A decrease in Elemental DPS on the PTR !?!

[Tesshin20]

Ive copied over my Shaman over to the PTR and compared my dps after 5 minutes hitting the ragdoll. On live I end up with 72k without using any long cds. Just the standard rotations and no consumables.

Under the same scenario I get 69k on the PTR. On the PTR I have a ms of 225 instead of 32 on live though.

Any others having the same experience?

[Teye]

So, high latency made your dps go down.


Nothing to see here guys, move along.

[mrinvisable2]

no i have increase about from 74 to 96k

edit: well that was before 2 set nerf

[zenga]

no i have increase about from 74 to 96k

edit: well that was before 2 set nerf

wait ... with the same gear?

[Deleted]

no i have increase about from 74 to 96k

edit: well that was before 2 set nerf

Where? On a dummy or raid? What ilvl?

Doing more than 94k in ilvl 503 on live.

[Tesshin20]

So, high latency made your dps go down.


Nothing to see here guys, move along.

Thanks for practically shutting down my thread :*(

---------- Post added 2013-02-24 at 02:55 PM ----------

BTW... the tooltip on live says my Lightning Bolt does 53k damage. On PTR it says 44k.

[Cairm]

It might does, but this isnt the actual damage of your lightning bolt. Considering it doesnt account secondary stats and other factors.

Also, PTR is higher on latency, wich means that the server takes more time to respond, leading to spells queue timer being longer.

[Tesshin20]

It might does, but this isnt the actual damage of your lightning bolt. Considering it doesnt account secondary stats and other factors.

Also, PTR is higher on latency, wich means that the server takes more time to respond, leading to spells queue timer being longer.

There is an option to turn so that the tooltip shows all those things taken into consideration. Which I have turned on.

[Namana]

Thanks for practically shutting down my thread :*(

Why get upset about it? That's pretty much the cause. Losing a little dps when your latency is 7x higher is to be expected.

BTW... the tooltip on live says my Lightning Bolt does 53k damage. On PTR it says 44k.

If that's what your tooltip says on live, then you were most likely looking at it during a temporary int buff (trinket proc, etc) or in a 5 man with the damage buff. The BIS heroic T14 sim has the average LB hit at 52k, and that's including when it was cast with potions, trinket procs, engineering gloves, etc, plus every buff/debuf. That your tooltip would say higher self buffed, without any procs is pretty unlikely.

The difference in what my LB tooltip says on live and PTR is about a 13% increase, and I'd imagine it's about that for everyone else too, because the only LB changes they've made is a buff.

[dreadmist]

You dps a dummy for 5 minutes and expect that to be an accurate assessment of your damage output? Gimme a break, this thread should be shut down, you could have any number of inaccuracies skewing your numbers wildly, the latency being only one of many factors. Wait till a new version of simcraft comes out and use that to gauge how things change, not some pitiful half-assed dummy test.
The sim programs repeat a standard 5-6 minute fight anywhere from 5000-25000 times to calculate DPS values, which yield results much less prone to RNG.

[Tesshin20]

You dps a dummy for 5 minutes and expect that to be an accurate assessment of your damage output? Gimme a break, this thread should be shut down, you could have any number of inaccuracies skewing your numbers wildly, the latency being only one of many factors. Wait till a new version of simcraft comes out and use that to gauge how things change, not some pitiful half-assed dummy test.
The sim programs repeat a standard 5-6 minute fight anywhere from 5000-25000 times to calculate DPS values, which yield results much less prone to RNG.

So 5 minutes isnt enough. What kind of nonsense is this. And why wait for simcrafts? They are arent accurate at all anyway.

You do have the mental capacity to understand the subject of the thread. Dont pretend otherwise. According to the shamanism and elemental fury buffs the results Ive got arent even remotely what it should be no matter what scenario you test it in.

---------- Post added 2013-02-24 at 04:50 PM ----------

Why get upset about it? That's pretty much the cause. Losing a little dps when your latency is 7x higher is to be expected.




If that's what your tooltip says on live, then you were most likely looking at it during a temporary int buff (trinket proc, etc) or in a 5 man with the damage buff. The BIS heroic T14 sim has the average LB hit at 52k, and that's including when it was cast with potions, trinket procs, engineering gloves, etc, plus every buff/debuf. That your tooltip would say higher self buffed, without any procs is pretty unlikely.

The difference in what my LB tooltip says on live and PTR is about a 13% increase, and I'd imagine it's about that for everyone else too, because the only LB changes they've made is a buff.

I am not that ignorant that you believe. The buffs, the gear passive and active is exactly the same. It was the exactly the same scenario except for the latency. When standing still as a turret infront of the dummy the latency should not give a decrease of that magnitude we are speaking of.

[Deleted]

So 5 minutes isnt enough. What kind of nonsense is this. And why wait for simcrafts? They are arent accurate at all anyway.

I can do a 5min test on a target dummy and end up with 85k dps.
Then I can another 5min test on a target dummy right after that and do 75k dps.

It doesnt mean I now do 10k dps less, it just means that I got screwed on RNG on the second test (less crits, less overloads, less lava surge procs, bad trinket proc timing, etc).

[Namana]

I am not that ignorant that you believe. The buffs, the gear passive and active is exactly the same. It was the exactly the same scenario except for the latency. When standing still as a turret infront of the dummy the latency should not give a decrease of that magnitude we are speaking of.

The alternative is that your LB (and yours alone) was nerfed 20% while everyone else got a 13% buff. So yes, despite what you wish to believe, you were misreading your tool tip somewhere.

Also, latency doesn't change what a tooltip says, but it would (along with what dreadmist and Zenvon mentioned) account for losing 3k dps at a dummy.

[Endus]

I am not that ignorant that you believe. The buffs, the gear passive and active is exactly the same. It was the exactly the same scenario except for the latency. When standing still as a turret infront of the dummy the latency should not give a decrease of that magnitude we are speaking of.

Everyone else is correct. 5 minutes isn't even close to a representative sample. When I was testing CL proc chances [i]and ONLY CL proc chances, no other factors at all), I did about 2 hours of nothing but spamming CL on dummy targets. And that got me enough data to be relatively sure about that one factor.

What you're talking about involves many more factors, and requires much longer testing, to eliminate RNG from the equation. This is why the sims use about 10,000 iterations to get reliable data; because you need hundreds of hours of data to draw statistical conclusions on this stuff.

A single 5 minute test on dummies is basically irrelevant. Especially when there's clearly major factors (latency) affecting your results that you failed to control for.

[PaintOnASign]

I am not that ignorant that you believe. The buffs, the gear passive and active is exactly the same. It was the exactly the same scenario except for the latency. When standing still as a turret infront of the dummy the latency should not give a decrease of that magnitude we are speaking of.

You're misusing, or do not fully understand, the word "exactly". I'd be much more amazed if your two test actually were exactly the same. That would be quite astounding.

[JKim]

DPS definitely should increase with the recent buffs. I've PTR raided most of the testings for the bosses and we are doing pretty well so far.

[happyzod]

Latency shouldn't matter with spell queing.

[Eruionmel]

Latency shouldn't matter with spell queing.

This.

Also, I went from 89k live to 102k PTR. So you're either lying about the context, or you seriously fucked up your rotation when you tried it on the PTR.

[Moshne]

Latency does matter with spell queuing. That isn't how spell queueing works.

That said, you need to look at the spell data, not the recount output. Log your LB casts, look at the average cast value. Then compare that to your average cast on live with the same buffs. You'll see an increase. Your methodology is just flawed, but you may be able to salvage your data if you kept in in a log rather than letting Recount/Skada do your interpretation for you.

[Eruionmel]

Latency does matter with spell queuing. That isn't how spell queueing works.

That said, you need to look at the spell data, not the recount output. Log your LB casts, look at the average cast value. Then compare that to your average cast on live with the same buffs. You'll see an increase. Your methodology is just flawed, but you may be able to salvage your data if you kept in in a log rather than letting Recount/Skada do your interpretation for you.

It matters, but not at something as low as 200 latency. As long as he's spamming whatever spell he wants to cast, 200 latency is going to have basically 0 DPS loss.