Hunter Spreadsheets -- In development - Page 92

Korikin · 02/05/08, 12:54 PM

Originally Posted by Cheeky

This whole mechanic is so fucking stupid

I hate not knowing WHY something works. And it's painful to try to explain to other hunters in my guild why they should switch macros if I can't tell them a single 100% for sure reason why it works other than "it just does". Plus, gawd knows this "mechanic" is going to be "fixed" sometime in the future with no warning or acknowledgment.

Rokh · 02/05/08, 1:12 PM

Nice work, Cheeky, looks promising. Are we sure that the cast time of Auto Shot isn't affected by Haste? I'm not sure how to go about collecting the data in the same manner as you did, but perhaps you/someone could log some tests without a quiver or any haste gear, then do the same with as much haste gear as possible?

Also remember that in 2.4 the combat log is being reworked, so hopefully the timing issues will be fixed, and we'll get a much more consistent and precise log to work with when doing things like this.

Cheeky · 02/05/08, 1:16 PM

Originally Posted by Rokh

Nice work, Cheeky, looks promising. Are we sure that the cast time of Auto Shot isn't affected by Haste? I'm not sure how to go about collecting the data in the same manner as you did, but perhaps you/someone could log some tests without a quiver or any haste gear, then do the same with as much haste gear as possible?

Also remember that in 2.4 the combat log is being reworked, so hopefully the timing issues will be fixed, and we'll get a much more consistent and precise log to work with when doing things like this.

I haven't done testing myself, but going with 0/5 Serpent's Swiftness and no quiver is the next logical step. In PMs Lactose suggested the range to mob (i.e. flight time) may be being applied to one shot and not the other, this would also complicate things, as well as make it completely un-modelable.

I've seen a little of the 2.4 combat log changes, but I don't remember hearing if the timestamps will all be server side (which is what we'll need.)

Rokh · 02/05/08, 2:06 PM

Just in some limited testing, it seems Auto Shot is affected by latency, and haste. I'm not sure if the speed of the weapon affects this, if quivers affects it, ect. It will take a lot of testing and data to rule out each of these, but I'm definitely noticing a difference between a 2.7 hasted weapon and a 3.0 un-hasted one, in just limited testing.

Cheeky, did your auto shot data come from the same variables each test, or were things like spec, weapon, and haste changed?

Cheeky · 02/05/08, 2:16 PM

Originally Posted by Rokh

Just in some limited testing, it seems Auto Shot is affected by latency, and haste. I'm not sure if the speed of the weapon affects this, if quivers affects it, ect. It will take a lot of testing and data to rule out each of these, but I'm definitely noticing a difference between a 2.7 hasted weapon and a 3.0 un-hasted one, in just limited testing.

Cheeky, did your auto shot data come from the same variables each test, or were things like spec, weapon, and haste changed?

The data was from the posted logs of 3 different Hunters covering 5 different tests. I have no idea what haste effects were in place, but I think all were BM Hunters.

We need controlled data to figure it out for sure.

Uncas · 02/05/08, 3:30 PM

Originally Posted by Cheeky

Doing all that, using a 0.15s latency and my gear, led to the following DPS totals (Hunter only):
[Serpent Spine Longbow] - 827 DPS, 17:33 Auto:Steady in regular, 22:28 in QS
[Sunfury Bow of the Phoenix] - 863 DPS, 20:30, 23:27
[Steelhawk Crossbow] - 840 DPS, 20:30, 24:26
[Wolfslayer Sniper Rifle] - 847 DPS, 20:30, 25:25
[Barrel-Blade Longrifle] - 848 DPS, 22:28, 25:25

3 - This whole mechanic is so fucking stupid I'm seriously thinking about giving up on modeling Hunters. I can't for the life of me figure out how any of this is "working as intended" and I really wish Blizzard would at least give us some indication of what they'd like our mechanics to be.

First: GD, I skipped testing sunfury and went straight for the serpent spine on Dr. Boom, maybe i should be using sunfury now

Two: re #3, Amen I spent an hour on the phone bitching to my friend (who doesnt even play WoW) about how this would be nice. And that I am not saying, tell me what gear and what macro, but explain to me how we are mechanicaly supposed to be doing DPS. and how we are supposed to actually achieve that.

Three: thanks for the 2 roll system, now I can get a better grasp on effects of losing +hit, With my current gear, it seems I should replace my survival favortie yellow (4agi 4 hit) with 8hit.

anjel · 02/05/08, 4:19 PM

Cheeky,

I have two questions regarding your spreadsheet (version 42):

1.) Currently specced standard raid BM, I am thinking of respeccing to survival where I'd like to use 'auto/steady/steady' rotation. Any chance you want to add this type of rotation into your spreadsheet? I had a chance to try this on the last PTR for about 2 hours on Dr. Boom. In my tests I was able to consistently beat my BM dps by a substantial amount (naturally not factoring in the pet damage nor the FI buff). I was going to run few more tests to record the numbers, but they brought down the PTR before I was able to do so. From memory, I believe the DPS difference between BM with auto/steady rotation and Survival with auto/steady/steady rotation was bout 100-150.

2.) I am not certain your spreadsheet works correctly to account for the thundering skyfire diamond. When I load my current gear, the difference in the reported dps with and without this metagem equipped is only about 1.5 dps, which seems to be rather low. What I find interesting is that with or without the metagem equipped, the shot rotation tab claims that the 50 shots are fired in the same time period.

Cheeky · 02/05/08, 4:37 PM

Originally Posted by anjel

Cheeky,

I have two questions regarding your spreadsheet (version 42):

1.) Currently specced standard raid BM, I am thinking of respeccing to survival where I'd like to use 'auto/steady/steady' rotation. Any chance you want to add this type of rotation into your spreadsheet? I had a chance to try this on the last PTR for about 2 hours on Dr. Boom. In my tests I was able to consistently beat my BM dps by a substantial amount (naturally not factoring in the pet damage nor the FI buff). I was going to run few more tests to record the numbers, but they brought down the PTR before I was able to do so. From memory, I believe the DPS difference between BM with auto/steady rotation and Survival with auto/steady/steady rotation was bout 100-150.

2.) I am not certain your spreadsheet works correctly to account for the thundering skyfire diamond. When I load my current gear, the difference in the reported dps with and without this metagem equipped is only about 1.5 dps, which seems to be rather low. What I find interesting is that with or without the metagem equipped, the shot rotation tab claims that the 50 shots are fired in the same time period.

1) I'm currently working on the model for this. Read up a few posts to see how that's going.

2) I don't model the [Design: Thundering Skyfire Diamond], because it would be very difficult to do, and by all accounts it's a much worse choice than the [Design: Relentless Earthstorm Diamond]. (This may or may not be true with the new macro.)

Uncas · 02/05/08, 5:30 PM

Simple additional suggestion:

Gear Tab, Add a row below Row 6 that shows the totals (aka, row 93), so that as you scroll and change gear/gems/enchants, the totals remain visible.

anjel · 02/05/08, 6:36 PM

Originally Posted by Cheeky

1) I'm currently working on the model for this. Read up a few posts to see how that's going.

2) I don't model the [Design: Thundering Skyfire Diamond], because it would be very difficult to do, and by all accounts it's a much worse choice than the [Design: Relentless Earthstorm Diamond]. (This may or may not be true with the new macro.)

Cheeky, thanks for the quick reply.

1.) I think you are referring to the post where you mention trying to model the new macro? I guess by modeling that you would in effect handle the auto/steady/steady rotation for gear/specs where effective weapon speed was above certain threshold, e.g. MM or Surv builds.

2.) Is it more difficult to model [Design: Thundering Skyfire Diamond] than IAoH or DST?

In any case, what do you think about these approximations/calculations (assuming pure auto/steady rotation with a BM build):

With no metagem your spreadsheet reports (using some random gear and AS/SS rotation):
hunter dps: 806.55

If I equip [Relentless Earthstorm Diamond] your spreadsheet reports (with the same gear & rotation):
hunter dps: 818.25

If I equip [Thundering Skyfire Diamond]:
I do the following calculations (for simple auto/steady rotation):
240 haste rating --> 240/15.7 ~= 15.3% faster attacks for 6 seconds
Assuming 1ppm, or approx. 10% uptime, you could say it averages out to about 1.53% faster attacks
That in turn (again, assuming AS/SS rotation) is about 1.53% extra dps
That would mean, hunter dps should be: 1.0153 * 806.55 = 818.89

Do you think I signifficantly oversimplified anything?

peterk0 · 02/06/08, 3:34 AM

I also did som testing on Dr. Boom. I focused on number of shot fired, instead of DPS, because I find it more clear and not luck (crit) based. What I did was this: I was counting the number of Auto Shots and Steady Shots fired during the 2 minute period, while changing the haste rating for each test. I also did it with the new 3:2 macro and with AotV and only with Leggionkiller (base speed 2.9)

My discoveries:
1, The number of Steady Shots was always the same, no matter if my haste rating was 0 or 95 (or between these numbers).
2, The number of Auto Shots varied based on haste, but in very small chunks i.e. with 95 haste (1.98 attack speed, instead of 2.1) and period of 2 minutes, it fired like 2-3 more Auto Shots. (The number of Auto Shot might also be influenced by latency, which at the time of testing was ~150 ms)

So my personal conlusion is that with new 3:2 macro, haste rating is not as good as for example crit (which is often lacking the haste items), because it only affects 1/3 of your shots. I think it also affects SS, but you need more like 95 haste.

I'm going to do some more testing with improved AotH and report my findings.

Hellraza · 02/06/08, 4:19 AM

What do you mean by "the new 3:2 macro" and a 3:2 rotation as a whole? I have seen it a lot lately, but I couldnt find a good example of this kind of rotation. There are also a lot of macros spread all over the forums.

I know that this is not the best place to ask this question so I want to apologize.

Belzi.ET · 02/06/08, 5:38 AM

Originally Posted by Hellraza

What do you mean by "the new 3:2 macro" and a 3:2 rotation as a whole? I have seen it a lot lately, but I couldnt find a good example of this kind of rotation. There are also a lot of macros spread all over the forums.

I know that this is not the best place to ask this question so I want to apologize.

This post by synergi quotes and links a post by howitzer from the WoW-forums.
Howitzer explains the theorie and technique of this new rotation model.
http://elitistjerks.com/f31/t11420-h...ing/p106/#2632

Cheeky · 02/06/08, 10:48 AM

Originally Posted by Hellraza

What do you mean by "the new 3:2 macro" and a 3:2 rotation as a whole? I have seen it a lot lately, but I couldnt find a good example of this kind of rotation. There are also a lot of macros spread all over the forums.

I know that this is not the best place to ask this question so I want to apologize.

In reality all the new macro results in is a Steady Shot spam. Most of it is just garbage, but what it tries to do is cast Steady Shot as often as possible, and let Auto Shots fire when they can in between chains. A lot of people see the pattern Auto, Steady, Steady, Auto, Steady (repeat), yielding 3 Steady Shots for every 2 Auto Shots.

In pouring over logs I think this only really works because latency isn't constant, and anytime it increases you have a greater chance of getting an Auto Shot through. The GCD on the client prevents lowering latency spikes from providing much of an effect.

Reinforce · 02/06/08, 12:51 PM

I've used this spreadsheet to check out how to max out my dps, and for the shot rotation it seems the 1:1 priority rotation is the best one. I get like 40 extra dps then the Steady/Auto Shot.

Now the problem is, I don't know the 1:1 priority macro. Is this just the Auto/Steady Shot "spam" macro, spamming this macro while also clicking on both arcane & multi whenever their cd is off?