 |
05/28/09, 10:00 AM
|
#826
|
|
Aloof Aggravator
Sutiru
Undead Warrior
No WoW Account
|
Originally Posted by Laraque
That is of course, assuming your statistical theory is accurate.
The writers of those simulators that were used to generate the stat weights will be the first to admit they are not perfect. For example: the way it handles misses, dodges, & parry is entirely an approximation. What they do is assume is a perfect rotation is used and everything hits every time; then multiply the damage by a reduction factor to simulate that some missed (a character with a 4% miss rate would multiple his damage by 0.96 to simulate 4% being misses). This de-values +hit & +exp stats because it does not take into account things like a messed up rotation because of it.
I am not saying the simulators are bad, in fact they are very useful and powerful. But rather that the data obtained from them is not perfect and should be taken as a good starting guideline and not strict rules of how to min/max your character because of it.
|
That approximation sounds less like how a simulator works (or should work in my mind) and more like the kind of approximation a spreadsheet or calculator uses. I would be extremely surprised if any of the premier simulators seen on these forums make such approximations, as cutting that kind of corner effectively defeats the purpose of simulators.
Now, I wouldn't be surprised if the simulators mindlessly continued the rotation after a miss or dodge. Implementing intelligent on the fly modification of rotations is nothing short of excruciating.
|
Originally Posted by Vectivus
... you could very well have a concerto, but the closest the average listener gets to hearing it is the interpretation as put on by a group of small children with those little rainbow-coloured xylophones.
|
Monte's LoL Blog
Monte's LoL Stream
|
|
|
05/28/09, 10:11 AM
|
#827
|
|
Don Flamenco
Draenei Death Knight
Chants Eternels (EU)
|
That one of the advantage of using a simulator vs a "simple" spreadsheet. My simulator, in case of miss or dodge, will adapt to this unexpected event and act according to the rotation or priority set up.
|
|
|
|
|
05/28/09, 10:48 AM
|
#828
|
|
Glass Joe
Draenei Death Knight
Silver Hand
|
Originally Posted by Darian_TruBlade
That approximation sounds less like how a simulator works (or should work in my mind) and more like the kind of approximation a spreadsheet or calculator uses. I would be extremely surprised if any of the premier simulators seen on these forums make such approximations, as cutting that kind of corner effectively defeats the purpose of simulators.
Now, I wouldn't be surprised if the simulators mindlessly continued the rotation after a miss or dodge. Implementing intelligent on the fly modification of rotations is nothing short of excruciating.
|
I believe most of the stat weights people use were derived by Methods. Please read his post here:
There Will Be Blood (DPS) - Meet the new thread/Same as the old thread
Edit: I just want to be clear that I am not criticizing Method's work, I am very appreciative for what he does for the community. I use his numbers myself for stat weights myself, although I do make expertise=str as my only change. He does good work, but all simulations or spreadsheets are just approximations and should never be taken as be-all-end-all data to use, IMO.
Last edited by Laraque : 05/28/09 at 10:54 AM.
|
|
|
|
|
05/28/09, 1:53 PM
|
#829
|
|
Don Flamenco
|
Originally Posted by Laraque
all simulations or spreadsheets are just approximations and should never be taken as be-all-end-all data to use, IMO.
|
It's important to remember why this is so. A properly coded simulation run for a long enough duration is the be-all-end-all on average DPS. Average DPS, however, is not the only metric that's important to in game raiding. It's not that a good sim doesn't account for everything that is necessary to maximize DPS, if it didn't it wouldn't be a very good sim, it is that maximizing DPS is not all that matters.
|
My vanity is justified.
|
|
|
05/28/09, 3:41 PM
|
#830
|
|
Piston Honda
Orc Death Knight
Dethecus
|
Originally Posted by tetracycloide
It's important to remember why this is so. A properly coded simulation run for a long enough duration is the be-all-end-all on average DPS. Average DPS, however, is not the only metric that's important to in game raiding. It's not that a good sim doesn't account for everything that is necessary to maximize DPS, if it didn't it wouldn't be a very good sim, it is that maximizing DPS is not all that matters.
|
While all this sim talk is interesting, it's not really any more relevant to blood dks than basically any other class/spec. You need to take from the sims and modeling what you want and go from there. However the last point you made here is an excellent on. Fights like XT-Hardmode, and Hodir HM require timing and burst that goes beyond what sims are currently designed to show. It would be an interesting addition to show things like max burst dps potential for a 5-10 second spread.
You can be #1 overall DPS on XT, but if your burst sucks you might still not make it to hardmode.
Coordinated bursts is also one of the pinnacles of Blood spec, monitoring raid buffs, trinket procs, available runes, RP, and getting them all to meet up right at one point can be difficult, but certainly rewarding when done correctly.
|
|
|
|
|
05/28/09, 4:05 PM
|
#831
|
|
Glass Joe
Gnome Death Knight
Zul'Jin
|
Ok..... Shifting gears from the exp/str and simulator debate since it seems to be boiling down to a personal preference anyways...
Why not discuss how to handle misses in the rotation? Aside from the lucky few who are fully hit & expertise capped, this is a scenario that we all go through, esp those with lower lvl gear. What are some strategies for handling miss/dodges in the rotation? This probably will go a longer way to improving your play than worrying about min/maxing red gems.
|
Now, I wouldn't be surprised if the simulators mindlessly continued the rotation after a miss or dodge. Implementing intelligent on the fly modification of rotations is nothing short of excruciating
|
I tend to favor not overbaking the cake on missed strikes - excruciating is not a word that comes to mind. I'm spec'd 2/2 into Epidemic and use a IT-PS-DS-HS-HS-DCdump-DS-HSx6-DCdump rotation which gives me more flexibility to handle misses. Since your runes are returned on a missed/dodged attack, I nearly always retry on missed strikes. I split my rotation into three phases, and then use DC dump time to adjust for errors:
1) IT-PS-DS---DS - is critical to set up diseases and death runes. If any of these attacks misses I retry them.
2) HS-HS - the first two heart strikes have no dependencies later in the rotation, so if I miss one, I just skip it.
3) HSx6 - When 4 death runes are up, I try to use them all. If there has been more than one miss in the rotation, the last HS may be missing one or both diseases. One diseaseless HS at the end of a rotation provides more dmg than DC, IT, or PS. (I think this holds even with the improved PS and vengeful heart-morbidity-dark death DC). I've been experimenting with this a bit lately, however, and would be interested if someone has a mathematical approach to answer whether the diseaseless strike is a net gain or loss compared with skipping it and immediately refreshing diseases. I feel it will depend on the length of the combat (ie. total number of rotations).
In movement heavy fights, where I'm not truly missing so much as out of range, the chance to miss strikes is significantly higher. In these fights disease ticks form a larger percentage of dmg so priority shifts away from using death runes in phase 3 to maintaining diseases at all times. This also simplifies timing of the rotation when you're busy worrying about movement and survival.
In summary, I retry if a missed strike has dependencies later in the rotation, while misses without dependencies are generally skipped. Any other strategies to speak of?
|
|
|
|
|
05/28/09, 8:09 PM
|
#832
|
|
Aloof Aggravator
Sutiru
Undead Warrior
No WoW Account
|

Originally Posted by Canadianloaf
I tend to favor not overbaking the cake on missed strikes - excruciating is not a word that comes to mind. I'm spec'd 2/2 into Epidemic and use a IT-PS-DS-HS-HS-DCdump-DS-HSx6-DCdump rotation which gives me more flexibility to handle misses. Since your runes are returned on a missed/dodged attack, I nearly always retry on missed strikes. I split my rotation into three phases, and then use DC dump time to adjust for errors:
1) IT-PS-DS---DS - is critical to set up diseases and death runes. If any of these attacks misses I retry them.
2) HS-HS - the first two heart strikes have no dependencies later in the rotation, so if I miss one, I just skip it.
3) HSx6 - When 4 death runes are up, I try to use them all. If there has been more than one miss in the rotation, the last HS may be missing one or both diseases. One diseaseless HS at the end of a rotation provides more dmg than DC, IT, or PS. (I think this holds even with the improved PS and vengeful heart-morbidity-dark death DC). I've been experimenting with this a bit lately, however, and would be interested if someone has a mathematical approach to answer whether the diseaseless strike is a net gain or loss compared with skipping it and immediately refreshing diseases. I feel it will depend on the length of the combat (ie. total number of rotations).
|
First, your rotation seems off. In fact, it seems fairly impossible without introducing significant wait times between rune cooldowns and complications from Blood Rune versus Death Rune timing. It looks like you're attempting to combine the old diseaseless rotation with the reality of needing diseases, and from my own experience I can't say that works very well.
Second, depending on the implementation of the simulator it can very well be excruciating to take simple player logic and translate it into something the simulator can act upon. Some priority based methods may not need any modification, but simulators based on using a set rotation would need extensive retooling in order to properly act upon a miss. It's not a matter of coding the specific logic, but of implementing functionality to allow players to input that logic themselves in the same way they do their rotations. It's not impossible, it's just a pain.
|
Originally Posted by Vectivus
... you could very well have a concerto, but the closest the average listener gets to hearing it is the interpretation as put on by a group of small children with those little rainbow-coloured xylophones.
|
Monte's LoL Blog
Monte's LoL Stream
|
|
|
05/29/09, 3:13 AM
|
#833
|
|
Von Kaiser
|
Originally Posted by Soilantgreen64
While all this sim talk is interesting, it's not really any more relevant to blood dks than basically any other class/spec. You need to take from the sims and modeling what you want and go from there. However the last point you made here is an excellent on. Fights like XT-Hardmode, and Hodir HM require timing and burst that goes beyond what sims are currently designed to show. It would be an interesting addition to show things like max burst dps potential for a 5-10 second spread.
You can be #1 overall DPS on XT, but if your burst sucks you might still not make it to hardmode.
Coordinated bursts is also one of the pinnacles of Blood spec, monitoring raid buffs, trinket procs, available runes, RP, and getting them all to meet up right at one point can be difficult, but certainly rewarding when done correctly.
|
We were attempting hardmode XT this week, I was bursting(and sustaining) around 8.9k-9.4k during the heart phase rather consistently. Though I noticed if I popped my army prior to it a bit I would always be low, regardless what my rune status was at which doesn't make any sense to me at all. Maybe I just got an unlucky string of crits? Who knows. Whenever we would get heart phase down, my personal damage would be anywhere from 395k~498k on attempts sans drw being included. I'll have to find my wws's to link what I was doing wrong. Though I will admit, having the burst capability I have was definitely needed for the fight.
Which actually lead me to a question myself. This thread is riddled with a ton of varying opinions on stats, specs and the like, but i'm still not sure of what stat or at what point armorpen is a primary stat to seek without crippling your strength. Would it be alright once i've gleened enough info from this thread that the op or maybe someone else could just re-make this thread so we have a setup more similar to the frost or unholy threads? This one has been without any kind of representation for eons now.
|
|
|
|
|
05/29/09, 5:37 AM
|
#834
|
|
Don Flamenco
Draenei Death Knight
Chants Eternels (EU)
|
Originally Posted by Darian_TruBlade
Second, depending on the implementation of the simulator it can very well be excruciating to take simple player logic and translate it into something the simulator can act upon. Some priority based methods may not need any modification, but simulators based on using a set rotation would need extensive retooling in order to properly act upon a miss. It's not a matter of coding the specific logic, but of implementing functionality to allow players to input that logic themselves in the same way they do their rotations. It's not impossible, it's just a pain.
|
I don't want to advertise for my simulator but on rotation mode, when you set up your rotation you have to set what the simulator have to do in case of miss or dogde.
Sample:
<Rotation>
<IcyTouch retry='1'></IcyTouch>
<BloodStrike retry='1'></BloodStrike>
<FrostStrike retry='0'></FrostStrike>
<Obliterate retry='1'></Obliterate>
<Obliterate retry='1'></Obliterate>
<FrostStrike retry='0'></FrostStrike>
<FrostStrike retry='0'></FrostStrike>
</Rotation>
|
|
|
|
|
05/29/09, 9:56 AM
|
#835
|
|
Piston Honda
Orc Death Knight
Dethecus
|

Originally Posted by Veritas17
We were attempting hardmode XT this week, I was bursting(and sustaining) around 8.9k-9.4k during the heart phase rather consistently. Though I noticed if I popped my army prior to it a bit I would always be low, regardless what my rune status was at which doesn't make any sense to me at all. Maybe I just got an unlucky string of crits? Who knows. Whenever we would get heart phase down, my personal damage would be anywhere from 395k~498k on attempts sans drw being included. I'll have to find my wws's to link what I was doing wrong. Though I will admit, having the burst capability I have was definitely needed for the fight.
Which actually lead me to a question myself. This thread is riddled with a ton of varying opinions on stats, specs and the like, but i'm still not sure of what stat or at what point armorpen is a primary stat to seek without crippling your strength. Would it be alright once i've gleened enough info from this thread that the op or maybe someone else could just re-make this thread so we have a setup more similar to the frost or unholy threads? This one has been without any kind of representation for eons now.
|
I am currently working on a complete rewrite of this thread, following Darkside's format for the frost thread. It is a pretty time-consuming process, and I want to get as much in it as possible. I hope to finish in the next few days.
As far as XT goes, your burst should probably be a little higher through the heart phase. I did 8.9k on XT this week Wow Web Stats, we did a pretty quick kill, but didn't do hardmode. I'll try to come up with a custom split for just the heart phase if I can.
Ok, for example, our last XT my burst through the first heart phase was 23k, doing 655k dmg (approx. 552k without DRW), Wow Web Stats
Doing custom splits like this can really be illuminating for where you can increase your DPS. For example, looking at my own performance, I started summoning my army way too late, and was still summoning 2-3 seconds into the fight, and didn't land my first melee strike until 4 seconds in. Thats losing out on almost 15% of my DPS time.
Last edited by Soilantgreen64 : 05/29/09 at 10:27 AM.
|
|
|
|
|
05/29/09, 2:21 PM
|
#836
|
|
Von Kaiser
|
Ah see that's where I was having my problem was with summoning into the first few seconds of the heart phase or not engaging it trying to set up all my buffs (I didn't have a blood macro set up yesterday until way, way late) so I kept missing out on a lot of precious dps time. My DRW was doing 95-98k so the timing definitly lends to it. Though I wonder if any of my loss had to do with lacking hit or anything of that nature. I sit at 263 so I figured I SHOULD be ok.
Would a section for the new post called burst segments or the best places to know when/how to use them be hand holding too much? I'd like to contribute to that if you would as I know there are a lot of people I see dpsing who just don't understand when to use their cooldowns. That and a macros section maybe?
|
|
|
|
|
05/29/09, 2:29 PM
|
#837
|
|
Glass Joe
Blood Elf Death Knight
Lightninghoof
|
Is crit rating just not important anymore with blood spec?
|
|
|
|
|
05/29/09, 4:39 PM
|
#838
|
|
Glass Joe
Orc Death Knight
Archimonde
|
Has anyone else with 4 pc done any extensive testing for OB vs. DS ?
|
|
|
|
|
05/29/09, 4:51 PM
|
#839
|
|
Von Kaiser
Dwarf Death Knight
Lightbringer
|
It's still nice to have crit. It's just further down on the priority list, worry more about strength, armor pen, hit, and expertise. Think of crit and haste as nice filler when you can't get the others previously listed.
|
|
|
|
|
05/30/09, 6:48 PM
|
#840
|
|
Glass Joe
Orc Death Knight
Eonar (EU)
|
Hi all,
I have been reading EJ for over a year now and it has always served me well. I have spent an absolute fortune on respecs between blood and unholy, and as much as i really want to go blood, i cant match the dps i do in unholy. I have read the posts in this forum over and over and I have tried the rotations....
IT>PS>HS>HS>DS>DS>DC>DC>HS>HS>HS>HS>HS>HS
Altlhough my rotation ends up on CD so i do IT>PS>HS>HS>DS>DC>DS>HS>HS>HS>HS>HS restart
I have taken stats of my last 10 man ulduar and i am hoping you guys with the knowledge can point out to me where i am going wrong.
Wow Web Stats
The World of Warcraft Armory
I realy do apologise in advance if there is something blatantly obvious.
Last edited by Nightorc : 05/31/09 at 4:24 AM.
|
|
|
|
|
|