I agree with most of what you're saying with regard to where the benefit from the Surprise Attacks talent comes from, and its relative value with regard to dodged finishers.
But to claim that it's difficult to reach or come close to the dodge cap through gear is misleading. Take 3 items; [Shoulderpads of the Stranger], [Gloves of the Searing Grip] and [Belt of One-Hundred Deaths]. Together they will give 53 Expertise Rating in 2.3, which becomes 13 Expertise. Add in the 10 from the Weapon Expertise tlent and you have 23 Expertise, which gives you a 5.75% lower chance to be dodged, on all attacks. Humans wielding swords or maces get an extra 1.25% (5 Expertise). That's up to 7% dodge negation from 3 items, a talent and a racial.
I know that the dodge cap itself is yet to be determined but it's thought to be roughly 6.5% against a level 73 mob (as stated in the first post of this thread). If that's right then it's certainly possible to achieve for some, and to get very close for others. The items in question are hardly unnatatinable either; the shoulders come from Hydross, the gloves from Al'ar and the belt from Vashj.
You're right. I was speaking from the perspective of a Night Elf without the benefit of an expertise racial, which means that, assuming that the base dodge rate is 6.5% (which seems like a reasonable guess), I would need 64 expertise to cap out, which is indeed challenging to obtain. Particularly since not all the items with Expertise on them are actually good; while the items you suggest do give a fair amount of Expertise, there tends not to be a lot of motivation to wear Gloves of the Searing Grip in practice, as they're inferior to Deathmantle for most purposes and, as such, are unlikely to be worn for very long (in particular, if you've done SSC/TK long enough to have the shoulders, the belt *and* the glove, you're about 95% likely to have picked up a pair of T5 gloves in the meantime).
So yes, as a human, you *can* reach the Expertise cap (although doing so is not necessarily advisable); as other races, it's somewhat harder, and even more questionable in it's merits.
I think that's fundamentally a question of what your guild's policy on epic gems is. My guild generally only gives them out for T6 and best-in-slot gear; as such, it would not be possible to get them for something like Skulker's Greaves. However, if your guild is less stingy with them, it may make sense.
That's kind of what I was asking. You look at several peices of gear withing BT/Hyjal and you can get upgrades from the later, harder bosses. As that is, shouldn't you have the best gems/enchants possible for your current gear to defeat the later bosses? In this case, the guild doesnt need me to perform to my peak ability gear-wise every week, they ahve the bosses on farm. As for their policy on gems, not sure. My question was mainly at how others feel about it. With the 10hit gems (mainly used for 10 spell crit on pallies and 10 resilience lol) I could upgrade my stats with a good 18hit rating taht would last me till t6 legs, probably not coming my way for a while.
Maybe I'll just go ghetto farm gems with some of the guys and ask for 3 lionseye as my share of the gems/money. Inherently, is there anything wrong, does it seem like "a waste of epic gems" if I'll be keeping them for a month or more at the very least?
So yes, as a human, you *can* reach the Expertise cap (although doing so is not necessarily advisable); as other races, it's somewhat harder, and even more questionable in it's merits.
Mmk, speaking of...
What kind/type of testing would we need to figure out a boss's base dodge/parry rates?
If it's possible, we should get some testing going. I'd def take some parses of my combat logs in BT/Hyjal instances if it would help out.
Well, even for a Night Elf, a 5.75% (including the 2.5% from Weapon Expertise) lower chance to be dodged can be achieved, which is great.
With regard to the gear, the Shoulderpads of the Stranger are actually better than the Deathmantle Shoulderpads (at least in patch 2.3, they're extremely close in 2.2) and the 2 pairs of gloves are so very close to each other that you may as well just take whichever you get first, unless replacing the Searing Grip gloves will net you the 4-piece Deathmantle bonus. That would be the only reason to make a point of taking Deathmantle pieces over the non-tier items.
If you're operating at tier 6 level, however, then obviously it would be foolish to not upgrade to the tier 6 equivalents, just for the sake of reaching the Expertise cap.
If Blizzard ever decide to include Rupture in the Deathmantle 2-piece bonus, however, then the Deathmantle pieces will be the clear best-in-slot items at tier 5 level.
--edit--
With regard to testing dodge rates I think the best we can do is something similar to the way the hit caps were tested. Get somebody with all the Expertise gear they can get (preferably including some human and some non-human characters) and see if they can ever record a long test without a dodge showing up (let's say against Venoxis). Slowly take away gear and talent points until you see a dodge. It's not the most accurate test in the world but it at least gives a minimum baseline for base dodge rates.
It's certainly better than simply seeing how many times you were dodged in a particular bossfight, since that could vary for any number of reasons, not the least of which is pure dumb luck.
What kind/type of testing would we need to figure out a boss's base dodge/parry rates?
If it's possible, we should get some testing going. I'd def take some parses of my combat logs in BT/Hyjal instances if it would help out.
Well, getting your Expertise rating into the 20's and fighting 73 bosses would be the way to go to test dodge. Parry will be harder, as it is thought to be much higher and gear can't get you to the 52 Expertise you might need to really narrow it down.
If you believe that 26 Expertise is the number you need (for 6.5% dodge):
Test as follows:
1) Equip gear to 25 Expertise until a boss dodges you, maybe a couple times to be sure. This will prove that the dodge rate is greater than 6.25%.
2) Equip gear to 26 Expertise and get as many reps in vs. 73 bosses showing that you never get a dodge.
3) With enough repetitions, this would prove the dodge rate to be between 6.25% and 6.5%, given that 6.5% makes a lot of theoretical sense, many would probably say its proven. You could follow up with a lot of reps at 25 Expertise trying to show that the observed dodge rate still appears to be about 0.25% at that level.
Now, we don't know that 6.5% is really the number, so depending on your results, you would add more Expertise if you are still experiencing dodges, and reduce your Expertise if you have a lot of reps with no dodges, until you are confident how much Expertise it takes.
And, I am speaking actual Expertise (not Expertise rating).
I will be modifying the first post sometime in the near future to reflect all of the patch 2.3 information. The "Preliminary 2.3 Information" section will be removed. Note, we still don't have any solid information on the proc rate of Executioner, unless I have missed something in the other thread. As such I cannot make an informed recommendation on whether to use it or not. Nearly everything else about the patch has been pretty well uncovered.
If anyone has any information they'd like to add that isn't covered in the "Preliminary 2.3 Information" section yet, please post here.
With regard to testing dodge rates I think the best we can do is something similar to the way the hit caps were tested. Get somebody with all the Expertise gear they can get (preferably including some human and some non-human characters) and see if they can ever record a long test without a dodge showing up (let's say against Venoxis). Slowly take away gear and talent points until you see a dodge. It's not the most accurate test in the world but it at least gives a minimum baseline for base dodge rates.
It's certainly better than simply seeing how many times you were dodged in a particular bossfight, since that could vary for any number of reasons, not the least of which is pure dumb luck.
Right idea, but it's actually much better to work up by increasing your Expertise while still getting dodges. This is because a single dodge proves you do not have enough Expertise (and hence that the dodge cap is higher) but no amount of repetitions without a dodge "proves" that one will never happen.
Right idea, but it's actually much better to work up by increasing your Expertise while still getting dodges. This is because a single dodge proves you do not have enough Expertise (and hence that the dodge cap is higher) but no amount of repetitions without a dodge "proves" that one will never happen.
Definitely true. Can't prove because of a lack of information.
I have the searing grip, gloves, vashj belt (which with human and talents gets me to liek 16.5 expertise) and I'll try with and without the gloves on VR. Its' a steady fight, long in duration usually and can easily be done. I'll check out what I get on there both with and without the gloves (about 1.1 expertise I think).
My hope is that with 6.5% dodge (assumed of course) I'll be able to mitigate a good 4-4.5% of the dodges, which on 700+ swings is quite a bit. The downside, of course, is that to maintain my miss percentage that I just finally got to (289 with 6% passive haste and 371 weapon skill), I'll have to ravage some items and socket with mostly 10hits, meaning lionseye will go up in value (hence I previously asked about socketing my Skulker's Greaves). We'll have to see after tuesday, I guess. :-(
I asked a while ago about the value of armor penetration, and whether someone could point me to the math or testing which verified the results the spreadsheets gave. The answer I got was that it's too complicated to explain the math, and no one replied with any test data.
Well, now I finally got a piece of gear with armor penetration on it (Shady Dealer's Pantaloons) so I did a very small bit of testing on my own. While this testing is far from perfect, it was enough to raise some questions in my mind, so I'd like to ask again if anyone has actually tested armor penetration more extensively.
My test was to equip a worn dagger and hit a level 64 spider (i.e. a convenient mob near shattrath that I could kill with this weapon using only white attacks) with various amounts of attack power (AtP) and armor penetration (ArP). I used a 1-point expose armor to get an additional "armor penetration" data point. Here's the data:
Typically, the number of hits was 30-40, and I threw out crits, so all the hits at a given level of AtP and ArP were N or N+1.
The relationship between AtP and average hit ought to be perfectly linear, and it is (to the extent one can say this with such a small number of data points). ArP shouldn't be linear, but it's close enough over this small range that we can get a rough answer by treating it as linear.
There are a number of ways to look at this data, but the simplest is to ask, approximately how much attack power or armor penetration does it take to hit 1% harder? The answer I get is that starting from 1735 AtP and 0 ArP, it takes 17.44 AtP or 152.8 ArP, so 1 AtP is worth 8.75 ArP. Since the spreadsheets come up with valuations in the 3.5-4 range, this testing result is quite surprising.
Of course, the spreadsheets model a level 73 raid boss with quite a bit more armor, but as I understand it, the value of armor pentration will be *less* against a higher level, higher armored opponent. Unfortunatly, testing in a raid environment with a wide range weapon, and mongoose, unleashed rage, and other effects which change my attack power going off unpredictably is all but impossible.
Other than the fact that I'm hitting a spider instead of a raid boss, am I making any other serious error in this analysis? I ask again, has anyone done any testing to verify the accuracy of our understanding?
Armor has less effect when fighting something higher level than you. What that means in this context is that each point of the spider's armor is less effective than it would be against an even con player. Thus, armor penetration is correspondingly less effective against lower level targets.
Right idea, but it's actually much better to work up by increasing your Expertise while still getting dodges. This is because a single dodge proves you do not have enough Expertise (and hence that the dodge cap is higher) but no amount of repetitions without a dodge "proves" that one will never happen.
Originally Posted by royaljester
Definitely true. Can't prove because of a lack of information.
I have the searing grip, gloves, vashj belt (which with human and talents gets me to liek 16.5 expertise) and I'll try with and without the gloves on VR. Its' a steady fight, long in duration usually and can easily be done. I'll check out what I get on there both with and without the gloves (about 1.1 expertise I think).
You're both right, my mistake
But the gloves give 4 Expertise (18 Rating) which becomes 1% dodge negation. It's still the smallest change you can make though.
Armor has less effect when fighting something higher level than you. What that means in this context is that each point of the spider's armor is less effective than it would be against an even con player. Thus, armor penetration is correspondingly less effective against lower level targets.
I believe this is incorrect. *My* armor has less effect when fighting something higher level, but that's not what I'm looking at. Although it is discussing mobs hitting the player, rather than the reverse, the formulas on wowwiki state clearly that the level of the mob being hit does not matter. Damage reduction for a given amount of armor depends only on the level of the mob (or in the case of my tests, the player) doing the hitting.
If you're right, then the common understanding (which I believe the formulas accurately represent) needs to be revised.
What *does* matter is the amount of armor which the target has. I'm not sure how to determine this accurately, perhaps by fitting a curve obtained with 0 through 5 point expose armor? In any case, as I said before, and assuming my understanding is correct, armor penetration has a bigger effect on targets with lower armor, and so assuming that raid bosses have more armor than the spider, the value of armor penetration on a boss is worse than my test calculated, not better.
EDIT: The curve fit says these spiders have 4985 armor, which seems kind of high, it is actually a bit more than the DPS spreadsheet assumes for bosses (7000 minus sunder and faerie fire = 3828) but the difference in the conclusion when I account for that is quite small. Still looking for the discrepancy between my tests and the spreadsheet.
EDIT2: Well, I think my tests don't mean anything. When I artificially add 165 ArP to the spreadsheet set up with my current gear, the buffed dps goes up 1%. It takes 49 AtP to boost buffed DPS 1%, so the ratio is 3.36:1. If I put in my test gear (worn dagger etc) and look at UNBUFFED dps, it takes 245 ArP and 28 AtP, a ratio of 8.75:1. I'm sure it's just a coincidence that this is exactly what my test showed. On the other hand, if I put in the rogue's dream gear, It takes 151 and 50. So, there's nothing to see here, move along.
Right idea, but it's actually much better to work up by increasing your Expertise while still getting dodges. This is because a single dodge proves you do not have enough Expertise (and hence that the dodge cap is higher) but no amount of repetitions without a dodge "proves" that one will never happen.
While this being true, you can establish a (more or less high) certainty that you will never dodge, if you have a sufficient large amount of data.
Have I missed someone? I keep seeing the Human racial being referred to as adding Expertise. I thought all the racial weapon skills were switched to critical chance.
Have I missed someone? I keep seeing the Human racial being referred to as adding Expertise. I thought all the racial weapon skills were switched to critical chance.
The change was reverted for some strange reason. (Human favoritism!)
Fans glory to the Gladiators,
Gods glory to the Heroes.
Your example Combat Mutilate build skips Imp KS and QR. Imp KS is certainly of debatable utility -- it's very nice where it can be used, but most significant fights it can't be -- but getting that deep into assassination without QR seems like a strange choice. The extra what, 10 DPS from Vile Poisons? has got to be worth less than surviving pounding -> melee orb, root in green goo with shock debuff, etc. Non-DPS-related intangibles are pretty much the only thing Combat Mutilate has going for it atm, skipping them seems a bit silly.
There's also the possibility the energy return on dodge actually makes QR higher DPS than Vile Poisons. As far as I know, nobody's done any math on it :P
Your example Combat Mutilate build skips Imp KS and QR. Imp KS is certainly of debatable utility -- it's very nice where it can be used, but most significant fights it can't be -- but getting that deep into assassination without QR seems like a strange choice. The extra what, 10 DPS from Vile Poisons? has got to be worth less than surviving pounding -> melee orb, root in green goo with shock debuff, etc. Non-DPS-related intangibles are pretty much the only thing Combat Mutilate has going for it atm, skipping them seems a bit silly.
There's also the possibility the energy return on dodge actually makes QR higher DPS than Vile Poisons. As far as I know, nobody's done any math on it :P
I had constructed the template from scratch with my roughly zero knowledge of Mutilate because I had forgotten that Oscarvil already provided a Mutilate template on the first page of this thread. I went back and substituted his build in.
I am wondering if you can put any EP value to the T4 2pc set bonus. Several of the new badge rewards are better than T4 pieces, and so are Arena S3 head/chest, and I can replace them if I want. However, you said "keep 2pc T4 until 4pc T5". The only T5 piece I am likely to get in a long time would be the T5 shoulders. Should I keep two T4 regardless? I gain around 27-37 EP for each T4 I can replace with Badge/S3 items. My spec is combat dagger if that makes any diff. Will it be worth losing 2pc T4?
Have I missed someone? I keep seeing the Human racial being referred to as adding Expertise. I thought all the racial weapon skills were switched to critical chance.
They dropped the +1% crit and decided to give 5 expertise.
I'm personally excited, because with that, talents and [Belt of One-Hundred Deaths] I'll have over 16 expertise or ~ 4% -dodge. I can't wait to see on fights like winterchill, akama and others how much extra dps will be possible once I get my hit rating closer to the new cap, which tbh, isn't really that hard once you factor in melee dranaei buff, hit-rating food and just stacking pure hit, rather than ap/agi/crit or whatever else kids these days think is best. lol
11/09/07, 2:59 PM
Similar Threads
