Kolusius

I assumed it wouldn't as well, but I've missed things like that before. I'll check it out now actually. But it is giving the penetration from gear to pets on the spreadsheet I believe.

 

Cheeky

Yep, it's probably a bug. I'll get it corrected.

Cheeky of Havoc

 

◊ Kurisu

Hey Cheeky, let me know if I am miscalculating here but according to the newest version of the sheet (as a troll using a bow) I get 100% hit at 96 rating, 96 is 6.07% but the cap is supposed to be 9%. As a troll I have 5 weapon skill which if I read correctly should be .50% (1 skill = .1 correct?). I used a setup with Valanos's Longbow and got 100% and I swapped it out for a Toxic Injector (I chose a random gun without hit) and dropped down to 97.09%.

SC Alts: Toremu - shaman

The other main
Kurisa- Druid

 

Glaurong

Right now the testing seems to indicate the +5 weapon skill is worth 3%. It seems like a huge amount but that is the latest theory.

Log Parser for BM Hunters (Right click, save as) - Updated 10/11/2007

 

◊ Kurisu

Ok, I guess I must have missed that in the other weapon skill thread, thanks for the clarification.

SC Alts: Toremu - shaman

The other main
Kurisa- Druid

 

galathian

Great spreadsheet BTW.

I put all my gear and talents in. Double and triple checked everything.
The only thing that is off is my mana and hp. Not too concerned about that atm.

I'm just doing some testing using the steady rotation.

Spreadsheet says my DPS should be around 860. Highest I can get is about 500.

In order for me to reach the 860, does it have to be for 50 shots?
I've not yet had a chance to try it out on a boss in kara or any other dungeon.

 

Belzi.ET

Make sure to insert a possible armor-absorb for your target.
As far as I remember the downloaded version from cheeky has a 0 armor-target.
There are evaluations going on (isn't it?) about mob-armor.

Just try to enter armor to get about 25-30% absorb.


You can go to Netherstorm an shot Dr. Boom for steady-testing. He's a little north from Area 52.
He's one of the best targets to test shot-rotation because he doesn't move nor defend himself =)

 

peteriong

Test Set Up:
Target: Lv 70 Void Conduit in NetherStorm
Aspect: Viper
Patch Version: 2.1.3 Live Server
Ping: Around 100~150 ms
UI: Recap

(a) Total proc times is monitored by myself
(b) Buff gain proc times is monitored by RECAP (Duraction Counts)
(c) Refresh proc times = Total proc times - Buff gan proc times
(d) Estimated buff up time = Buff gan proc times (Duration Counts by RECAP) * Average duration (Average Duration by RECAP)
(e) Combat time is monitored by RECAP. This is the duration between my first shot and last shot in each session.


[Test 1]
Macro sequence Auto+Steady Rotation, 15%+20% Haste, 1.96 Autoshot Speed

Total 5400 Shots(150 Shot each session), 2082 Auto Hit, 620 Auto Crit, 2083 Steady Hit, 615 Steady Crit, 0 Miss
Total proc times = 253
Buff gain proc times = 161
Refresh proc times = 92
Estimated buff up time = 2090.8 Sec
Total combat time = 5376 Sec

Proc chance = 4.69%
PPM = 253 / [(2082+620)*1.96] * 60 = 2.87
Buff up time percentage = 2090.8 / 5376 = 38.89%
Refresh/Total proc percentage = 36.36%


[Test 2]
Auto Only, 15%+20% Haste, 1.96 Autoshot Speed

Total 2700 Shots(150 Shot each session), 2060 Auto Hit, 640 Auto Crit, 0 Miss
Total proc times = 152
Buff gain proc times = 109
Refresh proc times = 43
Estimated buff up time = 1331.5 Sec
Total combat time = 5296 Sec

Proc chance = 5.63%
PPM = 152 / [(2060+640)*1.96] * 60 = 1.72
Buff up time percentage = 1331.5 / 5296 = 25.14%
Refresh/Total proc percentage = 28.29%


[Test 3]
Steady Only, 15%+20% Haste, 1.96 Autoshot Speed
(Working in Progress...)

 

Voxel

Macros execution in MS Excel 2007 is disabled by default.
To enable it, you have to add folder (where this document is) to the list of trusted locations:

• press "Office button" (icon in the top left corner) ->
• press "Excel options" button (in the bottom of the arrived page) ->
• choose "Trust Center" option (in the list on the left side) ->
• press "Trust Center Settings" button ->
• select "Trusted Locations" option (in the list on the left side) ->
• press "Add new location" button ->
• specify the path to the folder where Cheeky's file is stored on your local PC ->
• re-open your excel-file. Macros should work now

 

Cheeky

I really appreciate these tests. They are much better than what I have run in the past. I'm going to update the proc rate in the spreadsheet based on your final results when you total all of them together. I'm curious to see if the differences in proc rate based on Auto only versus Auto + Steady are maintained. The only thing I can think of that would account for that (based on what I know about ppm mechanics) would be if the Don had a 2-3 second internal cool down. Seems unlikely, but it would explain the results.

Can you montior the re-procs to see how close together you can get them? I'm not sure how to get refreshes to appear in a combat log so we have time stamps. I'm assuming there is an event for proc refresh, since UI mods pick up on it. Does anyone have the Lua skills to write a small AddOn that will dump them to the comabt log?

Cheeky of Havoc

 

galathian

Ok duh. I new it was something on my end. Adding the armor in brought things back down to reality. Thanks!

 

peteriong

I have all my options checked in my combatlog window. But there is no message when a refreshes occurs. I believe this is the reason why there is no UI for tracking refreshes. This makes massivly testing Santos very hard.

I didn't formly record the behavior of refreshes. From my experience, I always see "Buff up->count down to 9 Sec remains -> Reproc -> count down to 9 Sec remains -> Reproc". This leads me to believe there is no internal cooldown.

Here is my thinking about how to estimate Refresh/Total proc percentage if we only know auto shot speed and proc chance, using data of Test 2.

a) Assumes there are only two cases "Proc->Reproc" and "Proc->Reproc->Reproc".

b) In Test 2, 1 shot per 1.96 second, it means there are 5 shots between a Santos buff up and Santos buff off.

c) There is 5.63% chance for these 5 shots to proc. It means there is 1-(0.9437*0.9437*0.9437*0.9437*0.9437) = 25.15% chance to have at least one Reproc after a Proc.

d) Reproc extends the original buff time minimum 1.96 second maximum 9.8 second. The average buff extended time is 5.88 second, allowing 2 shots before the extended buff is off.

e) There is 5.63% chance for these 2 shots to proc. It means there is 1-(0.9437*0.9437) = 10.94% chance to have at least one Reproc after a "Proc->Reproc". And 0.9437*0.9437 = 89.06% chance to have no Reproc after a "Proc->Reproc".

f) Assumes RECAP records X Proc. There shoud be 0.8906*0.2515*X = 0.2240*X "Proc->Reproc" and 0.1094*0.2515*X = 0.0275*X "Proc->Reproc->Reproc".

g) It means 1*0.2240*X + 2*0.0275*X = 0.279*X cannot be recorded by RECAP. This is the estimated Refresh/Total proc percentage 27.9%

h) In Test 2 the actual Refresh/Total proc percentage is 28.29%. This is quite close to 27.9%.

 

peteriong

In this thread
Proc Per Minute Mechanics & Haste

There is another Santos test in #72. The result is UI counted 1.3 PPM under 2.25 auto shot speed.

To estimate the PPM with refreshes. Assumes proc chance is 1.72/60*2.25 = 6.45%. There are 5 shots between Buff up and off. Average buff extended time is 4.375 second, allowing 2 more shots. The estimated Refresh/Total proc percentage is 26.33%. It means the estimated PPM with refreshes is 1.64. This is close to the result in my Test 1.

I believe the PPM of Santos is between 1.64~1.72 under auto shot only rotation. But how the PPM changes after we fires steady shots between auto shots?

If we assume the proc chance from steady shot is the same with proc chance from auto shot under a specfic shot speed, we should expect the PPM under Auto/Steady rotation to be 1.64~1.72 * 2 = 3.28~3.44.

But Test 1 shows 2.87 PPM. This leads me to believe the proc chance from steady shot is lower than that from auto shot.

Comparing Test 1 and Test 2, firing steady shots every 1.96 second contributes 2.87-1.72 = 1.15 PPM to the total PPM. Proc chance from steady shot is 3.76%.

I hope results from Test 3 will verify this 3.76%/1.15 PPM results. But to have steady shot only rotation I have to fire steady shot every 1.5 GCD to prevent auto shot firing. This means that in Test 3 I am firing steady shot every 1.5 second, much faster than 1.96 second in Test 1.

 

◊ Lactose

Could it be something like the following?
Auto Attacks have normal proc rate, adjusted by hastes.
Special attacks have 0.5x [Auto Attack proc rate].

Look, Lactose, we'd rather you didn't eradicate the whole human race.
- Sam & Max

 

Glaurong

I believe you can eliminate steady shots from the PPM calculations. If we find the base PPM of the rifle we should be able to extrapolate the % chance to proc by our hasted auto shot speed and apply that same % chance to all steady shots fired.

The constant proc % and increase in PPM between your two latest tests support this theory.

Log Parser for BM Hunters (Right click, save as) - Updated 10/11/2007

 

Grondarg

I was testing out different specs with my gear and I found a possible bug. My current 18/43/0 spec is built with Endurance Training instead of imp AotH. If i move the 5 points in ET to Imp AotH my total dps actually drops from 1032.49 to 988.07. Since one has no effect on damage and the other is proc based there shouldnt be a difference. i get the same result when trying a 5/21/35 build i'm considering. I'm currently using version 22 of the spreadsheet.


I also have a feature request. Is it possible to add a page that calculates time to mana exhaustion? I know that you have one on the shot rotation page but that doesn't account for pot usage. Basically I would like to be able to see how long i can extend my mana pool accounting for mp5, pot usage timer, and shot rotation.

thanks

 

Cheeky

What rotation have you specified for Quick Shots? Most MM Hunters will see a DPS drop if they switch from a 1:1.5 rotation to a 1:1 rotation with Quick Shots.


Yeah, I'll work on that.

Cheeky of Havoc

 

Grondarg

Well that would explain it. It never occured to me that the quick shots tab was used for that. I sort of assumed that it was for testing a shot rotation while Quick Shots was active. Sorry for the misinformation then. Should the quick shots tab be used for all shot rotation testing only if you have imp Hawk instead of the shot rotation tab?




Thanks.

 

rockyourjon

Although it wasn't listed in the changes, I've noticed a difference in the DPS calculations for armor negation between version 27 and 28. I'm just curious what the math is now (it doesn't list the armor of the target as it did previously, I used to set it at 8000) and how good of an indicator it is. For example, for a 41/20/0 BM hunter like myself it currently says that Gronnstalker's Spauldors (w/ 10crit / 7agi gems) are about on par for DPS with Mantle of the Tireless Tracker, which I find strange. I would've imagined Blizzard intended Tier 6 to do more damage.

My guild is just starting to clear Black Temple/Hyjal, and I'm trying to pick out the gear I want before it drops, and I use a lot of the information on these forums and the spreadsheets to make decisions. Doing my best to be an informed PVE hunter.

Edit: I'm a huge noob, I just realized that all you did was move the location of that field (it's in the settings tab now, not the gear tab). The math looks more accurate now.

 

Ernesti

Hey Cheeky, great spreadsheet. One suggestion though, could it be possible to add other BM hunters and their pets FI to the group buffs?

 

Sylvarielle

Hello,

I cannot open any of those spreadsheets, because they are encoded using a password that contains characters that are not recognized by the French version of Excel. Excel just prompts an error message whenever I try to open them. Is it possible to encode them with a password that is more international ? Thanks ^^

 

Cheeky

Can you tell how to do that? I only have access to Excel 2003 for English. I have no way to test what you are looking for.

Cheeky of Havoc

 

Sylvarielle

mmmmh... I have no idea ? I can only assume that your spreadsheet is password protected (so that your calculations or formulas are not tampered with) and that the password you chose contains non UTF-8 characters, which could pose a problem for my little Excel.

Thanks for the fast answer however. Don't sweat it if you can't make it work, the spreadsheet looked like a pretty useful tool though ^^

 

Jaade

Using french excel and the spreadsheet works fine here (no password asked). Sylv you might want to check your security options (I'm no excel expert ^^).

Great tool, thanks a lot Cheeky.

 

Raba_Storm

In continuation to Post #637

Just downloaded version 28, and I'm still experiencing the same problem with incorrect data of item [Violet Signet of the Master Assassin]