Remember, that the weaponskill on the Belt would free up at least 1 point in WEX. I wouldnt mind 1 or 2 more Talent-Points in the Assassination-Tree if i can retain the +5 Weapon Skill with the Vashj Belt.
The Belt of One-Hundred Deaths seems to be better than the Belt of Deep Shadow, even if you have 2/2 WeX. At least according to the rogue DPS spreadsheet, which meanwhile has adopted the weapon skill formula (someway).
Correct me if I'm wrong, but the whole content of the thread is only focused on weapon skill impact on hit, that is supposed to be just a (minor? ) component of the whole. I've not been able to find a post where it's pointed out which impact it has on crit and dodge.
It's indeed outstanding the testing that lead to discover the gap at 354-355, and the whole analysis but to my eyes looks like we are still far from fully extimate the value of +weapon skill items (i see people giving conclusions over belt of one hundred deaths just based on hit part)
Long sessions with Venoxis running a combat logger.
1) 0/2 WEx, No Belt Of One Hundred Deaths (Baseline/ +0)
2) 1/2 WEx, No Belt Of One Hundred Deaths (+5)
3) 0/2 WEx, Belt Of One Hundred Deaths (+6)
4) 2/2 WEx, No Belt Of One Hundred Deaths (+10)
5) 1/2 WEx, Belt Of One Hundred Deaths (+11)
6) 2/2 WEx, Belt Of One Hundred Deaths (+16)
Fun one I couldn't do:
2/2 WEx, No Belt Of One Hundred Deaths, Human (+15)
Those all seem like enough data points to start getting some trends working. It tests all the edge cases and the one-afters, but nothing sub-5.
Tests will be performed by DPS'ing from the front of the mob and attacking nothing but Venoxis. The goal is to test the effects of weapon skill on parry and dodge.
I'm not sure I have a great way to get a sub-5 weapon skill boost to test for non-linearity. Are any of these tests superfluous? Is the boss incorrect? Any tips? I want to finally offer something to get the discussion flowing.
Last edited by Takkara : 10/03/07 at 5:12 PM.
Reason: Aldriana's recommendation added
I was wondering the same thing with the holy fire and the heal, as it's been a while since I ran ZG. However, I can't think of many other high level options that are easier to get data from with a small group of people. If you can give me one, I'll try to get there.
Venoxis is a bad mob to test. As discussed previously, dodge and parry numbers have been tending low vs. casting mobs. Seems, the act of casting prevents a mob from dodging or parrying making the dodge and parry numbers seem low. Second, Venoxis doesn't have that much life and goes too quickly into Phase 2. I remember with the Paladin testing, they were surprised how quickly she dropped into Phase 2. Renew didn't help her very much even with the starter weapons they were using.
Maybe Onyxia would work better? Has a fair amount of life, phase 1 consists of pretty basic abilities. Anyone happen to know if any of those abilities have a casting bar? One tank doing all the damage with a starter weapon could probably stand there quite awhile. Given the 1000's of swings we are going to need to figure out anything conclusively, Onyxia would probably be a better candidate.
Long sessions with Venoxis running a combat logger.
Two issues here:
- As already mentioned, Venoxis is a bad choice, since it seems that spell casting is preventing dodge and parries. I can't say for sure for this particular mob, but at least on Drek'Thar (channeled whirlwind) and Morogrim (Tidal Wave) I didn't encounter a single dodge/parry while they were casting, so I assume this is the case for most if not all channeling mobs.
- You need a *huge* sample size to be somewhat certain on the effects if you can't rule out an event and are only trying to see the difference between percentages.
As I posted earlier in this thread, I was beating up level 70 mobs with a fixed crit rate, and on 4000+ attacks my crit rate didn't match the tooltip by quite a large margin.
The best option would be a mob that doesn't hurt you, and has a quadzillion HP, so that you could run the test over night (a level 73 servant in Blaste Lands would be nice). But unfortunately that ain't possible.
Therefore we'd need several players performing the same test, preferrably with the same stats and same equipment (crit rating) and then merge the various results.
Garr might be a good choice, as they aren't so many mobs without (channeled) spellcasting out there.
Onyxia's fear certainly has a cast time... possibly her fire-breathing thing as well, though I'm not sure about that.
Realistically, the best options I can think of are some of the MC bosses - Garr, maybe, if you can find a way to deal with his adds.
But I don't remember Onyxia fearing at all in Phase 1, only Phase 3. The fire-breathing or knockback are the two I'm most worried about having a cast bar. The good thing about Onyxia is she is rather easy to get to. The only problem being the long quest chain to get access to her (and anyone who was raiding pre-expansion should already be keyed). I can't think of any easier to get to tank and spank fight than Phase 1.
Garr might be good if his debuff doesn't have a cast bar, have to be careful about positioning since I think he turns when he casts it like Golemagg.
Knockback most certainly doesn't have a cast time, but the fire breath might. It's been awhile :/.
Garr is pretty simple to deal with, a good protection warrior can control all of the adds and not worry about dying. I'd imagine any protection paladin that has a pulse can do the same. The adds are all level 60 and really don't hurt much at all, though you might need a couple shields if you're going to be there awhile.
Damn, I was wrong about that in a hard way, it looks like. I'm not sure about the name of Onyxia's Fire Breath spell. As for Garr's debuff, I only recall two AE abilities he did which did not target anyone in particular. I certainly can't remember him ever turning away, but I guess I could be wrong.
So I wonder which Wing Buffet Onyxia uses, since the last 2 on the list are instant casts. On wowwiki, they list a Knock Away and a Wing Buffet. Makes me wonder if Wing Buffet is only used if multiple targets are in range, but if Flame Breath has a casting time, that hurts a bit.
At the same time, it probably wouldn't be impossible to edit the combat logs (in Notepad or the like) and cut out attacks so many seconds before an ability. Which doesn't rule out using data from some bosses with a little manipulation. Now, if a boss is using such abilities right and left, it's probably not worth the edit time.
For example, it wouldn't be that hard to use "Find" and look for casts Flame Breath or whatever, then cut out attacks starting 5 seconds before. Of course if you are seeing dodges preceding a spell, you can also rule it out as a negative factor.
Garr certainly is sounding good, the only problem maybe being the number of people you need to come along. Onyxia might even be testable with just a tank and a healer with a tank and 2 healers being easier with easier trash. The main problem with MC will be the fire elemental spawns; have to have enough people to kill the adds.
Remember, that the weaponskill on the Belt would free up at least 1 point in WEX. I wouldnt mind 1 or 2 more Talent-Points in the Assassination-Tree if i can retain the +5 Weapon Skill with the Vashj Belt.
Since I'm human, is WEX a waste of talent points? It seems in general that the first 5 points of weapon skill are hugely valuable, and more than that is nice but not critical.
Using the calculator that was posted, it looks like +10 WEX from talents saves me about 16 hit rating. Seems those 2 talent points are better spent elsewhere. For reference, I don't PvP, and as far as raiding we're only as far as attempting Curator.
P.S. I know my talent build is sub-optimal, I plan to respec, just haven't decided exactly what to do yet.
WEx, even as a human, is point-for-point about as good as precision. On the whole, I consider it to be worth taking unless you urgently need the talent points for something else.
If the hit chance just depends on the difference in your weapon skill and the mob you're attacking's defense skill surely you _could_ test on a blasted lands mob if you were level 50-54 and attacking the appropriate level mob.
Obviously you'd need a pocket healer to keep you up but is there a reason why this wouldn't give good data? It would certainly save a trip to some lightforsaken dungeon somewhere to test on a boss mob.
You don't need any +skill gear to test it. As this could be a character genuinely leveling up weapon skill you could take data for each skill point increment. On any single character you can take tests for all the different weapons they can equip to maximize data per character. Also you could get a level 70 to tank the mob to save the lower level from damage.
When your weapon skill is not maxed out there is no real benefit in gathering these data sets, as you would have way too less attacks per point of weapon skill to be sure of anything. Even if you have 5 characters to do your tests with.
I leveled bow from 299 to ~345 with around 600-800 arrows on Keal'thas.
The problem with that approach is, you cant gather large samples, as you will gain skill ups during gathering data and the skillups not being tracked in the /combatlog. So it boils down to testing in steps of 5.
I might have misunderstood you though. I happen to have a druid around lvl 45 at the moment so if anyone can come up with a worthwhile test give me a shout.
You could get data with rogues getting 0, 1 or 2 Weapon Expertise or 1-5 in Mace Spec similar to how you got a lot of data for the original tests. Human rogues could get the 5-10 points in 1 skill increments, other rogues could get 0-5 in 1 skill increments.
[Edgemaster's Handguards] are level 44 also, so a Warrior could grab them and use points in Weapon Mastery to get increased skill levels as well.
10/03/07, 5:18 AM
Similar Threads
