2.3 Ele Changes - Final Attempt at Feedback

#0 - Nov. 5, 2007, 10:22 p.m.

UPDATE: It got announced on 10/11/07 by Eyonix that some of our coefficient will be coming back before 2.3 goes live. This is a success, as most Ele Shamans should now see the buff they were promised. Higher geared players will probably still see a nerf... so this isn't over, but this thread is.

Salthem - you want constructive, you're getting constructive.

The aim of this post is to :
- Show my own 3rd set of PTR testing
- Explain the changes for anyone who doesn't get why we are complaining,
- Try to sum up the issues
...and maybe, just maybe, get a dialogue with one of them blue guys going.

**Testing**

Previous Testing

I have previously done 2 rounds of PTR testing.

1st Test
10%+ loss of DPS, but flawed due to latency and bugged GCD.

2nd Test
http://forums.wow-europe.com/thread.html?topicId=1381118219&sid=1

Much more accurate, showing a 7.8% DPS loss.

New Testing

3rd Test
My longest and most conclusive test yet. I nuked for 115-second bursts (sometimes a bit shorter), added the results together and averaged. No Bloodlust, elemental mastery or trinkets abilities used. ToW and WoA totem up.

Gear was as I am wearing now, but T4 instead of T5 shoulders.

/Stopcasting and Quartz was used on the live realms, and only Lightning Bolt was cast.

Live
Test runs: 7
Total test time: 764 seconds
DPS: 1041.7
Crit rate: 35.4%
Castes per minute: 31.3

PTR
Test runs: 12
Total test time: 1219 seconds
DPS: 976.8
Crit rate: 37.2%
Castes per minute: 34.4

Conclusion: A loss of 6.2% of DPS

Its worth noting two points which may make this even worse:

i) The new "in-built" stopcasting system on the PTR really is working well now, and my gut feeling is that its superior to Quartz, and may be buffing my DPS, by letting me cast more.
ii) My crit rate was lower on the live realm, so the DPS loss should probably be bigger, if crit rates were equal.

I have tested for some time, and this is my third round of testing, so I have ironed out the kinks in my methods. This should therefore be reliable.

Other Testing

Testing by Niamee - respected regular of the Shaman forums. http://files.filefront.com/damtwmv/;8885685;/fileinfo.html

Shows a much, much worse nerf to DPS - this is partially to be expected as this is a scaling nerf (see below) and the gear is a lot better than mine :P

**Why We Are Unhappy**

1. The Loss of Lightning Bolt DPS
The changes to Lightning Bolt and the Lightning Mastery talent mean that our spell damage coefficient goes from 0.857 to 0.714. Or, in plain English... if you had +1000 spell damage, you will get +714 damage instead of +857 (ie a +143 damage nerf on each lightning bolt).

Even worse, because the concussion talent (+5% damage) is added after bonus spell damage, we get hit a bit more.

So adding all this up - a shaman with +1000 spell damage will go from hitting for 1533 per bolt, to 1383 per bolt. The damage loss for crits is obviously twice as bad, and Ele shamans tend to crit a lot.

Also, this nerf directly hurts our scaling, so the better your gear, the more it hurts. Its entirely possible that BT/MH Tier 6 geared Ele Shamans will lose 10%+ of their DPS - and thats just awful. (The Tier 6 set bonus also has another Concussion-style +5% dmg ability in it, which means the super well geared will see it a little worse still.)

2. The Loss of Overall DPS
The morning after the 2.3 changes were announced, Shamans were theorycrafting, and the majority showed a DPS nerf.

Its been clear to most of the community - literally since day 1 of this - that Lightning Overload would not add enough DPS to turn this into the buff we hoped for. The new version of Lightning Overload simply doesn't make back the damage we have lost on lightning bolt.

Hence we are looking at the sort of test results I am showing above (and wondering just how the developers could ever have come to see this as a buff - see below).

3. Lightning Capacitor Changes
Yes, it probably was a little overpowered - BUT we were kind of relying on it. The changed 2.3 version of the Capacitor is still good, but the loss of DPS from its new cooldown comes at a bad time for us.

Yes, it was a crutch, but now is a bad time to kick it away.

4. Loss of raid viability
Elemental shamans make the majority of their contribution through pure DPS. Our group buffs are not great (e.g. ToW) or can be done by another specc (e.g. Bloodlust). We don't want to hit like a lock, mage or rogue - but unless you improve our group buffs, we need good DPS to be worth a raid spot.

5. The loss of damage hurts PVP too.
The changes to Ele make Lightning Overload seem essential to regain some of the damage we lost, this is a direct nerf to 35/0/26 type PVP builds. By making us put more talent points into one tree to keep PVP viability it reduces our options, and our abilities as a hybrid.

Without going too off-topic, there are a lot of concerns already around our PVP ability (horrible vulnerability to counterspell locking out nature spells, lack of anti-cc - for example) if things go live as they are, they will make what some people already see as a bad situation worse.

6. Communication
I think this post sums it up very well:
http://forums.worldofwarcraft.com/thread.html?topicId=2765929939&sid=1

To summarise, our communication issues are:

a) Patch 2.3 was meant to be a buff to DPS - what happened?
How has your internal testing perceived this to be a buff to DPS? Its OK to talk about removing PTR bugs, but even the theory suggests that its a nerf.

What conditions was the testing done in? Can you admit a mistake was made? Or was it meant to be a nerf - in which case why??

If its not meant to be a nerf, can we have any reassurance that it will be fixed to a buff before 2.3 goes live?

b) We have had very little CM response, especially on the EU forums. Even a few "this is being looked at/passed on" posts would have been good, but we have been having to go to the US forums just for that.

**TLDR version**

Elemental shamans are losing a load of DPS, it will hurt a lot in PVE and PVP. We feel genuinely confused - and increasingly angry - from our communication with Blizzard, particularly that this was meant to be a buff.

I urge all shamans who would like to feedback here to:

a) Give your suggestions for anything I have missed
b) Keep this thread alive until we get some feedback (or hell freezes over, whatever).

Tharfor

#50 - Nov. 6, 2007, 6 p.m.

Hi there,

I’ve taken a little time out of moderation this evening to do a little theory-crafting of my own. I would like to say that I don’t normally do this sort of thing so I may have missed something fundamental which puts my theory out of the window completely; if you can point that out to me I will take it on-board. :)

Lightning bolt (Rank 12) has a cast time of 2.5 seconds and deals 563 to 643 damage.
(http://www.wowhead.com/?spell=25449)
The average damage of Lightning Bolt therefore is 603 (563+643 divided by 2); let’s call this 600 for talking sake.

The old Lightning Overload talent, with 5 points spent, gave a 5% chance to duplicate the spell cast exactly, right?
So in every 100 casts, 5 of them would be duplicated.
100 casts of a 600 damage spell = 60,000 damage
5 ‘procs’ in that 100 casts would deal an additional 600 damage, so 5*600 = 3000.
The total damage dealt in 100 casts would be 63,000 damage.
100 casts of a 2.5 second spell = 250 seconds.
63000 damage / 250 seconds = 252 damage per second.

The new Lightning Overload talent, with 5 points spent, will give a 20% chance to deal 50% damage again:
So in every 100 casts, 20 of them would do half damage (300 average).
20*300 = 6000.
The total damage dealt in 100 casts with this new proc rate is 66,000 damage.
66000 damage / 250 seconds = 264 damage per second.
264 / 252 = 1.0476 – a.k.a 104.76% or an increase of 4.76%.

This obviously ignores crit chance and spell damage gear and only looks at the Lightning Bolt in its’ purest form, but to me it seems clear that the base DPS is mathematically higher with the changes.

Chaos theory of course dictates that these numbers would never actually happen like this so let’s look at it pessimistically:

Under the old system, assume maximum damage dealt each cast (643).
100*643 = 64300
5*643 = 3215
Total Damage = 67,515
Damage per second = 270.06

Under the new system, assume minimum damage dealt each cast (563).
100*563 = 56300
20*281.5 = 5630
Total Damage = 61,930
Damage per second = 247.72
247.72 / 270.06 = 0.9173 – a.k.a. 91.73% which shows a DPS drop of 8.27%.

This proves to me that because each spell does not have a fixed damage value, but instead has a damage range, that there will always be an error margin; for Lightning Bolt rank 12 that appears to 8.27% variance...

So please, consider this dialogue open; however I promise nothing more at this time than that I will read what you have to say.

Tharfor

#65 - Nov. 6, 2007, 6:49 p.m.

Ok, here’s a quick number crunch concerning spell damage gear; again, assume that the average base damage is 600:

With a 2.5 second cast time (base), the spell damage contribution is 71.43%. Let’s use Mmootimus’ Armory profile (since this is his thread) and take the spell damage bonus of 865.
0.7143*865 = 617.8695 added damage per cast.
So each cast at 2.5 seconds is now 1217.8695 damage.
100 casts = 12,1786.95 and 250 seconds = 487.1478 damage per second.

With a 3.0 second cast time (base), the spell damage contribution is 85.71%.
0.8571*865 = 741.3915 added damage per cast.
So each cast at 3.0 seconds is now 1341.3915 damage.
100 casts = 13,4139.15 and 300 seconds (100 casts at 3 seconds) = 447.1305 damage per second.

487.1478 / 447.1305 = 1.0895 – a.k.a 108.95% or an increase of 8.95%.

This does not take into account casting time reductions from talents, only base casting time on Live and PTR; using the above numbers but with less total casting time we get:

Q u o t e:
So each cast at 2.5 seconds is now 1217.8695 damage.

100 casts = 12,1786.95 and 200 seconds (100 casts at 2.0 seconds) = 608.93475 damage per second.

and

Q u o t e:
So each cast at 3.0 seconds is now 1341.3915 damage.

100 casts = 13,4139.15 and 250 seconds (100 casts at 2.5 seconds) = 536.5566 damage per second.

Resulting in a 13.49% damage increase when you reduce your cast time through talents.

Tharfor

#80 - Nov. 6, 2007, 7:22 p.m.

Q u o t e:
Have it works now:
Because of the base cast time now of 3sec you get 3/3.5 = 85.7% benefit from spell dmg
You cast 100 LB of 600 dmg + 857dmg = 145,700dmg
5 procs in that = 5*(600+857) = 7,285 dmg
Total 152,985 divided by 200 sec (100*2sec) = 765 dps

After the changes:
Because of the new base cast time of 2.5 sec you get 2.5/3 = 71,4% benefit from spell dmg
You cast 100 LB of 600 dmg + 714 dmg = 131,400 dmg
20 procs in that = 20*(600+714)*50% = 13,140 dmg
Total 144,540 divided by 200 sec (100*2sec) = 723 dps

723/765 = 0.9451 a.k.a a decrease of 5.49%

I stand corrected.

I just ran the numbers again myself with the corrected oversight; this time including both Lightning Overload and the Spell Damage and I came to the same conclusion.

Don't hate me for doubting you, it just took my old brain a little while to catch up with you. :) Plus I haven't done any real calculations since high school!

So we agree that the changes will buff lower geared (presumably lower level mostly) Shaman, but as the bonus spell damage increases, the damage under the new system will decrease?

Tharfor

#96 - Nov. 6, 2007, 7:42 p.m.

Well my next point would be:
If we don't factor in Spell Damage gear, it appears to boost damage output.
If we factor in 1000+ Damage, it drops damage by 5%.

At what value of +Damage Gear does the damage 'boost' reach 100% and, more importantly, what can you be looking for on your items instead of +damage?

I'm thinking +Haste might be a factor here...

My original post used no +damage and showed an increase of about 5%, using +1000 damage gear we showed a decrease of about 5%.

I would hazard a guess that with +500 damage gear, the damage would be roughly equal to 'normal'.

And with +haste to bring the overall casting time of 100 bolts down to 150 seconds rather than 200 seconds I suspect that the damage would start to go up again.

I haven't ran these numbers myself yet as I have to get some moderation done whilst I'm posting here and I've built up a small backlog :)

I shall return to this thread when I can, but don't hate me if that turns out to be tomorrow. :(

Tharfor

#231 - Nov. 7, 2007, 7:16 p.m.

Firstly, regarding...

Q u o t e:
i admire you for this acknowledgment and standing corrected

...and many similar quotes. I would like to say that I do not think that it takes courage to admit when one is wrong, but rather it takes sheer stupidity to stick to your opinions when they have been proven wrong. It's nice that you are nice to me, but I haven't actually done anything yet :)

A lot of the time we Community folks don't have much more information than what you have, simply because we tell you everything that we do know. Often, you guys will know more than we do because you have time to test it on the PTRs - when I'm working I don't have that luxury, and when I get home at night I quite selfishly play the game for myself and my guild's progression. :)

Q u o t e:
the BURST damage nerf is even worse

I see a lot of discussion on the forums and one thing that has been made more and more clear to me is that each class has issues that they feel are a terrible injustice to them.

For example, we have people saying that one spec that they use for PvP is no good for PvE and if they spec for PvE, they're not as good at PvP. This may seem obvious to me, but for someone so deeply invested in their character and dedicated to every change that has happened to it since launch, it can be very easy to forget about the bigger picture and get stuck on the numbers.

When a screw in your clock starts to rattle, you can tighten it. However, a rattling screw does not affect the function of the clock; it will still tell you the time. If the person who made the screw knew that it was rattling, he might care a little more about it than the clockmaker who will have to dismantle and reassemble the clock for such a small issue.
(See I can do philosophy as well as mathematics!)

Maybe the burst damage was the reason for the change (note: "maybe", I don't read minds and reasons for changes are not always communicated) and maybe you have another talent configuration that will be 'better' for PvP and burst damage...
"...but why should I be bound to one talent configuration for each role?! I am a hybrid?!" I hear you cry!
Well, if you could do everything regardless of your talent points, why would you need to have talent points in the first place...?

Q u o t e:
To Tharfor (or Salthem, or any other CMs) - I hope you are starting to see why this is a bad nerf for us. I hope the theory is clear now. If not I will explain until it is.

I have to admit, doubt does enter my mind on a regular basis as I read these forums. I have rather a lot of blind faith in the development of this game simply because no matter how much people have screamed and shouted about some of the things in the past, it's always turned out fine according to the 'top-secret-master-plan'.

However, the fact that I have these doubts does not mean I am an unreasonable person. In fact, it would have been very easy for me just to make my original post something like "Blue tagging this because it's a good discussion thread. Carry on."

Back on topic:
I would like to repeat one of my original comments in this thread: "I promise nothing more at this time than that I will read what you have to say".
I wish I could tell you that I would hand deliver your concerns to the development office and not leave until they are acknowledged, but that's just not how things work I'm afraid. :(

I am compiling your points, as I have done for other classes in the past, but that is all I can say for certain at this time.