everything else the same including same type of mob with the same int the conclusion seemed to be that there might have been a lvl pentalty/bonus not necessarily this C rank thing. That said from a player perspective the 2 would be indistinguishable depending on how it was implemented but hard to tell from the testing that was done

4th they've shown a propensity and ability to make changes by zone generally mechanically. Namely they'd gotten rid of the pdif related lvl correction penalty in conveniently all the zones this update applied too. Which does include old non ilvl legion but does not include the 100+ mobs they added in other zones. Given we are in agreement that it's highly unlikely they adjusted all these mobs individually it seems likely they can do some things zone wide and in fact that was why they said HTFB mobs were just going to be inflated stat mobs since actually raising their lvls while keeping lvl correction would make them insane. Now say instead of skill it was just some kind of level correction factor like was guessed in the testing they could have adjusted for those zones. Again not saying this is what happened but it could fit as well from observation. It would also explain why 100+ mobs in other zones weren't on the list as well. That said this is SE and I fully acknowledge those mobs could be effected and they didn't care to mention/remember them

And lastly I keep finding it highly amusing the blog in questions title starts as unbearable triteness lol

Welcome to discussion with Saevel.

- Yesterday our OMEN group literally raped KEI like we never did before as SCH SCH BLM BRD GEO PLD.
- Another group of well geared playes using BLM BLM SCH SCH PLD GEO had trouble landing nukes at all.



Differences:

- INT songs
- Threnody
- Our GEO using no gearswap while the other did. Some .luas have been known to somehow f*ck up Bubble potency.


If anyone is testing high level NM again I propose to try unloading gearswap and using bubble set manually and see if there is any difference.

People typically have trouble with GEO gearswaps because they use ammos and ranged and don't set the opposite slot to be empty. So instead of trying to equip Dunna, it tries to equip both Dunna and Impatiens, which results in you having no instrument and getting a half potency geomancy.

Either way, good to know that it's still possible to mage-kill Kei. I want that Urushiiii!

not just this but the fact they cast geomancy and indicolor with a fast cast set in precast and essentially casting with zero potency since neither of these types of spells have a midcast. They are considered like a long casting JA, precast is where potency is needed to be equipped. A lot of people dont know this.

That's not true, but if it works for you then keep doing it I guess.

Also, my characters have been DCing at the end of their logout countdown since the update. It then boots me back to the lobby, where it lets me log back in.

Please elaborate your theory a little more.

Yup I was in the other group Tyche* is talking about (i think). It was sad really. The 2 BLMS got Outparsed by the Tupsi AG SCH hard, and even said Tupsi SCH ended up getting resisted a ton even with Am3 up. This was with Idris focus/malaise, entrust Lang. My BLM isn't near the top of the Ls nukers, but was more than good enough for Helms, and was overkill for Kei pre-update.

When u land 15 out of countless nukes for full dmg....and 12 of them were either under ES/Manafont/or Bolster/TR.....

Literally I had to take off all my MBD Gear and nuke mostly in a piecemeal Magic ACC set i made on the fly during the fight, and was still geting 1/2 resists. SCH did 45% of the DMG, when usually does about 20% on our Kei runs. Oh the NON Tupsi SCH who is very well geared? 2%. Yeah we won, but we got to Kei with 38mins on the timer, and exited with about 4mins left.

Basically...bring something else other than one GEO for support.

i tested this on lvl 1 mobs with geo-poison, i casted geo poison in full FC gear then full potency gear (master 900 skill) both with dunna. clearly the FC set caused way less dmg. then turned on gearswap and recasted, FC in precast then potency in mid. and the poison dmg was the same as full FC set all the way through. Im not the only one that doesnt use FC for geo and indi. And if i recall correctly im prety sure it was in the release notes for geo when geo first made its appearance.

like byrth said, it works for me. Maybe i did my test wrong and someone might want to repeat. but there was a distinct potency difference to me.

Interesting. Investigation to be done.

Casted indi-barrier on my 900 skill mule, it uses fastcast precast.

Before:948
After:1547

1547/948 = 63.1%

bgwiki lists potency as 62.8%

So I'm actually a bit stronger, likely due to rounding differences.

The only reason using fastcast would cause your bubble to be wrong is if you had a lag spike. Which isn't unique to GS, I've had sch macros using equipsets only misfire, and had to change my wait 1 to wait 2 just to be sure that wouldn't happen anymore. Lag spikes are rare, but sometimes it will rarely cause gear (talking like .001% kind of thing here) on at the wrong time no matter which method you use outside maybe doing everything manually but I could see even that screwing up if you weren't reacting to the lag and just going from muscle memory.

Anyway, I'd take byrth's word on it, he wrote GS and has some of the best understanding of how server-client communications work in the entire community.

I have 90 accuracy with no gear on at all. If I equip my geomancy midcast set and unload gearswap, then cast geo-precision and get naked, I have 190 accuracy (as it should be).

If I load gearswap and keep using:
geo-precision
gs equip naked
checkparam <me>
full circle

over and over, it tells me I have 190 accuracy every time.

guess i was wrong, i did do the test about a year ago, i probably had a mistake in gear from pre to mid back then.

I wish this camp wasn't such a *** to get to.

Am I the only one who won't take the "they don't have time for that" excuse? Isn't this a subscription based MMO?
This geo nerf broke a good portion of the endgame stuff, and when a BiS BLM complains about not being able to land nukes, even with the right support, it's pretty clear this was not intended. If they had taken the time to test the changes, they would have known.

I'm not saying they did manually adjust each mob, but given it's probably just 1-2 database values/formulae to change, it's entirely possible they did change each individually. Small NMs would only take 10-15 seconds each(copy/paste across the whole tier.. slightly different if a nm was meant to be particularly evasive or not). It's like 2 hours of work for 1 dev.

Saevel is probably right, but he loves to throw his assumptions around as fact without any evidence.

I'm very very skeptic on this, unless it's a windower/gearswap/packets issue from the last patch.

The only issues I've seen in precast for Geomancy is when people swap Ammo (if they're Dunna GEOs) or Weapon (if they're Idris GEO)
When you do that, if quickmagic procs and/or if a packet is lost, there's a chance you will end up midcasting with weapon/ammo you had in precasting.
Which is why I precast with Idris and Dunna.

Other than that I'm very skeptic with what you're saying. Should be easy to test with stuff you can check with /checkparam.


Use Checkparam
Use Eva bubble
wait 5 seconds
Use checkparam and check value
Use acc bubble
wait 5 seconds
use checkparam and check value


Rinse and repeat like 50 times, could automize that and use logger to see the values. With a large enough sample any possible anomaly would stand out.
If the values are always fitting what they should be, then the "no precast for geomancy" point becomes moot.
I can't rule it out but I'm very skeptic about it.