Two days of ToT logs on Raidbots, have at it

100 Human Mage
15015
The data we have suggests mages are probably low. But the sample sizes are way too small to prove it.

If you look at samples, only Jin'rokh has enough samples to really be worth discussing. Mages are low there. But T15 has 11 other bosses.


Blizzard claims mages are supposed to be strongest when they're stationary and Jin'rokh is the closest we have to a Patchwerk fight in this tier. Regardless, I'm giving it a couple of weeks to get a solid sample size before I judge class balance.


Very much this. Normally something like this wouldn't concern me, but when I heard the nature of that fight relative to the other raid bosses..

Well, yeah.
Reply Quote
100 Gnome Mage
18295
Unfortunately, waiting 2 weeks is not going to alter things that much. Between ptr testing, simcraft and now raidbots...How much more data is needed to show the obvious?

Blizzard over did the balancing for mages by alot. No amount of hand wringing or people waiting to see if things are going to change is going to alter that fact.

Its like dealing with Republicans and climate change. Reams upon reams of data stating that there is a problem. Response? the data is wrong...the way it was collected was wrong...the tools were wrong etc etc. Does not change the facts. As a result we suffer till they figure it out and by then
it will be to late.
Reply Quote
100 Draenei Shaman
5230
Unfortunately, waiting 2 weeks is not going to alter things that much. Between ptr testing, simcraft and now raidbots...How much more data is needed to show the obvious?

2 weeks of raidbots is needed to prove it.

But looking at mage performance on Jin'rokh normal modes, I agree that if I was a dev I would already be doing some due diligence.
Edited by Slant on 3/7/2013 1:29 PM PST
Reply Quote
It paints a picture. It's a blurry watercolor viewed through frosted glass... but it paints a picture.
Reply Quote
90 Human Mage
14040
Unfortunately, waiting 2 weeks is not going to alter things that much. Between ptr testing, simcraft and now raidbots...How much more data is needed to show the obvious?

Blizzard over did the balancing for mages by alot. No amount of hand wringing or people waiting to see if things are going to change is going to alter that fact.

Its like dealing with Republicans and climate change. Reams upon reams of data stating that there is a problem. Response? the data is wrong...the way it was collected was wrong...the tools were wrong etc etc. Does not change the facts. As a result we suffer till they figure it out and by then
it will be to late.


You flatter Blizzard with the word balancing. They are incapable of it, the rest of post I couldn't agree more with. This is a systemic problem Blizzard has. What can you do?
Reply Quote
90 Gnome Priest
11895
Just fyi, blizzard took in exactly two days of data after 5.1 dropped before nerfing the crap out of Fire.

That was also mid-tier, so there was quite a bit of data already that could show Fire doing too much. New tier has pretty much none of that, and with the magnitude of some of the other changes, it's hard to judge how things will pan out.
Reply Quote
100 Troll Mage
17020
03/07/2013 02:02 PMPosted by Skootalloo
Just fyi, blizzard took in exactly two days of data after 5.1 dropped before nerfing the crap out of Fire.

That was also mid-tier, so there was quite a bit of data already that could show Fire doing too much. New tier has pretty much none of that, and with the magnitude of some of the other changes, it's hard to judge how things will pan out.


Just to reiterate this point, people are still learning these fights.
Reply Quote
03/07/2013 02:06 PMPosted by Hiroran

That was also mid-tier, so there was quite a bit of data already that could show Fire doing too much. New tier has pretty much none of that, and with the magnitude of some of the other changes, it's hard to judge how things will pan out.


Just to reiterate this point, people are still learning these fights.

The first boss is as close to patchwerk as you can get, really. Not much movement.
Reply Quote
100 Goblin Warlock
18265
1) Most guilds do not post logs in the first few weeks of a raid

2) It hasn't even been 1 week

3) People are still learning fights.
Reply Quote
100 Night Elf Druid
11740
People are still doing the old fights. You could check logs from what you know to be good DPS checks in previous tiers. Also I think your better off deciding which logs or fights to look at than just trusting spec score. Good concept but without a detailed explanation of exactly how it was computed as opposed to the general explanation given it can't be trusted.
Reply Quote
100 Draenei Shaman
5230
There is a detailed explanation. It's not difficult to find, if you bother to look.

http://seriallos.wordpress.com/2011/04/13/413-spec-score-how-does-it-work/
Edited by Slant on 3/7/2013 2:33 PM PST
Reply Quote
100 Blood Elf Mage
18255
There is a detailed explanation. It's not difficult to find, if you bother to look.

http://seriallos.wordpress.com/2011/04/13/413-spec-score-how-does-it-work/


Ok, let's forget the "overall dps" setting since you don't like it.

Let's go straight to the actual damage done for individual fights.

Here is Jin'rokh the Breaker, nearly a patchwerk encounter:

10N: http://www.raidbots.com/dpsbot/Jin'rokh_the_Breaker/10N/all/7/60/default/
25N: http://www.raidbots.com/dpsbot/Jin'rokh_the_Breaker/25N/all/7/60/default/
Reply Quote
100 Draenei Mage
19910
Nobody knows how the devs' minds work, but it's safe to say they do internal testing and didn't nerf based upon data from 2 days of raiding. Because that would be crazy.


Internal testing should have caught the combustion changes rolling into 5.1. If there was any testing at all, the person in charge of looking at the results was completely oblivious.
Reply Quote
100 Blood Elf Mage
18255
03/07/2013 02:02 PMPosted by Skootalloo
Just fyi, blizzard took in exactly two days of data after 5.1 dropped before nerfing the crap out of Fire.

That was also mid-tier, so there was quite a bit of data already that could show Fire doing too much. New tier has pretty much none of that, and with the magnitude of some of the other changes, it's hard to judge how things will pan out.


This isn't really accurate. Before 5.1 went live, several blue posts and at least one tweet from GC said merely that they wanted to tone down Fire burst in PvP, specifically because of alleged multiple shattered Pyros.

As such, they briefly proposed an ill-advised three second cooldown on Pyro. One of the problems with this cooldown is that it would have severely nerfed our mastery, which is a rolling DoT called Ignite. And moreover, ignite was a large part of our main cooldown, Combustion. So at about the same time, they made a PTR change to Combustion, getting rid of the Pyro DoT contribution to Combution and raising Ignite's contribution to Combution from 50% to 100%. One of the blue posters specifically said the intent was to keep Combustion where it was before removing the Pyro DoT contribution.

Meanwhile, they realized how insane the Pyro cooldown was and got rid of it. Then we didn't hear another thing. No complaints about Fire being OP on the PTR. No additional proposed changes or nerfs in patch notes. Nothing.

5.1 goes live, and oops, the change to Combustion ends up being an accidental buff. But instead of reverting the changes, after two days, they nerf combustion to be merely 50% ignite, *and* they nerf critical mass to 1.25. The critical mass nerf was never brought up, mentioned, discussed, and certainly wasn't deployed on the PTR. Again, the only discussion regarding fire from the developers, GC included, were tweets about PvP.
Edited by Taymage on 3/7/2013 4:16 PM PST
Reply Quote
90 Undead Mage
11870
2 weeks of raidbots is needed to prove it.


To who? You can't use this data for proof to blizzard. They danced around the 5.1 fire nerfs repeatedly saying "bad data" "unsupported claims" "we're not seeing that".

I know that not ever person who raids submits logs, but 20k parses in 2 weeks in not what I would call insufficient data. There are only so many variables in this game - 20k iterations is still 2x the recommended "norm" for simC - but that's somehow not enough?

We aren't going to prove anything to anyone over at Blizzard.

If you're talking about proving it to the rest of the community, you're still fighting an uphill battle where the loudest and most misinformed are the ones that say things like "You were top dps by a mile over every other spec in T14" and "you're all just mad because you don't have 3 of the top 4 specs anymore like you always have in every tier" and my personal favorite "shut up you needed to be brought down to mortal levels".
Reply Quote
88 Draenei Mage
6610
To who? You can't use this data for proof to blizzard. They danced around the 5.1 fire nerfs repeatedly saying "bad data" "unsupported claims" "we're not seeing that".


And once you've got 2 weeks, people want 4. Then 2 months. Then 6 months. Then it's "lol Raidbots is srs bzns' etc.
Reply Quote
1) Most guilds do not post logs in the first few weeks of a raid

2) It hasn't even been 1 week

3) People are still learning fights.


All three of those things would effect all specs equally.

1. would only show higher for mages if all the top guilds that are hiding their logs are stacked full of super good mages, but I have a feeling that the guild posting and the guilds not posting logs have a similar/representative number of mages.

2. Sure, its been one week but in those data sets there are about 7,000 mages parses. 7000 is a significant number-- its not like those two days amount to 12 total mages posting logs.

3. This would only skew mages low on damage if there were some factor about mages that specifically made their DPS while learning the fights significantly more hindered than another class/spec.
If a Rogue and a Mage are learning the same fight and of similar skill level (and since he have 7,000ish parses per class we can assume similar aggregate skill levels) then both will show similar increases when the fight is fully learned. So while mage damage might go up, it will not go up significantly MORE than any other classes (who are already ahead of where we are).


And once you've got 2 weeks, people want 4. Then 2 months. Then 6 months. Then it's "lol Raidbots is srs bzns' etc.


And then it will be "Wait for (NAME OF EXPANSION AFTER PANDARIA) then hopefully mages will be fixed. . .
Edited by Alordis on 3/7/2013 4:32 PM PST
Reply Quote
100 Troll Mage
17020
03/07/2013 04:28 PMPosted by Rèquiem
To who? You can't use this data for proof to blizzard. They danced around the 5.1 fire nerfs repeatedly saying "bad data" "unsupported claims" "we're not seeing that".


And once you've got 2 weeks, people want 4. Then 2 months. Then 6 months. Then it's "lol Raidbots is srs bzns' etc.


Do you have a source for this because Slant has pretty much always stuck by a 2-3 week time period.
Reply Quote
MVP
90 Human Mage
10015
03/07/2013 04:28 PMPosted by Rèquiem
To who? You can't use this data for proof to blizzard. They danced around the 5.1 fire nerfs repeatedly saying "bad data" "unsupported claims" "we're not seeing that".


And once you've got 2 weeks, people want 4. Then 2 months. Then 6 months. Then it's "lol Raidbots is srs bzns' etc.


Raidbots is awful for this sort of thing (GC does a pretty good job of explaining why in the Totemspot interview), but I don't know that I've ever seen it be so thoroughly wrong as it would have to be in this case to explain the difference in internal and external perceptions.

I would be surprised if we don't hear more one way or the other by the middle of next week.
Reply Quote
90 Blood Elf Mage
6075
Raidbots is awful for this sort of thing (GC does a pretty good job of explaining why in the Totemspot interview), but I don't know that I've ever seen it be so thoroughly wrong as it would have to be in this case to explain the difference in internal and external perceptions.


I just read through it and I cannot see how sampling bias could effect all 3 specs simultaneously.
Nor the other points being relevant to all 3 being at the bottom.
Reply Quote

Please report any Code of Conduct violations, including:

Threats of violence. We take these seriously and will alert the proper authorities.

Posts containing personal information about other players. This includes physical addresses, e-mail addresses, phone numbers, and inappropriate photos and/or videos.

Harassing or discriminatory language. This will not be tolerated.

Forums Code of Conduct

Report Post # written by

Reason
Explain (256 characters max)

Reported!

[Close]