Please feel free to look over all the equations and see where I might be in error. I have no problem believing that a 0.2/2.9 (6.9%) increase in shot frequency can overcome 4.2/83.1 (5%) DPS difference in base weapon damage. The percentage is actually much lower when you factor in RAP, ammo, and scope.

I think I mathematically proved in a thread on the TKASomething.com forums that it would take lowering your RAP to -300 or so in order to have them equal out, stats on the weapons not taken into consideration. For any hunter, eliminating dead time in your rotation is a huge difference.

And thank you for the kind words.

Interestingly if you modify your latency in the Shot Rotation tab you begin to see the DPS values of the various weapons change. I was quite shocked at the original findings when it was set to the default of 0 latency, but when I changed it to account for my horrid Australian ping time, weapon selection fell more in line with what I expected.


Oh and Cheeky, thankyou so much for your continued work on this project.

Given that there are 2^n combinations of things that can happen, and given that dps can stop increasing over some haste regimes, perhaps you can just consider the dragonspine proc as something the user explicitly inserts into a shot rotation (with an associated cooldown related to its expected trigger frequency.)

I'd do the same thing for quickshots, too -- you would have to insert that as well. But what you have there isn't broken, so it's probably easier to leave it alone.

The idea is to model the damage from a particular series of events (shots selected, procs triggered) that the user chooses, and which the spreadsheet can choose to enforce with a cooldown matching the proc or shot frequency. If you take that approach, the only thing you need to model is the cooldown (don't even have to model a quickshot rotation any more -- just the quickshot's expected cooldown. This is longer with 1/5 talent and shorter with 5/5 talent, of course.)

I hope you saw my last post, Cheeky, and I also have another new question
For armor reducing items you list the armor penetration value in the sheet, but I just want to make sure by asking: atm it has no effect on the dps calculation, right?

Yes, I saw your previous post. Updated ammo stats and the bow will be in the next version. You can set the Quick Shots rotation to anything you'd like. Feel free to make it identical to your normal shot rotation. I'd just verify that the cool downs for Arcane and Multi shot aren't forcing you to wait till the point you are missing a potential extra auto shot.

Armor penetration should be working just fine. You can see the Effective Armor on the Calculations page. I now also support the 4-piece Beast Lord set bonus.

Ah, good to know. Great
Then in fact most armor reducing items are not as good as their stat-loaded counterpart.

I know that it makes a small difference depending on how much armor the mob has, but I tried to compare most items (Illidan ring, gronn shoulders and madness trinket) to their counterparts on all the likely different levels of mob armor and it never really surpassed the stat-item.

If you have a boss down to around 15% armor it starts to make sense, but its rather unlikely that this happens a lot.

Hello there Cheeky, quick question about your spreadsheet - after inputting all my gear spec and shot rotation I noticed that when un-talenting Imp AoTH I actually gained around 40 DPS. Is this the spreadsheet accounting for an altered shot rotation or am I just missing something entirely?

(Can email the spreadsheet as I have it if necessary)

For some MM builds, the hasted Auto/Steady shot rotation is less DPS than an unhasted 1:1.5 rotation. Try forcing a 1:1.5 rotation for Quick Shots and see if there is any DPS increase.

I provide the ability to specify both rotations. Play around with them, but IAotH may not be a good idea for everyone.

Hey cheeky I cant seem to be able to fill in the latency cell. Any idea what's causing this? I have just started playing around with the new excel 2007 so I am still playing with it trying to figure it out. any help is appreciated!

Also great job on the spreadsheet! keep up the good work.

It is an unlocked cell that accepts a decimal value between 0 and 1. Are you trying to set a latency greater than 1 second, or less than 0?

That should be valid. I just tested with Excel 2002, and it seems to be working fine. I do not have 2007 to test with, unfortunately.

Are you getting an error message of some kind? The cell is unprotected, so you could disable the Validation, and see if that helps.

Might it be using a German version of Excel or something? 0.2 doesn't work for me either but 0,2 does. Just a thought.

This one might be a bit buggy. Adding th [Dragonspine Trophy] involved a fair bit more changes than any other trinket would. Don't hesitate to point out mistakes if you find them.

Change Log:
- Added new gear (Quickstrider Moccasins, Cobrascale Hood, Primalstrike Belt, Primalstrike Bracers, Primalstrike Vest, Primal Intent set bonus, Black Bow of the Betrayer)
- Fixed a bug in the table that calculates effective rates of pet specials.
- Corrected gear stats (Pristine Fire Opal, Timeless Arrows, Timeless Shells)
- Modeled the proc of the Black Bow of the Betrayer.
- Serpent Sting now triggers the Global Cool Down.
- Fixed an issue with Furious Howl not accessing the correct ability rank.
- Added modeling for the Dragonspine Trophy. The user can input Shot Rotations for both under the effects of Dragonspine Torphy's haste as well as that combined with Quick Shots. The calculations tab computes the uptimes of the 4 potential states based on a user-supplied ppm rate of the Dragonspine Trophy.I am using a 1.7 ppm rate for the DST, but it is user settable right now, until I can confirm the numbers.
- Total Hunter DPS, MPS, and Shots per Second calculations are now done factoring in the effects of the Dragonspine Trophy as well.

Cheeky this sounds really stupid but how do I set 1:1 shot rotation and 1:1.5 shot rotations?

You can either hand select each and every shot in the different matrices, or use the buttons I provide (which require macros be enabled.)

The "MaxSpecials" button on the normal Shot Rotation tab seems to build an acceptable 1:1.5 Shot Rotation.

What latency value did you use?

The [Skyfire Hawk-Bow] is a 1.74 second rotation unhasted, and 1.51 with Quick Shots. Normal speed would be at about my limit to handle, and IAotH would be a waste of talent points for me.

It's a great weapon. It's on an almost perfect speed for a 1:1 cycle and MM/Survival Hunters. I never got one to drop, so I can't say what DPS I'd see with it. I do trust the math though, and believe it to be a far better weapon than some of the earlier epics.

The Consortium Rifle from Consortium revered goes even higher then the skyfire bow on the spreadsheet. It seems that Blizzard has failed at itemization again.

This is because of the way you have specified your latency/lag - the 2.4 weapon series under quick shot haste state (quiver/pouch, serpent swiftness, quick shots) gives you a tuned fire rate that eliminates any down time between steady shots (i.e. removes any wasted time from dealing damage with steady shot), and does not clip autoshot (i.e. the end of the steady shot and the beginning of the autoshot 0.5 seconds later does not exceed the normal autoshot frequency of the weapon). That is an autoshot cycle time of 1.51s and a steady shot time of 0.94s.

If you introduce human interface lag of 0.1 seconds (i.e. it takes you about 0.1s to tell the game you want to shoot your next shot) and a latency time of 0.2s (i.e. it takes your computer 200ms to tell the WoW server that you have requested a shot), then you are no longer have a perfectly tuned auto/steady shot timings. Latency does not translate directly into lost shot time in practice, but Cheeky has a latency factor (0-1) that simulates these delays (maybe not up to Australian standards over 1 second latency - but should work to model most states)

A couple of issues about version 26 of your spreadsheet that I've found so far, Cheeky:

(1) Quickstrider Mocassins is listed as having 15mp5 instead of 15 hit rating.
(2) If you set your Target Armor to a value under that of your gear's Armor Penetration amount, you still get an increase of DPS (and the Reduction goes negative).

I'll post more if I find any

Main post updated to include the previous 3 versions of Cheeky's spreadsheet (will be kept up-to-date, obviously). If you see any strange behaviour with items, etc, checking with these previous versions and comparing would be a benefit for tracking down potential bugs =)

I thought about it being a latency thing after I took off for work. I guess I had always ignored that part of the spreadsheet because I've seen mine jump around so much at times. It seems, as of late, to have settled to about 200-300ms. At these values, the Skyfire Hawk Bow was outpaced by the Steelhawk/Wolfslayers/Sunfury and I'm sure most everything else.

Interestingly (to me) at .2secs lag, Wolfslayer still beast out Sunfury as you claimed. At .3secs, however, the Sunfury once again outpaces the Wolfslayer. While I still agree to some extent that Blizz has really botched hunter itemization when I see that in a perfect world a SH/early Kara weapon destroys all of the others, in a non perfect world it seems to fall back into place the way Blizz intended. For my part, I think I will probably still try to pick up that Sunfury for fear that I would confidently use the Wolfslayer but find out that lag is gimping me. Regardless, thanks for the great spreadsheet Cheecky, it never ceases to amaze me with its amount of detail!