Elitist Jerks
Register
Blogs
Forums


Go Back   Elitist Jerks » Class Mechanics » Hunters

Reply
 
LinkBack Thread Tools
Old 08/08/09, 8:06 PM   #331
jmichaelp
Glass Joe
 
Night Elf Hunter
 
The Forgotten Coast
Hi alarge, thanks for the post.

As I mentioned, my sample size was over several months of 25-man raiding. I was hit-capped & raided with two non-capped hunters, no Draenai in the group. I no longer have data because I quit raiding for a couple of months.

The raid numbers showed we were about even on DPS, but when adding the other hunters to the spreadsheet, it showed I should have significantly higher DPS. I dropped my hit to the level of the other hunters to allow for better gear/gems, which the spreadsheet showed would lower my DPS. In reality, my DPS went up to where I was consistently the highest DPS.

In trying to come to a mathematical conclusion, I noticed that the calculations did not look to be smooth functions & in fact have an adjustment. I unlocked the spreadsheet & changed the boss level to 82.999, which changed the hit needed dramatically (by 50 points), other values are incremental, by 1-2 points. This 50-point change is close to the amount I previously altered to increase my DPS.

Offline
Reply With Quote
Old 08/10/09, 2:11 AM   #332
alarge
Piston Honda
 
alarge's Avatar
 
Night Elf Druid
 
Vek'nilash
Originally Posted by jmichaelp View Post
Hi alarge, thanks for the post.

As I mentioned, my sample size was over several months of 25-man raiding. I was hit-capped & raided with two non-capped hunters, no Draenai in the group. I no longer have data because I quit raiding for a couple of months.

The raid numbers showed we were about even on DPS, but when adding the other hunters to the spreadsheet, it showed I should have significantly higher DPS. I dropped my hit to the level of the other hunters to allow for better gear/gems, which the spreadsheet showed would lower my DPS. In reality, my DPS went up to where I was consistently the highest DPS.

In trying to come to a mathematical conclusion, I noticed that the calculations did not look to be smooth functions & in fact have an adjustment. I unlocked the spreadsheet & changed the boss level to 82.999, which changed the hit needed dramatically (by 50 points), other values are incremental, by 1-2 points. This 50-point change is close to the amount I previously altered to increase my DPS.
Wow, you had a 25-man raid with *no* Draenei? Wow. I doubt I've seen that since BC came out. So you had no shammies in your raid?

Hit is not linear with respect to level. It is stepwise. Note that 7% hit is sufficient for heroics -- it is only raid bosses that require the full 8%.

Finally, the sort of anecdotal situation you described is all too easily chalked up to you simply increasing your skill. If you had WWS logs, we could dig into the issue deeper...

United States Offline
Reply With Quote
Old 08/10/09, 4:56 AM   #333
Pijn
Piston Honda
 
Night Elf Hunter
 
Doomhammer (EU)
I'm not sure, but i think he meant no draenei in his group (draenei aura only works on partymembers, not raidmembers). And yes i have seen some raids without shamans as well, but thats a guild/server issue; good shamans hard to find for us.

Anyway, alarge we had some previous discussions about hit rating of course, i won't be going into them again. You are right, one needs actual math to support his claims. I personally just believe that being hitcapped is priority number one, for several reasons (that i all named before). A hunter who does an infinite amount of dps and is hitcapped, can choose to 'lose' 1% hit (33 hit rating) or the equivalent in agility/AP/ArP/crit (so 33 of each). He ll notice a incredible loss of 1% of infinity dps if he loses 33 hit rating, but will barely notice a thing when he loses 33 agi.

Now, infinite doesn't make any sense of course, but this just gives an idea of what i mean; the higher your gear, the more important hit rating becomes (and yes i do know you realize this as well). I strongly believe that in Ulduar gear and beyond, not being hit capped is a poor use of talents/gem/gear (even it is just by 1 hitrating = 0.03% hit). For this reason the spreadsheet and femaledwarf.com will recommend an Ulduar geared hunter to be hit capped; it simply is a too powerful stat.

Offline
Reply With Quote
Old 08/10/09, 2:09 PM   #334
alarge
Piston Honda
 
alarge's Avatar
 
Night Elf Druid
 
Vek'nilash
Originally Posted by Pijn View Post
I'm not sure, but i think he meant no draenei in his group (draenei aura only works on partymembers, not raidmembers). And yes i have seen some raids without shamans as well, but thats a guild/server issue; good shamans hard to find for us.

Anyway, alarge we had some previous discussions about hit rating of course, i won't be going into them again. You are right, one needs actual math to support his claims. I personally just believe that being hitcapped is priority number one, for several reasons (that i all named before). A hunter who does an infinite amount of dps and is hitcapped, can choose to 'lose' 1% hit (33 hit rating) or the equivalent in agility/AP/ArP/crit (so 33 of each). He ll notice a incredible loss of 1% of infinity dps if he loses 33 hit rating, but will barely notice a thing when he loses 33 agi.

Now, infinite doesn't make any sense of course, but this just gives an idea of what i mean; the higher your gear, the more important hit rating becomes (and yes i do know you realize this as well). I strongly believe that in Ulduar gear and beyond, not being hit capped is a poor use of talents/gem/gear (even it is just by 1 hitrating = 0.03% hit). For this reason the spreadsheet and femaledwarf.com will recommend an Ulduar geared hunter to be hit capped; it simply is a too powerful stat.
My bad on Heroic Presence. I thought it was raid-wide.

My argument with the OP is the same argument I had with you, albeit from the opposite side. Until you are capped, each point of hit rating has a particular impact on your dps. That impact is measurable and measured. It turns out that, for hunters, hit is almost always worth more per itemization point than any other stat. How much more is also quantifiable (though the answer is not fixed and depends upon current stats and ratings). It is almost never worth twice as much as agility, for example. Which implies that you'd never trade away 40 agility to get 20 hit. Because of the variability of hit on gear, enchants, and other factors, this isn't an academic question. It is not at all unusual to be (for example) at 259 hit rating and face a question of whether to gem a +8 hit/+8 agi gem ("wasting" 4 hit) or a +16 agi gem.

The simple answer I give is the same answer as others: do the math. If you can't do the math yourself (which is quite complex) then trust the modeling done by others (i.e. use the spreadsheet). Bugs can (and do) exist in the spreadsheet. But that is no rationale for ignoring it when you don't like the answers it gives you. It doesn't really matter *why* you ignore it -- either because you believe you have anecdotal evidence that others who ignore it do well or because you slavishly follow a pseudo-religious maxim. In either case, you've left the realm of theorycrafting and moved into the realm of subjective argument. You may even be correct -- but without concrete numbers to back up your position, it is impossible to have an objective debate on the issue.

United States Offline
Reply With Quote
Old 08/10/09, 9:39 PM   #335
jmichaelp
Glass Joe
 
Night Elf Hunter
 
The Forgotten Coast
I got some numbers off the heroic dummy using frostbite bullets, Aspect Dragonhawk, Trueshot Aura, Wolf, MM 7-57-7 & no other buffs. I shot until OOM & tested twice (accidentally changed ammo from light shot so I didn't average, but DPS difference was about the same.) The 260 hit setup should have 16 more DPS according to the spreadsheet, but as you can see was lower.

AS-autoshot, CS-chimera shot, SS-steady shot, ArS-arcane shot, CSS-chimera shot serpent, AiS-aimed shot, WQAS-wild quiver auto shot, SrS-serpent sting, PS-piercing shots.

Hit Rating 214
DPS-2,942
AS
Hit - 26
Crit - 10
Miss - 0

CS
Hit - 6
Crit - 2
Miss - 0

SS
Hit - 9
Crit - 6
Miss - 1

ArS
Hit - 4
Crit - 2
Miss - 1

CSS
Hit - 5
Crit - 3
Miss - 0

AiS
Hit - 5
Crit - 2
Miss - 0

WQAS
Hit - 3
Crit - 3
Miss - 0

SrS
Tick - 25

PS
Tick - 57


Hit Rating 260
DPS-2,828
AS
Hit - 21
Crit - 11
Miss - 0

CS
Hit - 5
Crit - 3
Miss - 0

SS
Hit - 9
Crit - 6
Miss - 0

ArS
Hit - 4
Crit - 3
Miss - 0

CSS
Hit - 6
Crit - 2
Miss - 0

AiS
Hit - 5
Crit - 2
Miss - 0

WQAS
Hit - 0
Crit - 2
Miss - 0

SrS
Tick - 24

PS
Tick - 53

Offline
Reply With Quote
Old 08/10/09, 10:47 PM   #336
Osinin
Glass Joe
 
Night Elf Hunter
 
Nagrand
Originally Posted by jmichaelp View Post
tested twice
This doesnt mean that much sorry, 2 samples if far from conclusive about either side of the argument.

Alarge even mentioned about this in their intial reply to you as a potential problem as to why you are seeing the results you think you are

Originally Posted by alarge View Post
Your conclusion is unlikely, given the degree of applicability of hit to the full range of damaging attacks -- as long as you are considering average dps over a significant period of time.

Originally Posted by alarge View Post
b. Your sample size is insufficient



I am not saying your are wrong or right, but 2 samples doesnt mean anything sorry.

Offline
Reply With Quote
Old 08/10/09, 10:59 PM   #337
Æpic
Glass Joe
 
Blood Elf Hunter
 
Rexxar
9 more shots

Originally Posted by jmichaelp View Post
I got some numbers off the heroic dummy using frostbite bullets, Aspect Dragonhawk, Trueshot Aura, Wolf, MM 7-57-7 & no other buffs. I shot until OOM & tested twice (accidentally changed ammo from light shot so I didn't average, but DPS difference was about the same.) The 260 hit setup should have 16 more DPS according to the spreadsheet, but as you can see was lower.

AS-autoshot, CS-chimera shot, SS-steady shot, ArS-arcane shot, CSS-chimera shot serpent, AiS-aimed shot, WQAS-wild quiver auto shot, SrS-serpent sting, PS-piercing shots.

Hit Rating 214
DPS-2,942
AS
Hit - 26
Crit - 10
Miss - 0

CS
Hit - 6
Crit - 2
Miss - 0

SS
Hit - 9
Crit - 6
Miss - 1

ArS
Hit - 4
Crit - 2
Miss - 1

CSS
Hit - 5
Crit - 3
Miss - 0

AiS
Hit - 5
Crit - 2
Miss - 0

WQAS
Hit - 3
Crit - 3
Miss - 0

SrS
Tick - 25

PS
Tick - 57


Hit Rating 260
DPS-2,828
AS
Hit - 21
Crit - 11
Miss - 0

CS
Hit - 5
Crit - 3
Miss - 0

SS
Hit - 9
Crit - 6
Miss - 0

ArS
Hit - 4
Crit - 3
Miss - 0

CSS
Hit - 6
Crit - 2
Miss - 0

AiS
Hit - 5
Crit - 2
Miss - 0

WQAS
Hit - 0
Crit - 2
Miss - 0

SrS
Tick - 24

PS
Tick - 53
You got 88 shots in not hit capped and 79 shots in hit capped. Your wild quiver procced more. If you go to outdps.com, you can get a better understanding of hit rating. I didn't go for hit cap for a long time, being in the same "i don't care about hit" boat as you, and then finally gemmed and enchanted for a little more hit to get to the cap (after a lot of pressure from friends and guildies). After raiding hit capped for a couple months, there has been a noticeable increase in dps.

Offline
Reply With Quote
Old 08/10/09, 11:50 PM   #338
jmichaelp
Glass Joe
 
Night Elf Hunter
 
The Forgotten Coast
Actually, the sample sizes are fine (and my previous data was over several months time) but I did it twice more with the same results (forgot to mention these were double-blind tests as well.) The wild quiver is less damage per shot, so having it proc more would only decrease DPS.

I was previously i the "hit cap is best group" until I saw that it didn't appear to be, tested it over months time & found it wasn't. I'd been creating a temporary modification to the spreadsheet myself to account for this, but with not having raided for months and with a lot of changes to the game/spreadsheet, I didn't know if this still held true. Apparently it does. Thanks for the link to outdps, it confirms I have the correct understanding of hit. Perhaps this is an anomaly given my particular set of equipment. I'll continue to test it before I put the work into permanently changing the spreadsheet.

Offline
Reply With Quote
Old 08/11/09, 8:40 AM   #339
Korghal
Von Kaiser
 
Orc Hunter
 
Ahn'Qiraj (EU)
Originally Posted by jmichaelp View Post
Actually, the sample sizes are fine (and my previous data was over several months time) but I did it twice more with the same results (forgot to mention these were double-blind tests as well.) The wild quiver is less damage per shot, so having it proc more would only decrease DPS.
Wild Quiver is actually free dps, it it impossible for it to decrease DPS. Also, why do you say that the sample sizes are fine? Try using the /roll command ingame 100 times and see if you have rolled every number from 1-100.

If you really want to test this, try stacking up 1000 pieces of ammo, and auto shoot at a target dummy until out of ammo. Remember to specc out of Wild Quiver, to decrease the impact of RNG. Do this two times, one with hit capped, one with less. Preferably you could switch a hitgem, in order to equal item budget (20 hit cs 20 agi). You might say that this is to simplify a complex problem, since I've left out all the other shots, but in a real raid situation you also have buffs and debuffs to consider. There is no real way of testing this unless you can gather a 25 man group and try it out.

Don't get me wrong. I'm not trying to convince you or anything, but I think that it is a bit undermining to make such a bold statement without providing enough facts. It only leads to confusion.

Offline
Reply With Quote
Old 08/11/09, 12:21 PM   #340
alarge
Piston Honda
 
alarge's Avatar
 
Night Elf Druid
 
Vek'nilash
Originally Posted by jmichaelp View Post
Actually, the sample sizes are fine (and my previous data was over several months time) but I did it twice more with the same results (forgot to mention these were double-blind tests as well.) The wild quiver is less damage per shot, so having it proc more would only decrease DPS.

I was previously i the "hit cap is best group" until I saw that it didn't appear to be, tested it over months time & found it wasn't. I'd been creating a temporary modification to the spreadsheet myself to account for this, but with not having raided for months and with a lot of changes to the game/spreadsheet, I didn't know if this still held true. Apparently it does. Thanks for the link to outdps, it confirms I have the correct understanding of hit. Perhaps this is an anomaly given my particular set of equipment. I'll continue to test it before I put the work into permanently changing the spreadsheet.
There is no way that your sample sizes are sufficient with this setup, particularly given the degree of random variability in the data due to things like crit rate. First off, you need to be able to assert the mean dps and standard deviation for each gear configuration. This will take you at least 10 trials of ~500 shots each. Once you've established a mean for each gear configuration, you can compare them. Only at that point can you establish:

a) Which setup results in higher DPS

b) Whether the difference in DPS is actually statistically significant (e.g., the difference is greater than what you would expect from normal variations in the fight).

I'm not a statistician, but there are some that post in these forums. They could give you a much better idea of how to properly conduct such a trial. Another option is to run these trials (virtually) through a simulator. I know that there is a WoW simulator that folks use for other classes, and I expect that someone has done the model for hunters. Given that the relative value of hit for hunters has never had a significant challenge in these forums, I suspect that the simulator has come to the same conclusion as the theorycraft model -- point for point, hit is worth more than any other stat until you cap.

Edit: Looks like there is a SimulationCraft post in this very forum. If you really don't believe the spreadsheet, you should check your setup against the simulation model and see what it says.

United States Offline
Reply With Quote
Old 08/20/09, 1:01 PM   #341
Sorean
Von Kaiser
 
Night Elf Hunter
 
Deathwing
Pet +hit

I am hoping to get a little clarification here since I haven't seen it mentioned anywhere else.

Past research and findings indicated that Focused Aim did not affect a pet's ability to hit which was causing a DPS loss for the pet.

The current patch notes state:

All pets now receive 40% of their master's resilience and 100% of their master's spell penetration. In addition, if a player is at their appropriate spell hit chance or hit chance maximum, their pet will be at the maximum for spell hit chance, hit chance, and expertise. If they are below the maximum, their pet will be proportionately below those maximums.
Am I to assume that Focused Aim will now properly scale to the pet as well?

Does this make obtaining hit cap through any means (gear + talents) no different vs. the old method of obtaining hit cap through gear only?

Offline
Reply With Quote
Old 08/20/09, 2:27 PM   #342
Lupius
Von Kaiser
 
Blood Elf Hunter
 
Black Dragonflight
The Focused Aim issue was fixed several patches ago. The newest patch simply scaled expertise for pets.

Offline
Reply With Quote
Old 08/25/09, 4:08 PM   #343
Ktharsis
Von Kaiser
 
Blood Elf Paladin
 
Area 52
Has the rounding issue been fixed, or does a 99.99% chance to hit for the hunter still translate into a 99.0% chance for the pet to hit?

Offline
Reply With Quote
Old 08/26/09, 4:52 AM   #344
Zeel
Von Kaiser
 
Zeel's Avatar
 
Troll Hunter
 
Vek'nilash (EU)
According to DPS Spreadsheet this issue has not been fixed.

Still there was a blue post about patch changes that stated:

In addition, if a player is at their appropriate spell hit chance or hit chance maximum, their pet will be at the maximum for spell hit chance, hit chance, and expertise. If they are below the maximum, their pet will be proportionately below those maximums.
This leads to conclusion that Blizzard's conception of "proportionately" differs from the commonly accepted one.

Offline
Reply With Quote
Old 08/26/09, 7:13 AM   #345
Silentcow
Glass Joe
 
Tauren Hunter
 
Frostwolf (EU)
Originally Posted by Ktharsis View Post
Has the rounding issue been fixed, or does a 99.99% chance to hit for the hunter still translate into a 99.0% chance for the pet to hit?
Afaik it is not fixed.
So 7,99% hit rating for you is 7% for your pet.

Offline
Reply With Quote
Reply

Go Back   Elitist Jerks » Class Mechanics » Hunters

Thread Tools

Similar Threads
Thread Thread Starter Forum Replies Last Post
Do GMs ever fix rating bugs? Howard Roark Player vs. Player 11 12/13/07 10:25 PM
Season Three Rating/Personal Rating Mechanics heel Player vs. Player 5 10/28/07 2:00 AM
Hit Rating/Crit Rating (Rogues) SmokeMT Class Mechanics 6 06/07/07 12:41 PM
hit rating and crit rating formulas Curly Public Discussion 28 12/04/06 10:45 AM
Rating? Erving Public Discussion 117 10/09/06 6:54 PM